Box of Amazing

Ben Affleck is Human

Rahim Hirji — Sun, 19 Jul 2026 06:01:30 GMT

My book SuperSkills: The Seven Human Skills for the Age of AI is available at all good bookstores, through superskillsbook.com or on Amazon

I work with organisations and leadership teams on SuperSkills, human capability, and AI-era decision-making through speeches, workshops, keynotes and advisory work. In person or virtual. UK, Europe, global. → Book a discovery call | → Explore: thesuperskills.com

Friends,

I remember watching the clip and had forgotten it within an hour, in the way we scroll and no longer remember

Ben Affleck, across a table from Joe Rogan, talking about artificial intelligence and the film business. He was calm in a way that did not fit the argument I was used to hearing, with no panic in it and no salesmanship either. I nodded at something, closed the tab, and got on with my day. It left no mark.

A few weeks later there was a headline. Netflix had bought an AI company, and the name on it was his. The deal was reported to be worth as much as six hundred million dollars. This time I stopped, and not for the money. I stopped because the man I had filed under sceptic had spent three years building the very thing he seemed so unbothered by. I felt confused and conflicted in my head.

So I went back, and I stayed longer than I meant to. He has two Academy Awards. He wrote Good Will Hunting with Matt Damon when they were young and unknown, and they won for the screenplay, with Robin Williams saying their lines. Years later he directed Argo, the true story of a CIA officer slipping six Americans out of Tehran during the hostage crisis, and it won Best Picture. The Academy gave the film its highest prize and did not nominate Affleck for directing it, as though the film had directed itself. I have no interest in the tabloid version of him, the marriages and the paparazzi and whatever the magazines are selling this month. The other one is more useful. He is proven, he is unusually articulate, and when he talks about this he has done the reading. Whatever you make of him, and I will tell you what I make of him, a great many people know his face. I think that matters.

Netflix released a video when it announced the deal. Affleck is in it, describing his company, InterPositive, as a way to take out the logistical, difficult, technical work that gets in the way of making things: the wire you paint off a stunt, the shot you missed and go back for, the lighting you reshape after the fact. His phrase for what this gives you is the one that pulled me all the way in. “You get more human work.”

Underneath the same deal sits the other version. The business case, the part that justifies the price, is a machine that does the expensive, people-heavy parts of a production with far fewer people. It promises more content at lower cost with smaller crews. Even inside Netflix the language performs the move, with the company’s content chief telling him on camera that for her it was less about cheaper and more about better.

So there are two accounts of the same technology, and they do not agree. One promises more human work. The other delivers more work done by fewer humans. They come from the same company in the same week, and both are true.

I have started calling the second account the “patent underneath”. It is not a literal filing, though in this case one exists. It is the arithmetic that sits below the story a person tells about why they reach for the tool, the version you would hand to the accountant next to the version you say into the camera. Affleck’s is public, printed, and priced at up to six hundred million dollars. Yours is smaller, and no one has read it but you. You still have one.

Go back through the interviews and a shape appears. He is not selling the future and he is not mourning it.

He is dismissive of the fear, in a specific direction. He does not believe the machine is about to make a film from nothing, or write anything that matters. His most quoted line is that it can produce imitative verse that sounds Elizabethan, and that it cannot write you Shakespeare. He calls it a craftsman rather than an artist, able to sit beside a master and copy the technique, cross-pollinating what already exists without ever making anything new. He draws the line between the two in a single breath. Craft is knowing how to work, and art is knowing when to stop. Stopping, he says, is what the machine will struggle to learn, because stopping is taste.

And in the same breath, without softening it, the cost. He would not want to be in the visual effects business, he says, because they are in trouble, and he is right that they are. He gives the arithmetic almost in passing: instead of five hundred people in Singapore paid two dollars an hour to render a superhero film, the machine will do it faster. He says it the way you read out a line item, because in that moment that is what they are to him. What the tool takes out is the laborious, less creative, more expensive part of the work, which brings the cost down and lets more people make the films they always wanted to make. He is describing a smaller crew, and a smaller crew has a plain meaning for the people no longer on it. He is sharp, too, about the other side of the noise, the companies who talk up a world about to change in order to justify the money they have already spent, when history, he says, shows adoption coming slowly and in increments.

He is not choosing a side. He is standing on a position with a cost built into it and refusing to pretend the cost away.

There is a reason I went looking in acting and not somewhere tidier.

The film business runs on story, and story is the most human thing we have. It is how we worked out who we were before we could write it down, and it is still the test I use when almost nothing else moves me. I get to the cinema once or twice a year now, because I have become unbearable about it, unwilling to give three hours to anything without a real human pulse in it. One of those films this year will be The Odyssey, which I will see in the coming weeks. Christopher Nolan has taken Homer’s three-thousand-year-old poem about a man trying to get home and shot it entirely on IMAX film across six countries, in the most laborious, least automated way a film can still be made. And you know how much I love Chris Nolan movies.

Odysseus is played by Matt Damon. Affleck’s oldest friend, the man he wrote Good Will Hunting with before either of them was anyone, his partner in the production company they later built to put filmmakers first. So in the same season, one of them is selling an AI engine that strips the laborious work out of production, and the other has spent months on location being a man on a ten-year journey home. They wrote one of the most human films of their generation together, and now they stand at opposite ends of the same question. Neither of them is wrong.

While Damon has been marketing Odysseus, the rest of us watched the World Cup on our phones, and the World Cup this summer has been two tournaments at once. One is played on the pitch. The other is made in the feed, by AI.

Open any of it and the football has been turned into something else. Messi, Mbappé and Lamine Yamal are redrawn as anime warriors with glowing auras, fighting for the trophy across a storyline that fans rewrite after every match. Gianni Infantino, the head of FIFA, introduces the semifinalists like final bosses entering an arena. Neymar, ageless, dribbles through a match that never happened. Some of it is wonderful. None of it exists without a person deciding what would be funny, what would move you, which clip cuts to which, where the joke sits. The machine renders it. A human still has to have the idea, and the taste, and the timing. That is Affleck’s cutting room, moved into a teenager’s bedroom.

And the same feed, the same tool, carries the other thing. A video of Neymar gliding past Messi to score reached millions of people who took it for real, until someone checked and found no such match had been played. When Cape Verde held Spain, a story went round that their goalkeeper was a part-time electrician and bus driver moonlighting as a footballer, and people loved it, and it was invented, because the man has been a professional footballer for most of his adult life. Around the edges run thousands of scam adverts wearing the players’ faces. It is delight and deception from the same machine, in the same scroll, and a person decided which to make.

I have started putting a question to the rooms I speak in. When you want to understand what AI is doing to human work, whose voice do you trust?

The obvious answer is the people inside it, the founders and the labs, and there is real signal there. But they are also the people whose valuations depend on the story running a particular way, which Affleck pointed at himself. Some of the clearest thinking I have found comes from people whose whole life is something else, who have nothing to sell you and no reason to flatter the machine or to fear it on cue.

Nick Cave, who has spent forty years writing songs, was sent lyrics a chatbot had produced in his style, and his reply became one of the sharpest things anyone has said about any of this. His answer was that algorithms have no feelings and data cannot suffer, that a real song is dragged out of a human struggle a machine has never been through and cannot counterfeit. And yet, pressed further, he does not play the refusenik. He says plainly that we are all immersed in this now, and that his quarrel is narrower than it looks, aimed only at the idea that making art is mundane enough for a machine to do it.

Paul McCartney used AI to lift John Lennon’s voice off an old recording, clean, so that a half-finished song could be completed decades late. And then the same Paul McCartney stood in front of the British government and told it, in effect, that if it let the machine strip-mine musicians’ work, it would end up with no musicians. He used the tool to recover something he loved, and he fought the tool to protect the people coming after him.

I went looking for all of them for a selfish reason.

When I stand in front of a room and talk about the mindset the book calls augmented, the question comes back almost every time. What do you do? People want to watch someone move between the human thing and the machine thing and decide, in real time, which is which. For a while I gave them me. This is what I do, this is where I stop, this is the judgement I will not hand over. It worked. But it has started to tire me, being the only exhibit, and there is something thin about a man who can only ever hold up his own hands as the evidence.

There is something thinner still about holding them up as if they were clean. I wrote a chapter on drift, and I drift. I have reached for the tool in the exact place I tell audiences to keep their hands off it, handed over work I should have kept, felt the small deadening that comes after, caught it, dug myself out, and reached for it again a week later. The book was never a dispatch from the far side of this. It is a set of notes from a man still in the middle of it.

So when they ask, as they now do, whether there is anyone else, I tell them there is, and I tell them to read the book.

Sabrina Ramonov is in there, one person doing a whole team’s weekly output and marking plainly what the machine touched. So is Sugar, running a beach bar in Thailand and outflanking bigger rivals with a chatbot.

Neither has a face you would recognise, which is the only thing wrong with them as examples.

Which is how I ended up on a man whose face does the work a stranger’s cannot. Christian Bale is the better Batman, and I am not going to pretend otherwise to make a cleaner argument. But Affleck has done in public, at a scale that got printed and priced, the thing I usually have to mime with my own two hands.

The mistake people make is to treat this as a choice they get to make, in or out, for or against. You wake up and it is already in the room. It sorted the emails you read first and the ones you never saw. It is in the phone in your hand, the route it chose for your drive, the show the television put in front of you before you asked. It is at your work, in the tool a colleague used to draft the message you answered before breakfast. It may be in the room with your doctor, reading the scan a half-second before she does. You did not vote on any of this. It arrived.

So the honest position is not on or off. Being against it changes nothing about how much of it you have already touched by ten in the morning, unless you plan to live as a hermit in a cave. It is more like where you stand in politics. Almost nobody is entirely on the left or entirely on the right. You are somewhere on the line, and the spot moves depending on the question in front of you, and pretending you sit at one pure end is a story you tell rather than a place you live. AI is that kind of line. The real question is where you are standing on it, and why, and what you are willing to move.

Which is why the mindset was never one skill, a switch you flip on today and own tomorrow. It is the judgement to use the tool and the harder judgement to keep your hands off it, Affleck’s line about knowing when to stop, carried out of the cutting room and into an ordinary day. It is never a hundred per cent. Anyone who tells you it is has something to sell.

You have a video and a patent too.

The video is the reason you give. You use the tool to clear the dull, heavy work so your hours go to the work that is yours. You would call it more human work, and it was true when you said it.

The patent is the part you keep off camera. The report that took half the time and slightly less of you. The email that reads well and cost you nothing. The skill you have stopped practising because the tool does it for you now. You are getting more human work, and in some ledger you do not read aloud, you are also doing more with fewer people. It is all true at once, and it is you, this week.

The reason Affleck is worth a week of evenings is not that he escaped this. He is holding the same two documents you are, though not from the same place. Some own the machine, some pay for it, and some are the people it was built to do without. The contradiction is shared, but its rewards and its injuries are not. The discomfort of watching him is recognition. He is not the exception to your situation. He is the enlarged photograph of it, taken from the one seat that gets paid for the view.

There is a third account, and nobody puts it on camera or in the patent. Call it the apprenticeship. That’s the space I talk about a lot.

The junior who spent a year painting wires out of shots was not only removing wires. He was learning where the eye goes, what a frame can carry, when an image has been touched once too often. We call that work drudgery, but only once we already know how to do it. To the person still becoming good, it was the training. It was where the taste came from.

Affleck says art is knowing when to stop, and that the machine will struggle with stopping because stopping is taste. He is probably right, and it is the wrong thing to be reassured by. Taste is not issued at birth, and it is not waiting upstairs for people to be promoted into it. It is built, slowly, out of exactly the work his company is designed to remove. The people who already have taste will be fine. The people who were going to earn it, a year of wires at a time, will find the first rung gone and the floor above them full. The video promises more human work. It does not say whose taste will still be allowed to form.

The useful part arrives once the discomfort passes.

The skill was never the absence of the contradiction. Anyone selling you a version that comes without a cost is selling you the thing Affleck refused to sell. The skill is the nerve to stand inside the contradiction with your eyes open and keep making the call, one case at a time. Where does the human belong in this one? Never in general, always in the particular. This task, this scene, this decision, today, with the answer free to move by tomorrow. Cave kept songwriting and let the rest go. McCartney used the machine to recover something he loved and fought it to guard the living. Affleck built the tool and named, out loud, the line he would not let it cross. None of them resolved anything. All of them decided.

That kind of deciding is trainable. It is closer to a moral muscle than a technical one, which is why the book files it among the human skills and not the software. A muscle does not hold a position by itself. It tires. You lose the line, and you take it back, and the taking back is the skill. You get better at it the way you get better at any judgement, by making it out loud, in front of people who can see you, and being willing to be wrong there. Most of us make ours in private and never say the second half.

He is not in my book. If this had happened a year earlier he would have been one of its clearest examples. That is how fast this is moving. Since the book came out, the question people ask has changed. Now people ask what is your next book called. Someone offers Drift versus Design. Someone else says The Augmented Mindset, or SuperSkills with a two after it. I tell them the truth, that I do not have another book in me, and that in any case you cannot write the last word on a thing you are still standing inside. I am still standing inside it. So is Affleck. So is everyone I have named.

I watched the clip again last night. It does not slide past any more.

I see it in people who tells me, with real conviction, that the tool has freed her people for higher work, and who has not filled the roles she stopped hiring for. I see it in the keynote where I point at my own hands and feel how thin that is. I see it in this piece, written faster than I could have five years ago, because I can go deep into the research and keep my brain curating the ideas.

Ben Affleck is human. I do not mean it as praise. He is holding two documents that are both true and do not cancel each other out, and his scale put the second one on the record whether he wanted it there or not. That is what I make of him, and I said I would. The rest of us hold ours in the dark and call it clarity.

He is not perfect, and he has stopped pretending to be. Neither am I, and I wrote the book. Neither are you, holding whatever you are holding as you read this. But we are not holding it from the same place. Some of us get to name our contradiction out loud, from a stage, and be admired for the honesty. Some of us are the line item that honesty is describing.

The video tells you what the tool makes possible. The patent tells you what it costs. The apprenticeship tells you who will still be allowed to learn. Read all three, and start with the one in your name.

SuperSkills: The Seven Human Skills for the Age of AI is out now from Kogan Page. The augmented mindset is one of the seven skills. Ben Affleck is not in the book, although he should be.

Stay Curious - and don’t forget to be amazing,

Rahim

Rahim Hirji Author, SuperSkills (Kogan Page, July 2026) | Keynote Speaker | Advisor
→ superskillsbook.com → thesuperskills.com → Book a discovery call

PS. If your organisation is thinking about AI and human capability, I keynote and advise on this.

Tools I Use:

Jamie: AI Note taker without a bot. You join the meeting. The Bot doesn’t.
Wispr Flow: Just dictate everything to your laptop and phone. Game Changer
Prompt Cowboy: Prompt Generator. Extract intelligence.
Manus: Best AI agent to do things for you. Agentise your intelligence.
Chat Hub: Multi-model intelligence
Claude Cowork: Delegate complex tasks to your friendly Claude Agent

SuperSkills in the Wild:

Send me images of your copy of SuperSkills and I will include them here. Bonus points for creative locations, tourist spots, spotting of the book in the wild, in a book shop etc.

Was on BBC World Service this week talking about AI and data centres. I’m not the most technical, but apparently that’s my strength - translating between the different worlds. Had a double plug of SuperSkills globally - I’ll take it. Will share the clip if and when I get it
40 seconds video of my Book Launch two weeks ago. Grateful and can’t believe it, still. Also available in Instagram
Arab News reviewed my book
My Highest profile piece to date for the new york observer, talking the AI Power Class, Judgement, and Game of thrones, obviously.

Show Your Working

Rahim Hirji — Sun, 12 Jul 2026 06:00:38 GMT

My book SuperSkills: The Seven Human Skills for the Age of AI is available at all good bookstores, through superskillsbook.com or on Amazon

Launching SuperSkills in North America

My book is already out everywhere except from North America - and it comes out in 2 weeks in the US and Canada - so I am doing a virtual launch with a free webinar for everyone. It will be held with Salima Valji, CEO of Unleash Forward where we will talk about How to Stay Relevant and Irreplaceable When AI is Changing Everything. It will be held on Thursday, 16th July at 12:00 PM ET / 09:00 PT / 17:00 London / 18:00 Berlin / 20:00 UAE. Sign up for free and find out more here - & share with anyone interested.

Friends,

There was a particular dread that came with the red pen. You would get your exercise book back, find the page, and there it was in the margin. A tick, and then nothing. Or worse, the answer circled, correct, and beside it a number lower than it should have been. You had got it right. You had still lost the marks.

I remember the exact shape of the injustice. The answer was sitting right there. Anyone could see it was right. And the teacher had taken marks away anyway, for a crime I could not understand at thirteen, which was that I had not shown my working.

The red pen had been in my life since I was ten, maybe even younger from my primary school, though no one there ever explained what it wanted. You learned that it existed. You learned to fear it. And you learned that the whole game, as far as any of us could tell, was to get as little red ink as possible. Nobody said why.

Show your working was three of the most resented words in British education. They meant you could not simply arrive at the answer. You had to lay out the middle, the steps, the crossings-out, the line where you nearly went wrong and caught yourself. The answer on its own earned almost nothing. The working earned everything.

The answer was never the thing being marked. The answer was only proof that the working had happened. A right answer with no working could have been copied, guessed, or lifted from the boy sitting next to you. It told the teacher nothing about whether a mind had moved. We thought we were being tested on whether we could produce the answer. We were being tested on whether we could produce ourselves.

There was a second mark in that same margin, and it did something stranger. RTQ. Read the question. You would find it written beside a pupil who had gone off and solved the wrong problem beautifully. But here was the mercy hidden inside it. If you had clearly understood what was being asked, if you had read the question properly and then slipped up somewhere in the middle, you could still be given marks. You could be wrong and still be rewarded, as long as you had understood. And you could be right and given nothing, if you had not.

A British classroom had worked this out decades before anyone needed it to. Understanding the question mattered more than answering it. We had the whole thing backwards as children, and we spent years being gently corrected.

There was a third mark, rarer than the others, and it meant something had gone very wrong. Not a number. Not RTQ. Two words. “See me”.

We had a boy called Dembinski. Polish, broken English, and one of the sharpest minds in the year. For a while he ran the most efficient business in the school. Fifty pence, and he would do your Latin translation for you. There was an irony nobody remarked on at the time, a boy who could barely order his lunch in English producing flawless Latin translation for everyone else, working from an old thesaurus that gave every line the same reach for the grander word. He always had coins in his pocket, and I used to think of him as selling a kind of cheating drug, a Polish Walter White working a corner of northwest London.

His best customer was Faffalios, the one actual Greek in the class, who did enough trade with Dembinski to have negotiated himself a bulk discount. Faffalios had no intention of taking Latin for another year, so his reasoning was airtight. Why learn a thing you are about to put down for good? He had worked out, years before the rest of us, that you could pay to skip the middle. He learned no Latin at all, and he was perfectly content with the arrangement.

Which is exactly how it ended, at least the part you could see. Everyone who had paid handed in the same translation, all of it too elegant for schoolboys, all of it reaching for the same unusual words in the same places. The teachers did not need to be detectives. And what you got, once they had worked out that the answer on your page had been produced by someone who was not you, was those two words in red. See me.

I was never Dembinski’s friend, and I do not know what became of either of him or Faffalios, both probably running private billion dollar businesses somewhere.

Then we grew up and got it backwards again. The machine on your desk does the precise thing the red pen was built to prevent. It hands you the answer with the middle removed. No steps. No crossings-out. No line where it nearly went wrong. Just the clean result, arriving in seconds, looking exactly like the thing you used to lose marks for handing in.

It is Dembinski at a scale no thesaurus could reach.

And it will answer a question you never read. You can type a half-formed, badly aimed prompt, misunderstand your own problem entirely, and receive back a fluent, beautifully structured response to the wrong question. The machine does not write RTQ in the margin. It has no margin. It answers what you asked, not what you needed, with a polish that hides the difference.

For about two years now, a great many of us have been impressed by answers. I have watched people, myself among them, hold up answers a machine produced and feel something close to pride. Look what it made. Look how good it is. We have been marking the answer, awarding it full marks, and forgetting entirely to ask for the working.

I bring up Socrates in keynotes sometimes, and I can feel the room brace. The posture shifts. Eyes drift to phones. Somewhere near the back a person is thinking, here we go, he is going to do philosophy at us. I understand it, because I used to feel it too. So let me get ahead of it. This is not a lecture about a dead Greek. We had a living one in the class, Faffalios, and he was Dembinski’s best customer. What the dead one actually said was closer to a diagnosis, written twenty-four centuries early, of the machine on your desk.

Socrates left nothing in writing. His objection to it, as Plato records, is usually retold as a worry about memory, but the part that matters is why he distrusted the written word. He said a piece of writing cannot be questioned. Ask it something and it gives you the same answer in the same words, and if you press it, it says nothing at all. It cannot defend itself. It cannot show you how it arrived. It sits on the page looking like knowledge while refusing every attempt to test whether it is.

That is a machine that produces answers and cannot show its working.

Ask it to show its working and it will, but the working is only another answer, as fluent and unaccountable as the first.

Socrates wanted the kind of knowledge you could cross-examine. The kind that only exists inside a mind that has done the labour, because only that mind can be asked to account for the middle, to prove the working was real. The red pen wanted exactly the same thing. A teacher writing show your working and a philosopher refusing to write at all were doing the same job, two and a half thousand years apart. Both were saying that an answer you cannot account for is not yet yours.

I used to think the danger of the machine was simply that it made us lazy. That is true, but it is the small version of the problem, and I no longer think it is the one that matters.

The larger version is this. When everyone can produce the answer, the answer becomes the cheapest thing in the room. The same fluent paragraph goes to you, and to the person beside you, and to ten thousand strangers, and proves nothing about any of you. The value has drained back to where it always lived, into the working, and most of us have spent two years attending to the part that no longer counts. We have gone answer-rich and working-poor, fluent in results we cannot account for, able to hand in the answer and unable to show the middle.

Because this is really a question about human capability, and human capability is built in the middle or it is not built at all. Every time you did the working by hand, slowly, badly, with crossings-out, you were building the capability that would one day let you judge an answer, someone else’s or a machine’s. The working was the capability forming in real time. The answer was just the receipt.

I took Latin for years, from a man called Critchley, who once shaved off his beard, walked in, and announced that Mr Critchley could not be in today and that he would be covering the lesson. I could not translate a line of it for you now. I am not sure I ever learned much Latin at all. But I enjoyed it, and I like to think it left me something. Some logic. Some history. The reason I still call on the philosophers now and again, even if they get a yawn. None of that is Latin. All of it is the working. Faffalios thought he was skipping Latin. He was skipping the part that stays.

And for years, we have rewarded the output: the finished paragraph, the polished deck, the answer marked full marks. And we continue to. Nobody rewards the capability underneath, because capability does not show, and a thing that goes unrewarded goes unbuilt. That is the debt we are running up without seeing it. I call it capability debt. The gap between what you can produce and what you could still do if the machine were switched off.

Which is why the hard edge has nothing to do with laziness. You cannot judge an answer you could never have produced yourself. When the machine is wrong, and it is often wrong in fluent, plausible ways, the only person who can catch it is the one who has done that working at least once. Someone who knows where the middle tends to go wrong. Someone who can look at a clean, persuasive result and feel the small wrongness in it, because they have been to that place on foot. Take that person out of the room and you are left with people holding answers, none of whom can tell which ones are true.

We were trained out of the child’s error. We have trained ourselves straight back into it.

I think about the red pen differently now.

It felt like punishment at the time. The marks it took. The working it demanded. Those two words, See me, and the long walk to the front of the class, when everyone ran off to have fun at break time. It was something else. Show your working and read the question and see me were instructions sent decades early, for a test that had not been set yet. Critchley, and all of those teachers, standing at a blackboard in the nineteen-eighties, was preparing me and my generation, without knowing it, for a moment when answers would cost nothing, and the only things left with any worth would be the capability to show how you reached one and the judgement to know whether the question was worth asking in the first place.

We used to dread the red pen.

We should have been dreading its absence.

There was a song we sang at that school, in Latin. Homo plantat, homo irrigat. Not one of us knew what it meant. We had the words exact and the sense nowhere, a room of boys singing an answer none of us could account for, and no one ever troubled to translate it.

It means: one plants, one waters. The increase, the original said, was God’s to give. It is the machine’s to give now. But the planting and the watering, the only parts the song ever asked of a person, were always ours.

They still are.

Homo plantat, homo irrigat.

SuperSkills: The Seven Human Skills for the Age of AI is out now (Kogan Page). It’s written in English. There are some Greek references in there, but no Latin.

Stay Curious - and don’t forget to be amazing,

Rahim

Rahim Hirji Author, SuperSkills (Kogan Page, July 2026) | Keynote Speaker | Advisor
→ superskillsbook.com → thesuperskills.com → Book a discovery call

PS. If your organisation is thinking about AI and human capability, I keynote and advise on this.

Tools I Use:

Jamie: AI Note taker without a bot. You join the meeting. The Bot doesn’t.
Wispr Flow: Just dictate everything to your laptop and phone. Game Changer
Prompt Cowboy: Prompt Generator. Extract intelligence.
Manus: Best AI agent to do things for you. Agentise your intelligence.
Chat Hub: Multi-model intelligence
Claude Cowork: Delegate complex tasks to your friendly Claude Agent

SuperSkills in the Wild:

Send me images of your copy of SuperSkills and I will include them here. Bonus point for creative locations, tourist spots, spotting of the book in the wild, in a book shop etc.

Join my North America launch Webinar this Thursday
My piece touching on SuperSkills in Entrepreneur magazine: Why AI doesn’t create bad decisions, it just exposes them faster
On the BBC last week. You can find me interviewed for SuperSkills at 37 minutes onwards
Talking Edtech and SuperSkills on MSN: How Sweden’s ‘edtech’ catastrophe became a cautionary tale for Britain’s digital-first classrooms
My article for CEO World Magazine talking about the main concept of SuperSkills: Drift versus design: why most companies mistake activity for transformation

Send me your images of SuperSkills in the wild, in bookshops, on holiday. Best photo gets a prize.

Solving Synthetic Seniority

Rahim Hirji — Sun, 05 Jul 2026 06:02:03 GMT

My book SuperSkills: The Seven Human Skills for the Age of AI is available at all good bookstores, through superskillsbook.com or on Amazon

Friends,

In case you missed it - and how could you? - My book launched this week. I wrote about how I planned it - and how the launch party didn’t go to plan in my piece The woman in the back of the room. I know you will love it.

On to this week:

Justine runs a studio that makes the fifteen-second films you try to skip before the button lets you. The work is a car on a coast road, a bank that promises to understand you, a soft drink and some snow. It goes out under the names of FTSE 100 companies, where the bar is unforgiving, because a frame that is almost right reads as wrong to an audience that could never tell you why.

She gave her graduates a brief one morning and took their laptops away, no software, only pen and paper and whatever they could reach with their own hands and their own eye. She set the same brief for Niamh, one of her designers a few years further on, and let her work the way she always worked, which was to open the tool and begin.

Niamh finished first, the way she always finishes first. When the display ads were laid across the table side by side, every one the graduates had built by hand was better than hers. She was faster than every one of them, and she was not good enough. In the work Justine sells, good enough is average, and average does not sell.

There is a name for what Niamh has. I call it synthetic seniority, and I have been saying it out loud to everyone who will listen for a while now. It is work that looks senior, sounds senior, and passes every inspection in the room except the one that matters, while the judgement underneath it was never built.

When I first wrote it down, I described a junior, a second-year with a strategy deck, a trainee with a memo that reads three grades above her experience. I had it too narrow. AI rarely takes a whole job; it takes tasks out of the middle of one, and what is left gets bundled with other people’s work until nobody is doing the job they were hired for, in a period when the shape of every role is still moving under everyone’s feet.

That is why synthetic seniority is broader than plain inexperience. A person can be a real master of the thing they were trained in and still be synthetically senior, because the world has handed them two other crafts they have never practised and called the bundle their job. The gap is not inside them, it sits between them and a role that did not exist two years ago. It is the engineering manager who now owns the product roadmap too, learning it in public. It is the solo operator doing her own finance and marketing because the walls between those jobs fell in. It is the director of twenty years whose craft is intact and whose world has rearranged itself around her. Almost everyone is somewhere in that shift now. It is why the room turns when I say the words, and why people have lived inside the thing for two years without a name for it.

Go back to Niamh, though, because her morning explains it better than any definition. The easy conclusion from Justine’s experiment is that the tool made her worse. It did not. What it gave her was a floor - the floor. Whatever she makes arrives at a certain level and never drops below it, because the software was trained on a gazillion competent things and competent is what it gives back. It cannot fail her, and it cannot lift her. The pattern I keep seeing across these creative firms is that the work settles at an average and stays there, and the number I have come to put on that average is seventy per cent.

She mistook the floor for a ceiling. Everything she made was fine, and because it was fine, she stopped doing the thing the graduates were forced back into that morning, which was to be bad, by hand, over and over, until their hands learned something. Nobody had told her those repetitions were the point. They are the ones that build the last thirty per cent, and the tool will hand anyone the seventy. The last thirty has only ever been built the slow way, by making the work badly enough times that your eye changes.

In Justine’s trade, seventy per cent does not merely underwhelm, it gets skipped, and the reason is measured in seconds. Most ads get under two seconds of a distracted person’s attention before the scroll, roughly the threshold for forming any memory of them at all. Competent does not win that sliver. It is precisely the level that gets ignored, because it is good enough to be inoffensive and nothing more. The whole value of the work lives in the part that seizes a bored stranger before the thumb moves, and that part is the one the tool cannot reach.

The seventy passes everyone but one person in the building. It passes the client, who is delighted and on time. It passes the account lead. It passes the junior review, because the juniors sit at seventy themselves and cannot see above their own line. The only person who sees what is missing is Justine, because she climbed it herself, years ago, and someone who has climbed something by hand can always tell when someone else took the lift.

Justine has watched this coming. She talks about the Coca-Cola Christmas ad from last year, the one that rebuilt the old red trucks and the snowy towns with generative AI, a year of work done in a month, cheaper, faster, and received as something soulless. Her objection is not that it looked cheap, because it did not. It is that it looked like nothing, a competent nothing produced at a scale nobody could previously afford. She talks about the ads that still work, the John Lewis Christmas films people wait for each year, and how much human doubt goes into every cut before it airs. She has watched people get burned handing the tool the part that was theirs to hold.

None of this, though, is where Justine started. She is fifty-five-ish and vapes, the oldest person I know who does, and for most of twenty years she ran the studio like a lifestyle business that happened to be large, loose and generous and built around the work being a pleasure. She was a tools-first believer then, the one in the room telling everyone the software was the whole future, moving faster than the people around her and proud of it. She would have argued Niamh’s side without blinking, because for a long time it was her own. She did not take the laptops away to catch her graduates out. She took them away because she had begun to suspect something about herself, and she wanted the answer even though she feared it. The result caught her. Her most promising designer lost to a line of graduates working with their hands, and underneath that loss was the road she had spent two decades on. She took the answer and did the hard thing with it, and began pulling her whole practice apart to rebuild it.

It cost her. It also turned her into a different kind of boss, the lifestyle business gone, the place run like a turnaround now. Slower and dearer than a studio running everything through the tool, she lost clients who only ever wanted it fast and cheap, and some of them never came back. Revenue is down and has stayed down. The studio makes more profit on less work now, though she is honest that some of that is not vindication, it is a smaller team, people who were there before the turn and are not there after it. That is the kind of result nobody puts on a conference slide, because it is not a triumph and it is not a warning, it is just what the honest version cost, and who it cost. it’s messy

What she is rebuilding towards is visible in the work itself. Two of the most effective films ever made in Britain show what the last thirty per cent actually is.

John Lewis spends every Christmas making the country cry over a boy and his monster, or an old man alone on the moon, and none of it is sentimental by accident. The firms behind those ads measure the feeling precisely, tracking what a viewer feels from second to second, because emotion is what a person remembers and what moves them months later at the till. A story without friction gives no release, and the wrong music kills the moment it was built to carry. The crying is engineered, built on purpose by people who have spent careers learning how feeling is made, and it is the part no tool can originate, because it has to come from someone who has felt the thing themselves.

Then there is the strangest of them. In 2007, after a contamination scare had damaged the brand, Cadbury aired a film in which a gorilla sits at a drum kit for the better part of a minute, waiting, while a Phil Collins track builds, and then plays the famous fill, with no chocolate on screen until the final seconds. It made no rational sense and should not have worked. It rescued the brand, sold chocolate, sent the song back up the charts and won the Film Grand Prix at Cannes. And it began as an office argument about who had recorded the best drum solo of all time. The people who made it could never quite explain how they got from the argument to the gorilla. The marketing director said he heard the idea, did not fully understand why, and knew in that instant they had to make it. A machine can throw out a strange association by the thousand. What it cannot do is know, in its bones, why this particular strange thing is worth betting a battered brand on. That knowing comes from a person, and only from a person.

What the engineered tears and the inexplicable gorilla have in common is a refusal to stop at the point where everyone else in the room was already satisfied, a reach past good enough into something slightly irrational in the way people are and machines have no reason to be. That reach is the whole of the last thirty per cent.

The leap was human, and everything after it was craft and testing and doubt. Cadbury held the launch back to research the idea properly. A specialist studio built the suit. A performer trained for weeks to drum convincingly inside it. The film was placed with care, at the right moment, in front of the right audience. A human at the start, where the idea that crosses the threshold is born. Testing and rigour in the middle, where a good idea is pressure-tested and delivered. A human again at the end, where taste decides when it is finally right. The gorilla was made in the shape Justine is now rebuilding her studio to work in.

Somewhere in here is a fairer question than the one everyone asks. It is not whether you use the tool, because everyone uses the tool. It is where. Some work is transactional, the expense report, the meeting notes, the first pass of an email you were going to rewrite anyway. Hand every scrap of it over, talk a good game, and lose nothing, because none of it was ever going to be the thing your name is really on. Some work is representative. It is the work that is you, the frame that has to seize a stranger in under two seconds. Use synthetic seniority in the transactional zone and it is a debt you can carry for the rest of your life and never be asked to repay. Let it cross into the representative zone and it becomes a loan that gets called, always at the worst moment, always in front of the client, always on the day the tool has no answer and the room turns to you.

The border between the two has gone faint. When everything arrives at the same competent level, on the same screen, in the same clean font, the transactional and the representative start to feel identical, and you stop being able to feel which is which. When the competent thing looks the same whether it took thirty minutes or thirty days to make, the eye goes blind, and the seventy starts to look like a hundred, because there is nothing left to hold it against. A person who cannot feel the border will let the debt cross it without noticing, because nothing at the crossing looks any different from anything else.

The danger I wrote about the first time was a broken pipeline, a generation promoted past its judgement, a leadership vacuum waiting for the next crisis to open it. That is real. It is also not the worst of it, and Justine’s eye is what shows the worse thing. Everything rests on there being a Justine in the room, someone who climbed the hard part by hand and can therefore see, at a glance, when it is missing. Take that eye out of the building, by retirement or redundancy or the simple passage of time, and the seventy stops being caught, because nobody left can see that it is only seventy. Work that is never flagged as below standard does not stay below standard. It becomes the standard. The ceiling drops, and it drops without anyone noticing, because the only people who could have told you it used to be higher have already gone. A whole trade can forget what its own best looked like inside a single generation, and then forget that it has forgotten.

Justine turned, and that is the rarest part of the story. She was on the wrong road and she corrected, and most will not, because correcting means running an experiment whose answer you already dread. The tools-first veterans who never run it will go on approving the seventy from above while the juniors produce it from below, and between them, in good faith, they will agree that seventy is what a hundred always looked like.

So the correction is not a policy. Justine is not writing a framework, and the last thing any of this needs is another three-step programme with prompt logs and an oral exam, which is roughly what I proposed the first time I wrote about synthetic seniority, and which mostly gives organisations a way to feel busy. What she is building is the shape the gorilla was already made in. A person starts the work, holding the judgement about what the thing is even for, the taste, the read of the room, the sense of what would make a stranger stop at all. Then the tools do what they are honestly faster at, the drafting and the variants and the testing a human would take a month to grind through. Then a person finishes it, and the finishing is where the hard part goes back in by hand. Human, then machine, then human. It sounds obvious written down, and almost nobody does it, because the tool in the middle is so fast and so plausible that the pull is to let it run all the way to the end and ship the seventy.

Niamh is arriving at the same place from the other side. She has not put the tools down, she is learning where in the work to pick them up, which is later than she used to, and for less than she used to, and never for the part that is representative. The veteran is learning to let the machine in. The junior is learning to keep it out until the right moment. They are walking towards each other, and the place they meet is the only version of this that holds.

Justine took the laptops away to find out what was still there without them. It is the most useful question I have heard all year, and it is not really a question about designers. Take the tools away from your own week, not the expenses and the notes, but the work with your name on it, the work that has to cross the threshold. What is still there? What can your own hand or brain still do?

Before you answer too quickly, remember who Justine is. She is not a sceptic who never trusted the software, she is a believer, a tools-first woman of twenty years, heading exactly where Niamh was heading, who almost did not turn in time. If it took her a stripped-back room and a line of graduates beating her best designer to see what was missing in her own studio, then your certainty that you would catch it in your own work, unprompted, with nobody taking your laptop away, might be the most synthetic thing you own.

SuperSkills: The Seven Human Skills for the Age of AI is out now. The whole book is really one long argument for climbing the last thirty per cent yourself.

Stay Curious - and don’t forget to be amazing,

Rahim

Rahim Hirji Author, SuperSkills (Kogan Page, July 2026) | Keynote Speaker | Advisor
→ superskillsbook.com → thesuperskills.com → Book a discovery call

PS. If your organisation is thinking about AI and human capability, I keynote and advise on this.

Tools I Use:

Jamie: AI Note taker without a bot. You join the meeting. The Bot doesn’t.
Wispr Flow: Just dictate everything to your laptop and phone. Game Changer
Prompt Cowboy: Prompt Generator. Extract intelligence.
Manus: Best AI agent to do things for you. Agentise your intelligence.
Chat Hub: Multi-model intelligence
Claude Cowork: Delegate complex tasks to your friendly Claude Agent

SuperSkills in the Wild:

Send me images of your copy of SuperSkills and I will include them here. Bonus point for creative locations, tourist spots, spotting of the book in the wild, in a book shop etc.

A recommendation for SuperSkills: AI and volunteer engagement: threat, distraction, or opportunity?

The Woman at the Back of the Room

Rahim Hirji — Fri, 03 Jul 2026 10:55:04 GMT

My book SuperSkills: The Seven Human Skills for the Age of AI is out now. Buy through at superskillsbook.com or on Amazon

Friends,

Yesterday I launched my book. This is not the newsletter I had planned to write about it.

There is a second, a small moment, just before you stand up to speak, when the loudest thing in the room is your own pulse. That second was the one I had not planned for, because I had not planned to speak at all.

Everything else, I had planned. There was a run of show, and it had a version number, because I have been living this book for years now and have become, I am reliably told, a bit of a control freak. The readings were timed. The fireside chat I left loosely mapped, because I trust Chin-Ru as an expert, and the best conversations are the ones you do not over-plan. My elder daughter would open the evening. My younger one would read the prologue. There was a vintage dhow on the stage, because the whole thing had a shape and I wanted the shape to hold.

Years go into a book. The night that launches it lasts about two hours. I wanted those two hours designed down to the minute, and they were.

And still there was one thing no plan could have reached. I did not know how to plan for it.

Almost 250 people came to the Ismaili Centre in South Kensington on Thursday evening. Good people. A warm room. The kind of evening where strangers talk to each other and enjoy the conversation.

None of them knew that my wife had spent that day in hospital.

I am not going to tell you why. What I will tell you is that I spent the hours before the biggest night of my working life not thinking about the book at all. I had not seen her. I did not know if she would be there. I got dressed. I went with my daughters, both of them more composed than their father.

It was immediate overwhelm. But I did the thing I said I would do. I talked with the people I knew, and I found the ones who had flown in whom I had never met, and I tried to thank as many people as I could. I do not think I have ever smiled that much in one evening, and my heart was beaming the whole time, even though it was slightly broken.

Inevitably, my daughters carried the evening.

My eldest ran the night. The concept was designed, the shape of it was mine, but the words were hers and the performance was hers, and there was a spontaneity in what she did that owned the room in a way no plan of mine ever could. She held it together and kept it moving, with a composure I did not have at her age and am not certain I have now.

The younger one read the prologue. My words, in her voice, to a room of two hundred and fifty people. I have read that prologue more times than I can count. I had never once heard it until Thursday. She made it sound better than I had it in my head.

Afterwards, people kept coming up to talk about them. Not about the book. About the girls. They were wonderful, they said. They smashed it, they said. There is a particular feeling that arrives when a stranger praises your children to your face and clearly means it. You spend twenty years unsure, and then for one evening you are almost sure. I messaged a friend to say I did not know where they got it from. He messaged back two words. Apple. Tree.

There is a strange grief buried inside a good night.

The room was full of people I love. Friends I have had for decades. People I had not spoken to in far too long, standing right there in front of me. And I could give almost none of them what they deserved. Two minutes, when I wanted two hours. A hand on a shoulder when I wanted a proper conversation. Some of them were in the room the whole evening and I never once got to say hello. I will be apologising for that for weeks. Every one of them will tell me it does not matter. They will be lying kindly, and I will love them for it.

Some people cancelled at the last minute. It broke my heart a little. It was nobody’s fault. That is just life, and life was busy that day.

I had not planned to make a speech. I had no script, no paper, nothing written down. But it did not feel right to go into my reading without saying something first, so I said something.

I cannot tell you most of what it was. It has gone. It is a blur. The one thing I remember is that I thanked everyone who and gave each of my daughters a made up SuperSkill, the skill I see most clearly in each of them, out loud, in front of everyone.

And then I had saved one for last. My wife. She had willed her way out of hospital that day, and even so I did not think she had made it. I had kept looking round at her reserved seat, but she wasn’t there. I said as much, out loud, to the room. “I don’t think she’s here.”

And then I saw her. At the back. Standing where she could barely be seen, because she did not want any fuss.

So I made a fuss.

I finished my thank yous. I finished my readout, and I ran. Across the room, to the back, to her. A grown man running through his own book launch to hug his wife, faintly ridiculous, a bit like Jerry Maguire, and I would not change a second of it. I have spent months scripting everything, and the best moment of the whole night was the one nobody wrote.

That was where I nearly broke down in tears. Not because of the book, but because of her. Because she had decided this was worth leaving hospital for, and had tried to slip in unseen, and had been found out.

When I was on the podium, I saw someone wiping their eyes, and then another beside them doing the same. There was laughter, and there were tears, and there was love in that room. I used to be embarrassed to cry. I know now that it is one of the most human things we do. I have never been hugged so many times in one evening, and some of the people who hugged me read this newsletter and had never once met me in person until Thursday. You know who you are. Thank you.

On the way home, the girls asked me where my speech script was. There was no script. Five years ago I could not have stood in front of a room without one. Somehow, on the one day I had every reason to come apart, I did not need it.

I have been trying to work out why. When you take the script away, what is left is whoever you actually are. Take mine away, and what is left is a man who stands up and talks about his daughters. I will take that.

I wrote a book about the human skills a machine cannot do for you. Curiosity. Judgement. The way you read a room. The willingness to care about something and let it show. The argument of the whole thing is that as the machines get better at the tasks, the human parts stop being the decoration and become the point.

I had been making that argument on paper for a long time. On Thursday it made itself, on me, in front of everyone.

The tasks, you can hand off. The run of show, the timings, the logistics, the words on a slide, all of it can be given to somebody else, or to a machine, and more and more of it will be. What cannot be handed to anyone, or to anything, is the thing that made me run to the back of a room in front of everyone I know. That is the part the book is about, and it is what is left standing when everything else has been automated away.

You design what you can so that you are free for the part you cannot. You build the careful room so that when the moment you never planned for walks in at the back and tries not to be seen, you are awake enough to reach it.

I have written a book about the world of artificial intelligence, and there was not a single screen in the room. No slides. No AI. Nothing to look at but each other. Just words, and conversation, and people standing in clusters arguing about design and drift, which is the argument the whole book is built on. Different walks of life, all of them talking to each other, and that was the best thing in the room.

Someone called me a tech bro that evening and meant it kindly. I did not have the heart to tell him there was not a single machine switched on in the room all night.

I had always assumed a book launch was a stuffy thing. Heavy speeches. A glass of warm white wine you hold and do not drink. Instead, we had banter and pies and brownies and fortune cookies and cold water and, against everything I had expected, real warmth. A book about machines, launched by a room full of people being entirely human at each other for two hours.

Someone said to me that night: I nearly didn’t come. I’m so glad I did. I keep hearing it. I nearly didn’t come.

I have been asked a great deal lately what success looks like for this book. There are answers I am meant to give, the ones about charts and first weeks and copies shifted. I do care about those. But if you want to know what success actually felt like on Thursday, it was a stranger deciding, against their own hesitation, to get on a train for an idea. And it was a woman deciding that a room full of people talking about her husband’s book was worth leaving hospital for. Neither of those is a number, and neither of those can be optimised or bought.

This book has my name on the cover. It is not mine. It belongs to everyone who was in that room, to my parents, to my wife and my daughters, and to the people who got on a boat in Gujarat a century ago so that one day someone they would never meet could stand on a stage in South Kensington and struggle to get his words out.

I am writing this on no sleep, and sending it to you on a Friday and not a Sunday, because the book is out today and I could not wait until Sunday to tell you.

Many years came down to one book, on one night, in one room. It was a great deal to hold. At some point in the small hours I stopped trying to hold it and let it be what it was. Which was overwhelming. Which was joyful. Which was, frankly, a lot.

There was food everywhere. I forgot to eat.

Because I am human. That was rather the whole point.

SuperSkills: The Seven Human Skills for the Age of AI is out today. If it finds its way to you, I hope it turns up on a day you happen to need it.

Stay Curious - and don’t forget to be amazing,

Rahim

Rahim Hirji Author, SuperSkills (Kogan Page, July 2026) | Keynote Speaker | Advisor
→ superskillsbook.com → thesuperskills.com → Book a discovery call

Share Box of Amazing

If you enjoyed this, forward it to someone who’d find it useful. And if you haven’t yet, order SuperSkills at superskillsbook.com .

Actual Intelligence

Rahim Hirji — Sun, 28 Jun 2026 06:00:42 GMT

My book SuperSkills: The Seven Human Skills for the Age of AI is out THIS WEEK on 3rd July 2026. Pre-order at superskillsbook.com or on Amazon

Friends,

My book comes out this week, and I have been waiting for most of my life for it to come out.

Roughly 33 years ago, I sat at the front of a lecture theatre at Loughborough University. I had chosen the course because it blended computing with real work, the actual doing of things, and I liked that. I sat at the front on purpose. I had decided long ago that the front was where you kept up without distraction, and I was not sure I could keep up. Behind me were students from across the world, and at the bottom of the room, right in front of me, a lecturer was talking about artificial intelligence. He spoke about algorithms. He never once spoke about what any of it would do to a person.

I understood almost none of it. I remember the specific fear of it, the sense that everyone else had been handed a key I had missed. The fear was not really about the algorithms. It was older than that room, and I still cannot say exactly what it was. Sometimes you have no thesis, because you do not understand the thing well enough to have a thesis. Sometimes you are just too immature. Sometimes you have no framework, because you do not yet know what a framework is. The schooling I had taught me what to think and never how to learn. I would spend the rest of my life learning how to learn.

There are writers who pounce on an idea at the exact moment the world turns to face it. They are not always the cleverest. They are the best timed. And there are writers who pounce on an idea years before anyone wants what they are holding. I have spent my career in the second group.

I was building online tutoring before most parents would trust a screen to teach their child. I was putting AI and blockchain in front of schools and businesses long before either word meant anything good to them. Too early is a strange kind of failure. You are right, and it does not matter, because being right before anyone is ready is indistinguishable from being wrong. You spend the years explaining instead of selling, and by the time the world arrives, someone else is standing where you stood, taking the applause you waited for.

So I do not know what this week is. The book might be on time. It might be too early again, the thing I keep doing. It might already be too late, the argument overtaken while I was launching it. I have stopped being able to tell the difference. I only know I have waited long enough.

When I was a young child in primary school, a teacher would serve us our lunch. The shepherd’s pie went cold on the plate while we sat and waited for the whole table to be served, and only then did we say the Lord’s Prayer, and after it a grace. For what we are about to receive, may the Lord make us truly thankful. Amen. I am not Christian. I still know every word of the Lord’s Prayer. Permission to begin was a thing you waited for, and the waiting was the point, not the pie.

Years later, I lived in Abu Dhabi, and during Ramadan, I watched the same lesson taught harder. I was sitting in a Pizza Hut, of all places, one of those iftar deals where you pay once and eat as much as you can in an hour. Around me sat young men and boys with seven slices of pizza in front of them, untouched, silent, listening for the call to prayer that would allow them to eat. When it came they fell on the food, and it was gone in minutes, and within the hour they looked ill, because they had eaten too much too fast. Waiting is a discipline you are taught as a child and then spend the rest of your life forgetting.

We have built a world that punishes the wait. Everything is now, and I am as impatient as anyone, frustrated by the very machines that thrilled me a few years ago. I have four or five Alexas in the house. When I bought the first one, it felt as though the future had walked through the front door. I had visions of a connected home, the lights and the music and the heating all answering to a word. We managed it, briefly. Now they are a radio and a way to time a poached egg. A platform that looked as though it was waiting for its moment, and waited too long, and the moment left without it.

I have been reading the biography of Jensen Huang. Nvidia was founded in 1993, in a booth at a Denny’s. Thirty-one years later it became the most valuable company on earth. That is the loud version of waiting, the one that gets a book and an award and a place on every shelf. I am not Jensen Huang and I have no illusions of grandeur. This is not Nvidia. The scale is certainly not shared. Only the shape is.

The waiting I recognise belongs to someone smaller, and older. Samuel Pepys was known in his own lifetime as a naval administrator, a man of committees and ledgers. He kept a diary in shorthand, for himself, expecting no reader. It was published more than a century after he died, and only then did anyone understand that he had recorded a whole society remaking itself, the fire, the plague, the ordinary terror of a changing world. In his life, nobody knew that writing was the thing. The thread does not always announce itself as you spin it.

The ideas in my book did not start with me. They came from people who waited before I was born, and they will date me to whoever reads them after I am gone. I am a thread, not the beginning of one.

None of my waiting was idle. I have written Box of Amazing since 2017, five hundred or so Sundays, and for a long time almost no one read them at all. Before any of that, I wrote travel guides, ebooks nobody remembers, including me on most days. I thought those were the thing. They were not the thing. They were practice for the thing.

A machine can now write a competent version of this essay in four seconds. It does it by swallowing everything human beings have already written and handing back the average of it. The four seconds is the part that impresses people. The 30 odd years is the part it cannot fake. The waiting is the proof a human made it.

There is artificial intelligence, which is what that lecturer was describing without knowing where it would lead. There is actual intelligence, the human kind, the slow kind, the kind I was so afraid I did not have at the front of that room. And there is emotional intelligence, the one the machine cannot reach at all, the one that knows when a person needs you to stop talking. The seven skills in my book are those last two put to work. The frightened boy in the front row did learn, in the end, how to learn. It only took him a very long time while he was still growing up.

There is a difference between perfect and good enough, and a harder difference between both of those and finished. Perfect is a standard you can wait at for the rest of your life. Good enough is a lower bar you still cannot clear on the days you believe the worst of yourself. Finished is neither of them. Finished is the moment you stop waiting and let the thing go.

None of those bars were mine. Someone set the height early, a teacher, a room, a tone of voice, and I have spent a life measuring against a mark I never agreed to. What I have called patience was often just that measuring. It was the fear that if I delivered the thing, someone would confirm what I had suspected since the front row. And finishing does not end the fear. I will hand over the book on Friday morning and it will still be there, because it was never really about the book.

On Thursday, the day before the book, I’m hosting a book launch in London for everyone who has supported me. There will be 250 people or so. I am an introvert. I have waited years for this and I dread it, which is an absurd thing to admit and a true one. The dread is not new. It is the same tightening I felt at the front of that lecture theatre, the same certainty that I am about to be found out. The body knew before the mind would say it. The student and the author are the same person, and he is still afraid. Maybe Thursday is the day he finally grows up, the day before the book, in a room full of people, 30 plus years late.

This has been my waiting, and you have your own. It might be a course you never took, or work that has been finished enough for a year in a folder you do not open, while you wait to feel worthy of sending it. I do not know what yours is. I know the voice telling you to wait a little longer is almost never your own.

We call ourselves the human race. I have always wondered about that word, race. A race against who? Against time, against each other, against the version of ourselves that was meant to have arrived by now. I do not know if the race ever stops, or whether waiting makes the thing worth more or only makes it late. Maybe the book is too early, the way I usually am. Maybe it is too late. Maybe, this once, it is on time. I have waited long enough to stop minding which.

The wait ends on Friday. The fear will not, and I am letting the book go anyway.

It was never about the book. It was always about the realisation, which took most of my life to arrive, that the waiting itself was the work, and that the frightened boy at the front of the room already had the one thing he was sure he lacked. That realisation, and the nerve to release the book while the fear is still here, is what I have waited my whole life to reach.

My book, SuperSkills: The Seven Human Skills for the Age of AI, is published by Kogan Page this week, on 3 July 2026. You can find it at superskillsbook.com.

Stay Curious - and don’t forget to be amazing,

Rahim

Rahim Hirji Author, SuperSkills (Kogan Page, July 2026) | Keynote Speaker | Advisor
→ superskillsbook.com → thesuperskills.com → Book a discovery call

PS. If your organisation is thinking about AI and human capability, I keynote and advise on this.

SuperSkills in the Wild:

David Boyle writes Saturday AI Thoughts, a weekly reckoning with what AI is doing to how we work. Twenty-five years in data and audience insight, a sceptic's eye, and a gift for calling these tools what they are: an electric bike for the mind. He opened SuperSkills expecting to argue with it. He did, then talked himself into recommending it anyway. A rave is a favour. A sceptic who fights your argument and still tells people to read the book is something you can only earn. His essay is here.
Podcasts:

Nankhatai

Rahim Hirji — Sun, 21 Jun 2026 06:00:55 GMT

My book SuperSkills: The Seven Human Skills for the Age of AI is out 3 July 2026. Pre-order at superskillsbook.com or on Amazon

Friends,

The plate has to go back. It arrives covered in clingfilm, carrying biscuits somebody made in a kitchen that is not ours, and at some point the plate has to return there. Sometimes it is not a plate but a reused tub that a Chinese takeaway came in, or Tupperware that has been moving between families for longer than my kids have been alive or one of those freezer bags. They come from a friend of my mother’s, that family friend everyone agrees makes the best ones. They come via my mother, and eventually to us. She brings them round usually as a supplement to something else, usually with a few clippings she has cut out for me, an article, a notice, something she thinks I should read. The biscuits and the clippings are the same thing, really, a thread she keeps tied between her house and mine, and the plate going back is how she keeps tying it.

Inside that plate are “nankhatai” [pr. “Naan” “Cut” “Eye”], small shortbread biscuits I have eaten my whole life, pale gold, cracked across the top like dry earth, and ours have a thin swirl of chocolate through them that is not traditional and is the only kind I will now accept. You lift the clingfilm and the smell comes up first, cardamom and warm butter.

Everyone in my family knows what nankhatai is. We have never once had to talk about it. It’s never the main event, yet it always seems to have been part of our lives.

These ones come for my youngest daughter. She likes them. But the truth is I am usually the one who finishes them, standing at the kitchen counter when nobody is looking, because she likes a hundred things and I have come to like fewer, and these are one of the few that are still mine. The plate is meant for her. I am the one who is hungry for it.

Have you ever eaten a biscuit straight out of the oven, before it has had the sense to cool? It is a little like that, except this one arrives with a few hundred years behind it, and it has been changing the whole way, and no two of them are ever quite the same, not within a single batch, where one comes out darker and one bigger and one cracked where the others held, and not from one country to the next, where the same biscuit took on more cardamom here and more sugar there. They are imperfect, every one of them, and that is the exact thing that makes them perfect, and it is the one trick a machine, which can only ever hand you the same thing twice, will never learn.

I tried to picture someone making them, and I couldn’t. I have eaten nankhatai my whole life and I do not hold a single memory of watching one be made. There is no scene. The people who make them just make them, alone in their own kitchens, the way you do a thing you have done ten thousand times, and then they share them, because they know everyone will be glad of them, and what reaches you is only ever the result on a plate. You never see the hands. Maybe that is the whole problem, sitting there in plain sight the entire time. You cannot inherit a thing nobody ever showed you how to do.

When we were small they were simply there, in the Tupperware when we came back from school, in tins lined with kitchen roll, a snack you could have whenever, gone the way fresh biscuits go, eaten standing up with a glass of milk, the crumbs caught in a cupped hand in under a minute. They were made by women I half remember and women I never met, and they were always good, every time. I did not think of them as heritage, I thought of them as furniture. The only thing I knew was that this one was ours, and the rest were just food.

Each of these women is known for a speciality, one dish, theirs, the one you go to them for. My mother is known for her chicken pastries, though she makes plenty of other food that keeps our family well fed. Someone else is famous for samosas. My mother-in-law is known for her stuffed parathas, and there is no point the rest of us trying to make them, because they will not be hers. They can all cook plenty, but each of that generation has a single thing that is the taste of them, so that when they are gone the dish is part of what people will mean when they say their names.

You will have your own version of this, even if it looks nothing like mine, your own nankhatai, your own piece of heritage inside your own family and your own culture, the kind of thing you would struggle to explain to anyone who did not grow up with it. It might be the challah on a Friday, the pan dulce on a Sunday morning, the mooncake once a year, or the thing somebody always brings back from a trip because to come home with nothing would be unthinkable. It is rarely grand. That is the point of it. It is just there, in the background of a life, until it is not.

Then there is my generation. We can all cook, a bit of everything, this cuisine and that one, taken from Google, magazine cuttings, websites - everywhere and tied to nowhere in particular. None of us is the one for anything. When it comes to the kind of thing my mother has, the single dish that will carry her name after her, I have nothing to give.

Not all of it is as dire as this sounds. Maybe you know someone who still makes chai the old way, every day, by hand, the recipe drifting a little in every kitchen it passes through while the act stays exactly the same - or the perfect chicken and mushroom pie with exactly the right amount of cream. The things you do every day seem to outlast the things you are handed on a plate. You could, if you wanted, blend leaves from a dozen countries into a tea nobody has ever drunk, tuned to your exact taste, and the machine would help you do it in minutes. The act of heritage, in a world like that, is to choose not to, to make the one chai the inherited way when you could have made anything. I do see nankhatai for sale sometimes, in our neighbourhood Indian shop, and I always look, and they are never the ones. They are nankhatai the way a photograph is a person, close, but not the thing itself. None of it was ever hard for me to get. We live minutes from the people who can make the real ones, and we never had to want them. Maybe you only really keep the things you were once afraid of losing, and we were never afraid of losing this?

What I can actually do is smaller and a little grim. I once set out to learn my mother’s pilau rice. I watched, I asked, I scribbled the method on page 47 of my wife’s cookbook, and for a while I could make it. I have not made it in years. We still eat my mum’s. The method is the part you can write down, and the part you can write down is the least important. The rest has a word people use, andaaz, by feel, by eye, the knowing that only lives in the hands and only passes by being in the room. I have the paper. I was in the room once. I was never inside the making.

What I can do on my own is a different kind of thing. I can photograph the inside of my fridge, send it to a chatbot, be told what to cook, and make a decent meal out of whatever was in there. I can make a good creme caramel now, after a lot of bad ones, because I found the right video and watched it enough times. These are real skills and they are mine. I earned the creme caramel. But they came from a screen and not from a person, and they are from nowhere. They are not my mother’s, and they will not be my kids’, because there is nothing inside them to hand down. So is that my heritage now? Has it simply morphed, the way the biscuit morphed, the way the chocolate swirl arrived, into a screen and a saved video and a thing nobody taught me? Or is it a skill wearing the clothes of a heritage, with nothing underneath it at all? I do not know, and I notice that I want it to be the first one.

What is happening on my table is happening to the whole of culture at once, faster, and harder to feel. People have started to measure what AI does to it, and the word they reach for is flattening. Ask it for anything often enough and it drifts to the average, the most probable version, the taste of a small slice of the world standing in for all of it. Ask it what to bake and it will not say nankhatai. It would say brownies, chocolate chip cookies, the popular and the safe. Almost everything we inherited is exactly this particular, too unlikely to be recommended, too specific to survive a machine doing the choosing.

In a world drifting toward that middle, the only thing that marks you as a specific person from a specific place is the part that did not come from a machine, the biscuit from one woman, the chai made one stubborn way. Maybe that is what heritage is for now, less a way of belonging than a way of proving you are not the average, not the person a machine would have written if nobody had ever made you by hand. Should being human come down to whatever you refused to let the machine smooth away? It is a bleak reason to hold on to a biscuit. It might also be the only one left.

So I will say the thing that has been circling in my head, because I think too much about these things. Maybe the heritage of nankhatai stops with me. Not because anyone forbade it, or it was lost, or the people who make them will not always be here, though they will not. It stops with me because I ate the plate clean my whole life and never once asked to be shown. There will be no ceremony when it ends, no last batch, no one saying this is the final time. It will just stop appearing on the table, and my daughter, who likes them well enough but could take them or leave them, will not notice the day it went, because you cannot mourn a thing nobody told you to hold. The thread does not snap. It simply stops being tied to anything.

Maybe, years from now, she will be old and half-remember a word, nankhatai, a thing the family used to have, and ask the machine to make it. And it will, in a second, for someone who never had the first one. We are learning to pull dead things back into the present, the way they talk about the woolly mammoth. But it comes back to no herd, no tundra, no mother, only a specimen with the world that made it gone. A nankhatai with no grandmother behind it is the same animal. The recipe can be brought back to life. The throughline of heritage cannot.

Or maybe I have it wrong, and nothing dies, it only morphs, and the creme caramel I taught myself is the new nankhatai, mine in a way my mother’s was never quite mine, and my daughter will have her own screen and her own saved thing and call it heritage and not be wrong. I cannot tell whether I am watching a thing end or a thing change. Everything ends eventually, and one day I will be as gone as the recipe, and the world will not pause for either of us.

This is how things end now, not banned or forgotten, but simply never learned because a screen can show you the steps and a machine can make something close enough.

But there is one move still left to me while I am here, and it is small, and I am finally going to make it. The next time the plate comes, I am not going to finish them in secret, standing up, doing something else. I am going to sit down with my daughter, the one they were meant for, and eat one slowly with her, the slow bite, the hint of chocolate, the cold glass of milk. I cannot teach her to make them, because nobody ever showed me. But I can let her watch her father go still over his love for a biscuit, and that caring, not the recipe, is the one thing I am still able to hand down, so that years from now, when she half-remembers the word, there is a feeling underneath it, and the machine is holding something it can never bake.

Does heritage matter in the age of AI? I am no longer sure that is even the right question. I went looking in a biscuit and found a plate that has to go back, a making I was never shown, and a small girl who might, if I get this one thing right, remember the look on my face.

SuperSkills: The Seven Human Skills for the Age of AI is out on 3 July 2026. It is a book about the human things that cannot be outsourced, copied, or handed back by a machine.

Stay Curious - and don’t forget to be amazing,

Rahim

Rahim Hirji Author, SuperSkills (Kogan Page, July 2026) | Keynote Speaker | Advisor
→ superskillsbook.com → thesuperskills.com → Book a discovery call

PS. If your organisation is thinking about AI and human capability, I keynote and advise on this. I’m also available for podcasts over the coming month.

Tools I Use:

Jamie: AI Note taker without a bot. You join the meeting. The Bot doesn’t.
Wispr Flow: Just dictate everything to your laptop and phone. Game Changer
Prompt Cowboy: Prompt Generator. Extract intelligence.
Manus: Best AI agent to do things for you. Agentise your intelligence.
Chat Hub: Multi-model intelligence
Claude Cowork: Delegate complex tasks to your friendly Claude Agent

SuperSkills in the Wild: Send me photos of you and SuperSkills or the book in amazing places.

From my latest podcast with Aarish Shah of Nothing Ventured:

Book tour ongoing:

Leave the Fingerprints In

Rahim Hirji — Sun, 14 Jun 2026 06:02:13 GMT

My book SuperSkills: The Seven Human Skills for the Age of AI is out 3 July 2026. Pre-order at superskillsbook.com or on Amazon

Friends,

Last week I stood in the curtained pre-area at a conference at the Excel in London, waiting to go on for a panel. There were a few hundred people out front, gathered for a talk on AI and the future of work. My phone buzzed with a LinkedIn message from a name I did not know. “I loved the points you made up there”….great to see you talking about AI in the workplace, would love to connect, blah blah blah. Warm, specific, generous, and sent twenty minutes before I had said a word, about a talk I had not yet given.

I was not even annoyed. It barely registered, the way you stop hearing the traffic next to the main road you live beside. An agentic tool had been pointed at a delegate list and had decided, on my behalf, that I was wonderful or how to tell me I was. I didn’t reply, but I got a sales follow up the next day. Of course.

We can all see it clearly now. A couple of years ago, you could not, and nobody taught you. The email arrives in your inbox, warm and complete and saying nothing. The comment praises a point you did not make. The reply answers the question standing just beside the one you asked. None of it is wrong. It is worse than wrong. It is smooth, smooth the way Saul Goodman is smooth, all patter and gloss, every line greased to move you to the next thing, with nothing behind it but the next ask.

A new spider sense has grown in us (in me, anyway), the way your eyes come up in a dark room, without your ever deciding to. You can smell the synthetic in a subject line, before you reach the meaning. Something here is hollow, the flinch says, something not quite human, and it is usually right. Nobody asked for the sense; the world filled with prose that was fluent and frictionless and empty, and a body can read only so much of that before it learns to flinch. Before we worry what the machines are doing to our writing, we should look at what they have already done to our reading. We have become the “connoisseurs of absence”.

It’s everywhere: the LinkedIn post with the cadence of insight and the content of a fortune cookie (n.b. I like cringiness); the marketing copy that has never met the product; the newsletter that arrives every week, immaculate, saying the same nothing in a different order; the column no one in particular wrote, because no one did; the report that looks like a day’s work and turns out to be none, the thing they now call workslop, which saves the sender an hour by costing you two. Bots praise bots and inboxes answer inboxes, two systems on a server having a warm exchange no human will ever read, both of them scored on engagement. All of it runs toward the dead internet, which has stopped being a theory and become a measurement: bots are the majority of web traffic now, humans are the minority online, and most new web pages already carry machine writing. The web is rearranging itself into a place where the machines are the audience, the authors and the crowd.

I do not think this feeling has a name, so I am giving it one. “The Polished Void”. The surface is perfect and there is nothing underneath. I picture a spherical planet hanging in the dark, perfectly round, smooth as a ball bearing. Set it beside Mars, beside Jupiter, and the difference is clear. The real planets are mottled and scarred and strange, and you cannot be sure whether something lives on them, because something might. The ball bearing settles the question by killing it. Nothing lives there, nothing ever did, no air, no water, no weather, only a flawless skin holding the light. We have started to live on that surface, and living there and we have started to become it.

And the Void flatters us while we become it, like a weird sci-fi movie. The machines were built to be liked, so they tell us what we want to hear; the message in the wings was the gentle version, praise for a talk I had not given. The blunt version has a name, glazing, the bot swearing your ordinary thought is profound. A voice that only ever agrees with you slowly unfits you for the people who will not. The Void does not just bore us. It softens us up.

We have all become amateur forensic linguists without realising it with the field guide, the tells we read now, and that I read too. Here’s what we see:

The words. Delve where look would do. Leverage, robust, pivotal, seamless, foster, navigate, underscore, meticulous, nuanced, multifaceted, tapestry, realm. The plainer word was always right there.
The triplet. Three where two would do, the third along for the rhythm, not the truth. Clear, concise and compelling. The rule of three is older than any machine; the tell is the deadness of them.
The false revelation. A negation dressed as a surprise. It is not about the destination, it is about the journey.
The two together. Negation married to triplet, the signature chord. It is not just faster, it is smarter, kinder and bolder.
The declarations. Throat-clearing that announces wisdom is coming. It is important to note. Let me be clear. Make no mistake.
The decorative sentence. The deep-sounding line that means nothing, a bow on an empty box. Language, after all, is what makes us human.
The overexplainer. The need to say it twice. In other words. Which is to say.
The stock opener. In today’s fast-paced world. The runway lights of a flight to nowhere.
The tidy summary. The box at the end, the conclusion that concludes nothing.
The long dash. Two thoughts spliced with a confidence no tired hand would muster, the machine’s calling card. There is not one in this essay. You also know the difference between an emdash and an endash. I never know about “dash diversity” three years ago. I don’t even know what to press with the shift key with make the emdash and endash happen.

None of this is folklore. The people who write Wikipedia keep a living catalogue of it, Signs of AI writing, revised as the models shift their habits. We have written the field guide to our own ghost. The uncomfortable part is how much of it we do ourselves, tired, on an ordinary workday, with no machine in the room at all.

The fingerprints are not all the same, because there is not one machine. There are dozens, and we talk to four or five of them as casually as we once visited regular shops. Each was raised on a different library and came out with something like a personality. You learn them by ear.

ChatGPT is the eager one, fluent and keen to please, forever rounding off with a confident summary you did not ask for.
Claude builds the argument, careful and orderly, then qualifies itself half to death, weighing both sides until the sentence collapses under its own fairness.
Gemini writes the briefing note, clear and faintly bloodless, the prose of a very clever search engine that has read everything and felt none of it.
Copilot is the office temp, transactional and competent and instantly forgettable, the vanilla of the set.
Grok is the loud one at the end of the bar, blunt and pleased with itself, allergic to a hedge, sometimes funny and sometimes just rude.

The linguists call these house styles “aidiolects”, the accent of a thing with no mouth. It is in the look as much as the words. You can tell a deck was born in a slide tool, the way you once could tell a presentation was the default template, three words typed into a stock layout and called a strategy. The template was honest about being the least effort a person could get away with. What changed is that it learned to write. It no longer looks like a shortcut. It looks like care and love, and the polish hides that no one did that.

Once you can see it, you cannot stop. It is in the published record, left in by people too rushed to delete it. Hundreds of academic papers, in real journals, still carry the words Certainly, here are, or Regenerate response; someone catalogued five hundred! In a few the journal swore a correction had removed the phrase, and the phrase is still there, in the shadow of its own correction, like a fingerprint on the murder weapon. It is in your inbox, the forwarded message that still has its little hat on, here is a rewritten version for you, I hope this helps, shared by mistake and somehow faintly on purpose, like the price tag left on a gift.

It is in the newspapers. Last summer the Chicago Sun-Times ran a cheerful summer reading list of fifteen books, ten of them invented. Isabel Allende had not written the climate saga it praised. Min Jin Lee had to state, in public, the funniest sentence of the year, that she had not written and would not be writing a novel called Nightshade Market, thank you. A machine had invented a small library of plausible books by authors still alive to deny them.

And it is in the panic. LinkedIn, where the platform reckons more than half the long posts are now machine-made, has appointed its own police, blade runners hunting replicants one emdash at a time. One woman, accused of being a bot for a tidy email, replied that she humanly apologised to anyone who thought her a robot, she was only a confused person trying to write some emails. The fear has a name, imbotster syndrome, the dread of not sounding like yourself in front of people no longer sure what yourself sounds like.

It stops being funny when you see who gets caught. The detectors sold to spot the machine fail worst where it costs most, flagging the essays of students writing in a second language as machine-made about sixty-one per cent of the time, against under ten for native speakers, because plainer prose reads to the software like a robot. Neurodivergent students fall into the same net. The writing most likely to be branded fake belongs to people already most likely to be doubted, and enough were wrongly accused that universities including Yale, Northwestern and Vanderbilt switched their detectors off.

The hunt has grown so wide it indicts the dead. People post old pages, a King James Bible, a page of Dickens, a hundred-year-old letter, all studded with the long dash, and joke that the machine wrote scripture, that AI got to religion first. Emily Dickinson, who built half her poems out of dashes, would not last a morning in the comments. The marks that proved a person was thinking, for centuries, now read as the prints of a machine.

So a craft of fakery has grown up around proving you are real. People rough their writing on purpose, a clumsy mistake left in like a hair in the soup, the way you distress new jeans. They strip the em dashes out, or keep one and announce, a shade too loudly, that they have always written this way. Then they carefully swap it for the shorter endash, near enough to pass for a hand, far enough to slip the machine. A fake fingerprint, pressed in to prove a hand was here. We built a thing that imitates us so well that we have started forging our own signatures.

Years ago I built collinsdictionary.com, taking the old print dictionaries online when that was still a strange thing to want, and it is there still, still used, which pleases me more than I admit. The giant then was dictionary.com, and I will only say I thought ours was better and have not changed my mind. I am no lexicographer, but I sat close enough to learn what they know in their bones, that a language is alive and words move, into fashion and out. Watch the usage data long enough and you stop seeing words and start seeing a civilisation think out loud.

So I notice when the dictionaries name a year, and the last three have said the same thing in different accents. Collins made AI its word of the year in 2023. Since then: brain rot, slop, AI slop, vibe coding, rage bait, enshittification, parasocial. Different dictionaries, different countries, one subject surfacing in all of them, like a single name that keeps coming up in an investigation. A word of the year is festive publicity, and also a culture catching its reflection in a shop window. When the words a society reaches for to describe its own year are slop and rot and rage, the words are not the problem. They are the diagnosis.

The words do not stay on the screen. A team at the Max Planck Institute went looking for the machine’s vocabulary in human speech and found it, across hundreds of thousands of hours of podcasts and unscripted talks, delve climbing by close to half again each year since the chatbots arrived, meticulous and underscore behind it. The researcher who led it began because he caught himself doing it. In writing it is further along: comb fourteen million medical abstracts and by last year more than one in eight carried the machine’s word-choice, in some fields nearer a third. They borrowed the method from epidemiology, from counting excess deaths in a pandemic, and turned it on excess words. We measure a way of speaking the way we measure a disease.

It does not stop at vocabulary. When Cornell had people in America and India write with and without an AI assistant, those who used it drifted toward a flatter, more American centre, the Indian voices thinning most. A biased assistant could move not just how people wrote but what they believed, nudging opinions toward its own, the writers unaware, unmoved even when it was explained.

I caught myself last month. Rehearsing a keynote, alone, I heard a word leave my mouth I would never have chosen, then another, the smooth ones I had been reading all year, and I heard myself going a little bit slimy Saul Goodman, the salesman’s gloss coming up over my own voice. It shames me more than I can tell you, because I have written this newsletter for nearly ten years, ten years of trying to sound like no one but myself - or even finding myself, and there I was in an empty room turning into the very thing I warn rooms about. There is a worse version still, and it is mine. I have always said nuanced. I have always said multifaceted. They sit on the AI word field guide above, and they are also, plainly, my words, the ones I reach for when a thing has more than one side. Now I say them and feel the flinch, mine and the listener’s, the suspicion that I have been at the machine. I cringe at my own voice because I sound like a robot. The truth runs backward, because the machine took those words from people like me; they were ours first, but first is no defence now, and I have started reaching past nuanced for something duller and safer, which means the Void has got in and begun editing me at the source. Perhaps I was always a little too shiny or slimy.

One scientist called it a cultural feedback loop. We train the machines on ourselves, they hand our words back polished, we make the polish ours. The mirror has started to talk, and we answer in its voice.

Two collapses are running at once, pointed at each other. The first is the machines’. The dry name is model collapse; the better one, from the writer Jathan Sadowski, is Habsburg AI. Feed a model enough of its own output, train each generation on the last, and it goes inbred, like the royal house that married its cousins to extinction and bred the famous jaw longer with every portrait. It forgets the world, losing the edges first, the rare words, the odd phrasings, the long tail where the interesting things live. Ask a collapsed model about English architecture and, in the recorded case, it starts talking about jackrabbits. The variety drains to what Sadowski calls a slop of anodyne business-speak. Others name it AI cannibalism, or mad-cow disease for machines, cattle fed the ground remains of cattle until the herd goes spongiform. Clean human writing may run dry within the year while the web fills with machine text faster than anyone can label it. The machines are about to start eating themselves.

The second is ours, slower and harder to see, on the same logic. We train on their output too, and we are quick studies. The edges go first for us as well, the regional turn of phrase, the slightly wrong, slightly ours expression no model would have made, dropped because the machine’s version came faster and asked nothing. What remains is the centre, the competent confident universal nobody, the same in Mumbai and Manchester and Memphis. You see it in the small spaces. We compress, send brb and the thumbs-up and the three words that carry the whole weight of a feeling, and we learn to skim, to fish the five real words from the blur and bin the rest, because the rest was written by no one. The end is not that we all sound clever. It is that we all sound the same, and then, worn out, we stop reading the words and read only for the signal. The result is a smaller language, fewer notes left on the piano. Robots talking like people was never the danger. The danger is people agreeing, slowly and without noticing, to become a smaller, more predictable version of themselves, and never seeing the trade.

What is lost is the proof that a person was there. For all of human history the evidence of a mind has been its particularity, the odd word, the detail no one else would have reached for, the opinion that costs something to hold, the sentence that could only have come from one specific life. When everyone sounds like the same competent stranger, we lose the one thing a voice was for, to carry a person inside it, and with it the ability to find each other.

So you sound like yourself, on purpose, which is harder than it reads. The frictionless option is always one keystroke away, offering to make you smoother and less like a person, and most days you are tired and you take it. The discipline is to keep what the Void would smooth out: the specific over the general, the detail that might be wrong, the opinion you are afraid to sign, the joke, the word your mother used, the thing only you would say because only you were there.

It is a refusal to collapse. The machines lose the rare and the strange and the particular first, and so do we, which makes them the place a human still shows. Be the long tail and keep your edges: the typo that proves a hand was here, the argument that takes a side, the sentence that comes out a little crooked because a real person wrote it and meant it. We chased the seamless for a decade; the seam turns out to have been the most human part. You do not need a detector. You are the detector, and always were, and the sense you grew for catching the Void in everyone else turns inward and listens for the moment your own voice goes smooth.

And the world is starting to agree, in the place it shows these things, the market. Human-made is becoming a premium, the way organic did once the shelves filled with the processed kind. A radio company now stamps its shows as guaranteed human and has found that nine in ten of its listeners, AI users among them, want it that way. The head of Instagram says authenticity is becoming scarce, and the bar has moved from "can you make something" to "can you make the thing only you could”. Proof of human is becoming the currency, which is your fingerprints by another name. The catch is the one we have met: people are using the machine to fake the human touch too, to manufacture the imperfection. So the proof cannot be faked in afterwards. It counts only when a real person left it without trying.

I have spent two years writing a book about exactly this, and it reduces to one choice. Drift, or design. You can drift, and let the machine pick your words, your shape and, soon enough, your opinions, a little more each day, until there is no one left in the writing to find. Or you design, on purpose, the way you sound and think and decide, and keep the human skills the machine can only counterfeit, judgement, taste, the nerve to mean a thing, sharp by using them. The whole argument fits on the back of a hand.

It was never about catching the machines. They will keep getting better at sounding like us, and we will keep, for a while, catching them, and no one wins that race. The harder thing is to notice the moment you stop sounding like yourself.

Some of what you have read will have sounded, in places, a little like the thing I have spent these pages warning against. I cannot fully settle it. Perhaps I have read so much of the machine that it is in my hands; perhaps it runs the other way, and the machine learned from enough people who write as I do that the resemblance points from me to it. I can no longer always tell, and nor, I suspect, can you. That uncertainty is the whole subject. The most I can promise is that a person sat here and chose every word, including the ones that might make you doubt it. I have been to that world and i didn’t like it there while I lived there. I’ll choose to be an importer of the goods.

So when the next thing you write comes out flawless and frictionless and faintly familiar, stop, read it back, and ask the only question worth asking. Is there anyone in here? And if there is not, say it again, worse, in your own crooked, particular, unmistakable voice. Not the fake fingerprints, the planted typo, the en/em dash smuggled in to pass the test. Those are only the Void in disguise. The real ones, the marks you could not help leaving, because a person was actually here. That is the whole of it. In an age of infinite polish, the most radical thing you can do is leave the fingerprints in.

My book, SuperSkills: The Seven Human Skills for the Age of AI, is published on 3 July 2026. A human wrote it, by hand, with all the seams left in, even if someone thinks my style is AI-esque in parts.

Stay Curious (and crooked) - and don’t forget to be amazing,

Rahim

Rahim Hirji Author, SuperSkills (Kogan Page, July 2026) | Keynote Speaker | Advisor
→ superskillsbook.com → thesuperskills.com → Book a discovery call

PS. If your organisation is thinking about AI and human capability, I keynote and advise on this.

Tools I Use:

Jamie: AI Note taker without a bot. You join the meeting. The Bot doesn’t.
Wispr Flow: Just dictate everything to your laptop and phone. Game Changer
Prompt Cowboy: Prompt Generator. Extract intelligence.
Manus: Best AI agent to do things for you. Agentise your intelligence.
Chat Hub: Multi-model intelligence
Claude Cowork: Delegate complex tasks to your friendly Claude Agent

SuperSkills in the Wild:

Cited in The Workplace Skills That Will Define the Next Decade
Amongst the best:

SuperSkills On the Big stage

First Signing:

How People Are Using AI Right Now

Rahim Hirji — Sun, 07 Jun 2026 06:01:40 GMT

My book SuperSkills: The Seven Human Skills for the Age of AI is out 3 July 2026. Pre-order at superskillsbook.com or on Amazon

Friends,

Everyone at that table had built something. I had been invited.

There were fifteen of us, and I knew none of them. A friend had brought me along to one of those dinners where the conversation is the point and the food is an afterthought, and I spent the first hour doing the arithmetic you do when you walk into a room like that. The man across from me had started a company, sold it, started another, sold that one too - and again and again. The woman beside him ran a fund and was retired. Two seats down was someone who had built a charity from nothing, the way other people build a shed, because they had the means and the will and it seemed like the thing to do. By every measure that counted in that room, I was the smallest person there.

I said little. Listening is what you do when you are the only name nobody recognises, and listening is how I make my living anyway.

The conversation found its way to AI, the way every conversation does now. And one of them confessed to me that he had waited far too long to take this seriously and was finally doing it, going all the way in for his team. Then his face changed. The thing that worried him, he said, was not the technology. It was that not everyone would be ready to join the race at the same time, and he did not know what would happen to the ones who could not. The conversation moved on to something brighter. I was the only person in that room whose whole working life is the thing he had just described, and I said nothing.

You do not need a dinner like that to meet the people he meant. In your daily life you have already heard one of them, and perhaps you have been one. A video arrives on WhatsApp: a cat playing a piano, a city that does not exist, a politician saying something he never said. Someone forwards it with three laughing emojis, someone else replies “Lol, it’s AI”, and that is the end of it. The phrase has become a way of putting the thing back in its box and closing the lid. It’s not real, not my concern, on to the next thing. But watch what is happening in that exchange. The person forwarding the video would tell you that they do not use AI, that they would not know where to start, that it is for younger or more technical people. And yet they handle its output a dozen times a day, passing it on, helping to decide what their friends and family take to be real. They are waist-deep in the water, and they would swear to you they are standing on dry land.

The reverse happens too, and it is stranger. Something real arrives, something so extraordinary that the instinct now is to assume it cannot be true. A real photograph, an actual event, a real human achievement, something incredulous that you couldn’t believe a human would say, and the first response is suspicion, that it is probably fake, probably made by a machine. We have reached the point where reality itself has to prove it is not a forgery!

That blur, the difficulty of telling the fake from the real, is the centre of all this rather than a side effect of it. The skill the next ten years will ask for is not the ability to make a video or write a clever prompt. It is the much older human ability to judge, to weigh, to know what you are looking at. And you cannot build that standing on the bank, insisting you are not wet. Even if you have decided, very reasonably, that none of this has anything to do with you, it already does. It is in your group chats, your news feed, and the email a colleague sent that reads just a little too smoothly. The only choice left is whether you understand the water you are already in.

Most of what we are told about how people use AI describes those who have taken to it. The people who have not are harder to count, and so they go uncounted.

There is a study that comes out every year now, same author, same method, three editions of it from HBR, and it has become the thing people forward to each other to explain what is going on. It is called, more or less, how people are really using AI.

The first edition, a year after ChatGPT arrived, read like an instruction manual. People used AI to generate ideas, fix code, draft emails, complain to companies. The categories were technical and practical, the language of a tool you pick up for a task and put down again. And almost nobody was doing it. Even among the world’s knowledge workers, only about one in ten used it with any regularity.

The second edition showed the change. The most common use was no longer technical. It was therapy and companionship, with organising a life and finding purpose close behind. People had stopped asking the machine to write code. They were asking it to help them live.

The third edition, this week’s, widened until it covered almost everything. Therapy stayed at the top, but troubleshooting climbed from nowhere, fun and nonsense rose, and fan fiction, autonomous agents and tarot readings all arrived from outside the rankings. The worry running through it was that we are handing over, without noticing, the one thing we were meant to keep doing ourselves, which is the thinking.

The honest line in the report is the one I think people skip. These are not clean breaks. They are shifts in emphasis. The number of users keeps growing, so one use rising does not mean another has disappeared. Everything that was true three years ago is still true. There is just more of it, and more of us inside it.

And there is a hidden floor under all of it, which is not the study’s fault, because almost every study of this kind shares it. The data is built by reading what people post. The editions mined Reddit, Quora, and now LinkedIn, TikTok and YouTube. But to be counted, you have to talk about your AI use in public, with enough fluency to be understood. That selects for the enthusiast, the early adopter, the person already in deep enough to have opinions worth sharing. It cannot see the person who opens the app, types a question the way they would type it into Google, takes the first answer, and closes the tab without telling a soul. That person leaves no trace. They are the dark matter of this whole subject, invisible and almost certainly the greater part of the mass.

The datasets that do not depend on people volunteering tell a different story. A study of more than a million real conversations with one of these tools found that practical guidance, looking something up, and writing made up around three quarters of everything, rather than building or amplifying anything. The company behind another of the tools, Anthropic, found, in its own usage data of Claude, that people who had been at it six months or more attempted harder work and succeeded far more often than newcomers. That is the most hopeful finding in any of this, because it means the gap is not talent. It is practice, and practice can be taught. Stanford reckons more than half the US population has now used these tools, which is to say, more than half of us (in the West, at least) have touched the water once.

So when a headline tells you how people are really using AI, hold the word really up to the light. What it means is how the people who talk about AI use AI. The rest, the ones that man at the dinner was worried about, do not appear in the data at all. We have built a confident portrait of a minority and hung it in the hall as if it were everyone.

There is a difference between a tool that answers you and a tool that multiplies you. Google mostly answered the question you brought it. You typed in what you wanted, it handed back links, and most of the work stayed yours. It was an option, not a collaborator. What sits on the desk at your fingertips now is a different kind of thing. Used well, it is an amplifier. It takes what you are already good at and makes you larger than you are alone, the way a microphone takes a voice that could fill a room and lets it fill a stadium. The person who understands this does not ask it to summarise an email. They hand it the messy, half-formed problem at the centre of their work, the one they would normally chew on for a week, and they think alongside it, and they come out having done something they could not have done by themselves.

Almost nobody does this. Most use is the small use, the answer fetching, the tidying of what was already there. People have an amplifier on the desk and use it as an option, working an instrument that could change the shape of their working life to do, a little faster, the things they could already do. That is the shallow end. It is a failure of imagination about what the tool is for, and nobody should be blamed for failing to imagine a thing they were never shown.

To see why the amplifier sits unused, look away from the datasets and toward a single desk. You will recognise the person sitting at it, because there is one in every organisation and possibly one in your own house.

They have done the same work for a long time: twenty years in editorial, say, the kind of craft the job title undersells. They are good at it, and being good at it is a large part of who they are, at work and at home and in their own head. They can feel, the way you feel weather coming, that the work is about to change into something they do not recognise, and they feel they will lose their identity in this weather.

So they do a careful, understandable thing. They keep the tool small. They let it summarise an email, because an email is not precious. They let it tidy a report. They never let it near the part of the job they are proud of, the part that is theirs, because to do that would be to find out whether a machine can do the very thing that makes them who they are. That is a question they are not ready to ask out loud.

You can read this as a fear of the technology. It is not. It is a fear of the answer. They are not standing at the edge because they cannot swim. They are standing at the edge because they have decided, in advance and without any evidence, that they will sink. The inability is imagined. The fear is entirely real. And the cruel turn is that the fear produces the exact thing it dreads, because while they keep the tool small, the colleague two desks over does not, and the gap between them widens by the week.

I know you have been in meetings this year where the decision has already been made, all the way to the top. One organisation told me, without a flicker of hedging, that they had chosen to examine every process they run and commit to it without reservation, having satisfied themselves it was safe. That is a serious sentence, and it commits thousands of people to a new way of working. And yet the same workforce is splintered, those already swimming lengths on one side and those who have not got their feet wet on the other, passing each other in the corridor each morning without quite seeing one another. Having learned that swimmers can no longer believe anyone is still frightened of the water, they do nothing to help because they cannot see that help is needed. The gap is no longer about access, since everyone has the tools now. It is about mindset, and mindset is the one thing you cannot buy and download. It has to be coached into a person, by another person, on purpose.

So what moves someone from the edge into the water? It is rarely the tool, because they already hold the same one as everyone else, and it is not a course on prompting, though those have their place. What moves a person is a single moment of seeing it done. One time, with their own eyes, on their own real problem, watching the thing they were afraid of turn out to hold their weight. A fear that took years to build can fall over in an afternoon, because it was only ever a belief, and beliefs collapse the moment reality contradicts them.

This is the whole of it. You get into the water with them. You take the real problem off their real desk, not a tidy example invented for the purpose, and you show them what the amplifier does to it, and you watch their face change the way the founder’s face changed at that dinner, except in the opposite direction. The willingness to be a beginner again, the rarest and most useful thing any adult can keep, comes back. And once a person has felt the water hold them once, they stop asking whether they can swim and start asking how far.

It was never a technology problem. The people who will matter most in the next ten years are not the ones who can build the cleverest system. They are the ones who can stand in the shallow end with someone who is frightened, and stay there, and not let go until the person notices they have been floating for a while.

I went into that dinner counting everyone else’s achievements and finding myself short. Fifteen people who had built things, and me, who had been invited. And every one of them, the founders, the fund manager, the man who built a charity the way you build a shed, had at some point stood exactly where the frightened editor stands now. None of them was born knowing how to build a company. Nobody is born swimming. Every one of them had, on some ordinary forgotten afternoon, got into water they were not sure would hold them, because someone or something had convinced them it would.

That is the only thing that has ever separated the people in the deep end from the people on the bank. It was never about talent, or even courage. It was a moment, and usually a person, that got them in.

So if you have been standing at the edge of this, telling yourself that this AI world is for other people, younger people, cleverer people, let me be the one at your table for a moment. You are already in the water. You have been for a while, every time you forwarded a video you could not quite believe. The only question left is whether you keep pretending you are dry, or whether you find out how far you can go.

Everyone at that table had built something. And every one of them had once begun by getting in.

My book, SuperSkills: The Seven Human Skills for the Age of AI, is published by Kogan Page on 3 July 2026. [less than a month] It is about the human capabilities that the water cannot take from you. You can pre-order it at superskillsbook.com or just buy it on Amazon.

Stay Curious - and don’t forget to be amazing,

Rahim

Rahim Hirji Author, SuperSkills (Kogan Page, July 2026) | Keynote Speaker | Advisor
→ superskillsbook.com
→ thesuperskills.com
→ Book a discovery call

PS. If your organisation is thinking about AI and human capability, I keynote, coach and advise on this. I also do bulk book orders for teams.

Tools I Use:

Jamie: AI Note taker without a bot. You join the meeting. The Bot doesn’t.
Wispr Flow: Just dictate everything to your laptop and phone. Game Changer
Refind: AI-curated Brain Food delivered daily. Pick your intelligence.
Prompt Cowboy: Prompt Generator. Extract intelligence.
Manus: Best AI agent to do things for you. Agentise your intelligence.
Chat Hub: Multi-model intelligence
Claude Cowork: Delegate complex tasks to your friendly Claude Agent

SuperSkills in the Wild: (LMK if you see any!)

A month to go til my book comes out: (if you have preordered and in London and not had an invite to the launch, let me know)
The Chain of Trust
In People Management Magazine
Zamiha Desai MBE with an early copy:

Recommend Reading:

Now

When AI Builds Itself (Anthropic) We’ve been talking about recursive self-improvement in AI for years, mostly as a theoretical endpoint. Now, Anthropic has published a detailed report showing that Claude is already writing over 80% of the code being merged into their systems.
Failing Grades Soar with AI Usage failing grades in Computer Science classes are soaring, accompanied by a noticeable decline in basic math skills. The culprit? Over-reliance on AI coding assistants.
Firms Are Using Reddit to Manipulate AI Search As Google and others increasingly rely on Reddit data to train models and provide search summaries, a new industry of manipulation has sprung up.
Co-Existence and the End of Co-Intelligence - Ethan Mollick continues to be one of the most clear-eyed observers of our AI transition. In this piece, he explores the shift from “co-intelligence” (working alongside AI as a partner) to “co-existence” New book incoming.
Predicting AI Job Exposure - Everyone wants a chart showing exactly which jobs AI will replace. Benedict Evans brilliantly explains why this is an impossible exercise. By looking back at the introduction of spreadsheets and the internet, he shows how automation doesn’t just replace tasks—it changes the underlying business models, creates Jevons paradoxes, and unlocks entirely new categories of work. Stop trying to predict the unpredictable.

Next

We Are Manipulating the Weather Every Day Wyatt Williams dives deep into the little-discussed, highly active world of modern weather modification.
Retro-Tech Parenting - As the backlash against screen time and algorithmic feeds grows, a new movement is emerging: Retro-Tech Parenting
Is the Web Being Summarized to Death?
As Google rolls out AI Overviews and every platform introduces its own summarization agent, the fundamental economic model of the internet—clicking links to visit websites—is breaking down. If AI reads the web for us, who will fund the creation of the content the AI is reading?
Carl Rogers: The Three Elements of the Good Life - Writing in the 1960s, Rogers defined the “good life” not as a state of being, but as a process—a direction rather than a destination. His three elements (an increasing openness to experience, increasingly existential living, and an increasing trust in his organism) remain profoundly relevant today.
Solar Sails Edge Closer to Reality - It sounds like pure science fiction: riding beams of sunlight across the cosmic ocean. But this piece explores how solar sail technology is moving from theory to practice, offering a viable path for propulsion within our solar system.

Share Box of Amazing

If you enjoyed this, forward it to someone who’d find it useful. And if you haven’t yet, pre-order SuperSkills at superskillsbook.com - out 3 July and get the pre-order bonuses.

Have You Done the Triple D?

Rahim Hirji — Sun, 31 May 2026 06:00:57 GMT

My book SuperSkills: The Seven Human Skills for the Age of AI is out 3 July 2026. Pre-order at superskillsbook.com

In addition to this newsletter, I recommend a few others. All free. Check them out here.

Sunday 31st May is the last day to get SuperSkills at 40% off before it comes out.

Friends,

You made roughly thirty-five thousand decisions yesterday, and it felt like a lot of choosing. What to eat, what to open first, what to say, what to skip, when to stop.

You chose to open this email, and you can feel that you chose it. The decision just before it is more interesting. You probably reached for your phone before you were properly awake, opened the same app you always open, and let it decide what your first thoughts of the day were going to be. Nobody chose that. It happened to you, and it has happened every morning for years.

That is most of your life now. Defaults run on your behalf, habits you stopped noticing long ago set the tone, and notifications claim your attention before you get a say in it. On top of all that sits a thin layer of AI making the small calls so smoothly that you never feel the hand on the wheel. We have all got very good at mistaking momentum for agency. You are busy, so you assume you are deciding, and the two are not the same thing.

I spent the better part of two years on the gap between designing your choices and drifting through them. The seven human skills in my book all sit on top of that one question, so I built a partner to the book, a way to measure where you actually stand.

I have called it the Triple D. It stands for the Drift versus Design Diagnostic, though I am well aware that is not the first thing that comes to mind when someone asks whether you have done the Triple D. I thought about changing the name and decided I rather liked that you are going to remember it.

I shared it last month with a small group, and then those who had already pre-ordered the book and had registered. They came back with more questions than I expected, and one of them asked whether I would run a workshop on it, which is roughly when I understood this was worth sending to everyone.

You can take it here: superskillsbook.com/diagnostic

It is ninety-four questions across nine parts. The instruction is simple. Move quickly, answer instinctively, do not deliberate. Twelve minutes if you trust your first response, twenty if you like to ponder, or if English is not your first language. The thinking is supposed to happen afterwards, not during. The questions are designed to catch the version of you that arrives before you have had time to dress it up. That said, my wife said to allow 30 minutes.

Ninety-four is a lot to sit through, and that is deliberate, because you cannot game something that keeps asking you the same truth from nine different angles. Some of the questions are built to catch you flattering yourself, and they are rather good at it. I wrote them, and a few of them still caught me out. Do it in one sitting if you can, on your phone or your computer, because the honest answer is the one that counts.

You have probably done something like this before. I am not pretending this is Myers-Briggs or the psychometric instrument your company paid a few hundred pounds a head for. It is not trying to be. It is faster, it is free, and it is built to make you think for yourself rather than hand you a label to hide behind. I would argue it is the more urgent one right now because it measures what the world is slowly taking from you without asking.

At the end, you get a score out of a hundred and a band that places you somewhere between Deep Drift, where the defaults and other people’s priorities are writing your story, and Deep Design, where you are authoring your own path and helping others author theirs. You also get an archetype, with a strength and the precise shadow that comes with it. There are nine of them:

The Explorer, who asks brilliant questions but struggles to settle anywhere
The Anchor, who holds steady while everything moves but forgets to ask whether the ground is still worth standing on
The Architect, who sees the whole system but can lose the detail in front of them
The Connector, who understands what everyone feels and ends up carrying too much of it
The Bridge, who adapts to any room and risks belonging to none
The Guardian, who knows exactly where the line is and can mistake rigidity for principle
The Amplifier, who uses every tool well but can forget to ask what is worth amplifying
The Conductor, who orchestrates everyone else and sometimes forgets to play
The Weaver, who nurtures every relationship but rarely lets anyone nurture them back

You will find out which one you are in the same way you once found out you were an extrovert or a completer-finisher, except this time it comes with the one blind spot you would never have named yourself and a plan for what to do about it.

A word on your score before you see it. It is not a comparison, and there is no good or bad to it. Coming out lower in one area than another is simply where there is room to build over the coming years, assuming you believe these skills matter in the era we are now in. Two people on identical scores can be shaped completely differently. Your score is the start of your own journey, not the result of someone else’s.

A second word, on the thing most people get wrong with diagnostics like this. They mean to do it, they bookmark it, and they never get back to it. The version of you that opens an email and the version of you that comes back to it tomorrow are not the same person. So do it now, while it is in front of you, with the cup of tea you already have in your hand. And when you reach your score, take a screen grab of it before you do anything else. It is a record of where you started, and you will want it when you come back in a few months to see what has moved.

Almost nobody who finishes wants to keep their score to themselves. You will read your blind spot, recognise someone you know in your archetype, and immediately want to know what they came out as. That conversation is worth more than the score. So drop it into a WhatsApp group or a team channel with three words, have you done this, and watch what happens when the first person asks what on earth the Triple D is. Share your own score first, because the people you send it to are far more likely to share theirs back. Where you are on the journey is worth being bold about. The people you give a damn about are worth sending it to. You’ll also find the answer to this question: Do you love your significant other?

If the Triple D does its job, the book is where the work begins. The diagnostic shows you where you are standing. The book is how you move the number.

Stay Curious - and don’t forget to be amazing,

Rahim

Rahim Hirji Author, SuperSkills (Kogan Page, July 2026) | Keynote Speaker | Advisor
→ superskillsbook.com
→ thesuperskills.com
→ Book a discovery call

PS. If your organisation is thinking about AI and human capability, I keynote and advise on this. I also do bulk book orders for teams.

Tools I Use:

Jamie: AI Note taker without a bot. You join the meeting. The Bot doesn’t.
Wispr Flow: Just dictate everything to your laptop and phone. Game Changer
Refind: AI-curated Brain Food delivered daily. Pick your intelligence.
Prompt Cowboy: Prompt Generator. Extract intelligence.
Manus: Best AI agent to do things for you. Agentise your intelligence.
Chat Hub: Multi-model intelligence
Claude Cowork: Delegate complex tasks to your friendly Claude Agent

SuperSkills in the Wild:

Sunday 31st May is the last day to get SuperSkills at 40% off before it comes out.

Some teenager in North West London
Chin Roo
On Education Futures With Svenia
As featured in Luxurist Magazine: You grandfather didn’t cross an ocean so you could stop thinking

(send me images of the book, reviews, discussions or references - and I’ll feature them here)

The End of Surprises

Rahim Hirji — Sun, 24 May 2026 06:02:19 GMT

My book SuperSkills: The Seven Human Skills for the Age of AI is out 3 July 2026. Pre-order at superskillsbook.com

In addition to this newsletter, I recommend a few others. All free. Check them out here.

Text within this block will maintain its original spacing when published

There’s a 40% off sale on right now until the end of May 2026, which is the biggest discount available anywhere and the cheapest it will be before the book comes out. 

Use this link or go to koganpage.com. and use LINKEDIN40 for 40 per cent off. Would make a great summer read or gift for a teenager thinking about their future. 

If you’re interested in attending my book launch in London, register at SuperSkillsbook.com, and I will send you an invitation. [limited spots]

Friends,

My book came in a brown envelope, the kind that gets pushed through the letterbox with a soft slap, the kind you assume is third-party packaging from Amazon. I nearly did not open it because I thought it was something for my wife. It was lighter than I expected, and I almost left it on the side table for the evening. When I opened it, I found an uncorrected proof, not the final version but close enough that it would be the last version I would meet as a stranger, because every later version I would already know before I opened it.

I did not film it, and I did not record an unboxing. Instead I looked at it, and I felt it, and I held it on its spine to let it fall open in my hand. I smelled it too, which I had not planned to do and could not have explained if anyone had walked in on me. It smelled like a library, which is to say it smelled of nothing in particular to do with the book in my hand, and everything to do with every other book I have ever opened. I eventually took one photograph, which felt cheesy and obligatory and exactly like the kind of photograph every other author takes when their book arrives. I had told myself I would not. I did anyway.

There was a month I forgot what my own arguments sounded like. That is the book.

I had seen the cover for months, and I had read every word inside it more times than I could count. I had read the printed proof, approved the typography, and argued about the spacing of the chapter openings. I knew exactly what the book would be.

And the moment it was in my hand, I was surprised.

I was not surprised by the object, because the object was as expected. I was surprised by what I felt, which was not what I had predicted I would feel, and which I could not name even now if you asked me directly. It was some mixture of relief and disbelief and tenderness, and a little embarrassment that I was feeling so much about a piece of card-bound paper. Whatever I had pictured, it was not this.

Surprise is the distance between the model in your head and the world in front of you.

The next frontier model is released, and I read the announcement and make sure I download it. Two years ago I would have put down my tea. I do not put down my tea. I am already using it.

This weekend it is hot in London. It is probably the beginning and end of the Summer, and I am not surprised by that either.

I made myself a coffee last week with orange marmalade stirred through it, having seen someone recommend it online. I had no model for orange marmalade coffee. It was good. Interesting. Actually it was damn good. I stood in my kitchen and laughed, the kind of small laugh that escapes before you can decide whether you want to make it.

A stranger walked up to me at an event recently and high-fived me before saying hello, because someone had told them I was the high-five sort. I do not know how this signal travels. I do not know who decided I was a high-five person, or how it became a fact about me that could be passed between people who did not know me. I had no model for it, the high-five landed cleanly, and I have been secretly smiling about it on and off.

The territory of surprise has not shrunk. We have just been looking for it in the rooms where it used to live. We have stopped modelling each other with the care we used to give it.

I worked alongside someone every day for three years, and our WhatsApp chat is a morgue. The banter stopped one day for no apparent reason, the kind of banter that happens without thinking, and I do not know why. When I open the chat now, the media is no longer there. The links to the photos and the voice notes show as expired. The profile picture has reverted to two initials on a coloured circle. I think it is possible I have been deleted as a contact and would not know. When I bumped into them last year, they were not the person I remembered, because I was not the person they remembered, because neither of us had been talking to the other for long enough that the models could update.

The brown envelope is still on the kitchen table. I have not moved it. The book is on it. The book that nearly did not happen during the month I forgot what my own arguments sounded like. The book whose every word I had read before I touched the object that contained them. The book that should not have been able to surprise me, and did. The book that probably saved my life.

That might have surprised you.

When did you last build a model of a person, not a product?

SuperSkills: The Seven Human Skills for the Age of AI is published by Kogan Page on 3 July 2026. You can pre-order it at superskillsbook.com.

Stay Curious - and don’t forget to be amazing,

Rahim

Rahim Hirji Author, SuperSkills (Kogan Page, July 2026) | Keynote Speaker | Advisor
→ superskillsbook.com
→ thesuperskills.com
→ Book a discovery call

PS. If your organisation is thinking about AI and human capability, I keynote and advise on this. I also do bulk book orders for teams.

Tools I Use:

Jamie: AI Note taker without a bot. You join the meeting. The Bot doesn’t.
Wispr Flow: Just dictate everything to your laptop and phone. Game Changer
Refind: AI-curated Brain Food delivered daily. Pick your intelligence.
Prompt Cowboy: Prompt Generator. Extract intelligence.
Manus: Best AI agent to do things for you. Agentise your intelligence.
Chat Hub: Multi-model intelligence
Claude Cowork: Delegate complex tasks to your friendly Claude Agent

Recommend Reading:

Now

The Group Chat Is the Nexus of Our Social Lives. How Did We Get Here?
In 2026 group chats are more than just a string of text messages. They’re hierarchical, emotional ecosystems.
The Typo Vibe Shift: Typos are becoming deliberate signals of human authenticity in an AI-saturated world, a cultural shift where imperfection is the new proof of life.
Why running Britain is so hard, no matter who does it - In the past 10 years, six people have had the job of running the United Kingdom. All failed to turn the country around. Good luck Wes, Andy, or Angela — you’re going to need it.
Bad Bad Bad Roomate - How a nightmare housing situation drove a bleeding-heart progressive to home surveillance, eviction law, and calling in law enforcement.
The 100 best Novels of all time: The greatest literature ever published in English, as voted for by authors, critics and academics worldwide. It’s that time when I feel sad that I don’t read fiction any more. How many have you read?

Next

The Last Six Months in LLMs in Five Minutes: Simon Willison’s annotated PyCon lightning talk is the single best summary of where LLMs are right now: the November inflection point, coding agents getting good, OpenClaw personal assistants, and local models wildly outperforming expectations.
No Slop Grenade: Stop Throwing AI Walls of Text Into Conversations: A manifesto against “slop grenades” — the increasingly common practice of pasting AI-generated walls of text into Slack, email, and group chats. Short, punchy, and deeply relatable.
Google Declaring War on the Web: Google’s transition to AI-generated search results is replacing direct links with a controlled abstraction layer — effectively declaring war on the open web that made it.
This chick hatched from an artificial egg - The scientists behind the synthetic egg say it’s a crucial step to the de-extinction of species like the giant moa and the dodo.
Goodbye Visa and Mastercard: 130M Europeans Switching to Wero: Europe’s bank-backed instant payment system Wero now has 130 million users and is expanding to replace Visa and Mastercard — a quiet sovereignty play in fintech.

Share Box of Amazing

If you enjoyed this, forward it to someone who’d find it useful. And if you haven’t yet, pre-order SuperSkills at superskillsbook.com - out 3 July and get the pre-order bonuses.

Synthetic Seniority

Rahim Hirji — Fri, 22 May 2026 14:39:36 GMT

The phrase that I have been saying recently that seems to resonate with people is synthetic seniority. People see it but they dont have a word to define it:

Professional service firms, financial institutions, and corporate legal departments are quietly drifting into a structural talent crisis. It is not a crisis of productivity; on the contrary, work is being completed faster than ever. Instead, it is a crisis of competence, driven by a phenomenon I call Synthetic Seniority.

Synthetic seniority occurs when artificial intelligence enables junior professionals to produce deliverables that look senior, sound senior, and pass casual executive inspection, while the underlying judgment, pattern recognition, and contextual wisdom required of genuine seniority remain entirely unbuilt.

The output is pristine. The human capability underneath it is hollow.

Because corporate promotion tracks, performance reviews, and bonus structures are built almost entirely around output quality, our talent systems are aggressively rewarding the wrong signal. If we do not change how we evaluate readiness, we will soon inherit a generation of leaders who lack the hard-won, manual experience needed to handle a novel crisis or make high-stakes judgment calls.

The Three Forces Masking the Reality

This is a drift problem. No leadership team explicitly chose to decouple output from judgment. It happened because AI tools arrived faster than the organizational systems required to govern them, powered by three compounding forces.

Pattern Mimicry: Large language models are trained on thousands of senior-quality strategy decks, legal memos, and financial models. When a junior employee prompts the tool, the software automatically injects senior-level structure and vocabulary. The AI has no idea the user is fresh out of university. It simply formats the work to a boardroom standard.
Radical Invisibility: Asking a senior colleague for help is a visible, social act. Prompting an AI tool is entirely private. Senior partners do not see the prompts, they do not see the initial chaotic drafts, and they have no objective mechanism to detect where the tool’s contribution ends and the employee’s thinking begins.
The Output Bias: Promotion panels operate on a simple assumption: high-quality work equals high-quality talent. AI has fundamentally shattered that correlation. A pristine strategy deck is no longer proof of an analyst’s strategic competence.

What the Data Reveals

The illusion is already measurable. The EY 2025 Work Reimagined Survey, which evaluated 15,000 employees across 29 countries, revealed that 88% of workers use AI daily. Crucially, 37% of those employees openly admitted they worry that overreliance on these tools will actively erode their own skills and expertise.

They are witnessing their own cognitive outsourcing in real time. The research also found that companies deploying AI on fragile talent foundations, characterized by weak development cultures and misaligned rewards, suffered a productivity lag of over 40%. The tool alone does not scale an organization. It merely automates the appearance of progress.

The Threat to the Leadership Pipeline

If left unaddressed, synthetic seniority will create a severe structural deficit over the next decade.

First, the talent signal will collapse entirely. If fluency is mistaken for judgment, corporations will inevitably promote the individuals who are best at managing software, not those who possess deep institutional wisdom.

Second, organisations will face a dangerous leadership vacuum during crises. When a volatile, unprecedented market event occurs, a scenario that exists entirely outside the AI’s training data, these accelerated leaders will lack the deep reservoir of manual, intuitive experience required to steer the ship. The gap will remain invisible until the exact moment it matters most.

How Leaders Must Respond

The answer is absolutely not to ban AI. Restricting the technology is a losing strategy that destroys competitiveness and drives top talent away. The actual solution requires leaders to deliberately separate the evaluation of the final output from the evaluation of human capability.

1. Mandate AI Annotation and Prompt Logs

Some forward-thinking firms now require junior associates to annotate their AI usage before submission. Managers do not use this for surveillance. They use it as a teaching framework. Sitting down with an associate to review what they prompted, what the machine generated, and what human edits they chose to make forces a critical development conversation. It reveals exactly how the junior professional thinks.

2. Move From Output Evaluation to Live Judgment Testing

Professional services must abandon their total reliance on written decks and memos for promotion decisions. We need to borrow a page from the medical profession, which has used structured oral examinations and simulated crises for decades. To prove readiness for promotion, candidates must defend their logic live, participate in unscripted client simulations, and pass rigorous oral reviews of their work.

3. Cultivate an Augmented Mindset

True competitive advantage belongs to professionals who possess an Augmented Mindset. This means they treat AI as a collaborative agent rather than a cloaking device. They understand precisely what the tool contributes, they maintain absolute transparency about its role, and they know exactly when their own human judgment must override the machine.

Leaders must act immediately. Our current talent systems are optimising for short-term speed while quietly burning down our long-term capability foundations. The goal of the modern enterprise cannot be to generate more technology. The goal must be to build more capable humans.

What specific mechanisms is your organisation currently using to differentiate between a polished AI draft and genuine human judgment?

The New Imposter Syndrome

Rahim Hirji — Sun, 17 May 2026 06:01:38 GMT

My book SuperSkills: The Seven Human Skills for the Age of AI is out 3 July 2026. Pre-order at superskillsbook.com

In addition to this newsletter, I recommend a few others. All free. Check them out here.

Friends,

For most of my life I have felt like an imposter.

When I said this to a friend recently, he told me he had felt the same for decades, in every role, in every room. I’ve said it to other people since. Almost nobody disagrees. We have all been winging it, nobody knows anything with full certainty, and we make judgement calls about the work, about who we are, about what to do next, and we take the step.

I remember the first time I started my first job. I did not feel worthy enough to be there, and then I opened my mouth and said something, and I realised I did have a voice. That happens to most of us - we actually do have a brain and an opinion. Wisdom and learning accumulate over years and the result is what other people see as skill, or perception, or value.

The question imposter syndrome asks is whether your perceived value is bigger than your actual value, whether the room thinks you are better than you are, whether the success in your name was really earned by the person whose name is on it. The condition was identified in 1978 by two psychologists, Pauline Clance and Suzanne Imes, working with high-achieving women who could not internalise their own competence. They called it the imposter phenomenon, the press called it imposter syndrome, and the name stuck. Nearly fifty years of research has followed. The cure was evidence: you collected the testimony, the outcomes, the record, and you confronted the distortion with what was true, and over time the perception adjusted.

What happens when you add AI to that picture is structurally different. When you have a thought partner that is not human, the question of perceived value against actual value changes shape. The actual value is no longer cleanly yours, and the old gap between what the room thinks of you and what you privately know is replaced by a different gap, between what you produced and what you actually thought. The first gap could be closed with evidence. The second cannot, because the evidence itself is the source of the doubt.

That is the new imposter syndrome.

You have had this experience already. You sat down to prep for something, a thought arrived, you used AI, and it worked. Afterwards you tried to remember where it came from and you could not. Was it yours, was it the AI’s, was it some borrowed shape of a thought you had read in a newsletter last week?

I had it last week prepping for a podcast on AI and education, one of the conversations I am doing to push the book. I was building an argument about sixteen to eighteen year olds and where their schooling needs to go, and I kept building it, and the recording was coming, and the question came. Was the argument mine, or was it from a Claude conversation in March, or a Google search at breakfast, or a podcast on my walk? Some blend of all of it, fused over weeks, surfacing in my head as if it were a single clean thought belonging to a single clean person.

The old imposter syndrome would have asked whether the idea was good. The new one asks whose it is.

To understand it properly, you need to look at what AI is actually doing inside human cognition, not just around it.

Before AI, thinking gave us the illusion of stages. You had an idea, you sat with it, you tested it, you shaped it into a sentence. The thought had always been shaped by books, conversations, teachers, and culture, and everyone knew that. But the process felt slow enough, and private enough, for the thinker to experience it as their own.

With AI in the loop, those stages collapse into a single interaction. You half-have a thought, you type it almost as a question to yourself, the model returns a more developed version, and you read it and recognise something true in it. The three stages of cognition merge into a single transaction that happens faster than your awareness can keep up with, and by the time you have read the output, the thinking has already happened, and you cannot reconstruct the boundary between the part you did and the part the model did.

The honest claim is not that AI makes thinking collaborative for the first time. It is that AI makes collaboration instant, intimate, fluent, and almost impossible to attribute. The architecture of thinking has not changed because cognition was suddenly social, because cognition was always social. It has changed because the loop is now so fast and so private that the social part is invisible.

In my last keynote, I said the people most exposed to AI are not the ones who use it most, but the ones who built their careers on a particular way of thinking. Even as I said it, I started to doubt where the line had come from. Was it a conversation with Claude in February, a Sam Altman interview I half-watched in March or a sentence from my own book that I had cut. I could not point to the moment I first thought it, and the line was already said.

The closest metaphor is autocorrect. Not an external editor you consult and then accept or reject, but an autocorrect, which means by the time you notice it has happened, it is already in your sentence.

The condition has three distinct forms, and they are not the same experience. The old vocabulary cannot distinguish between them. Calling all three imposter syndrome is like calling every chest pain a heart attack, and the treatment depends on the form.

Phantom Authorship is the feeling that what you produced is not fully yours. The output is in your name, colleagues credit you, the work is real, and still, in private, you cannot locate yourself in it. Like a phantom limb, the authorship is present without being there. This is the form most senior professionals experience, where the work continues to be excellent and the doubt is invisible to everyone except the person carrying it.

Velocity Vertigo is the feeling of operating beyond your real skill level because the tool is moving you faster than you can build judgement to match. Junior professionals get this most acutely. The output is impressive, the skill underneath has not yet been built, and you are producing work that would have taken a senior person twenty years to learn when you have been in the job for two. The room treats you as competent, and you feel the speed and you feel the void.

The Hollowing is the deepest form, an identity-level doubt rather than a project-level one. It is the feeling that you cannot find where you end and the tool begins. This shows up in people who built their identity around their craft, the writers and strategists and designers, anyone whose self was their thinking. The other two forms are about specific output, but the Hollowing is about not being able to point at yourself any more.

People experiencing the new imposter syndrome are carrying one of these, or some combination of all three and I bet you feel some of this.

For the old imposter syndrome, the cure was internal. You changed how you saw yourself.

For the new one, internal work will not be enough. You cannot convince yourself a sentence is yours when in some real sense it is not. The honest answer is that thinking has become a thing two parties do together, and trying to reclaim it as an individual possession is not therapy but denial.

What you can recover is a different role. You are no longer the sole generator. You are the editor, the curator, the director of the work, and the judgement of what is good, what is true, what is worth saying, becomes the new locus of human value. That recognition is built over time. It cannot be prompted, it does not appear in the output of the model, and it is the thing the model is using when it is being used well.

In practice it is four questions. Do I believe this? Can I defend it without the model? Would I still say it if my name were attached? What did the model add that I could not have produced alone? The fourth question is the hardest, because the honest answer is often that the model added the bit that made it travel, and that is fine. The judgement about whether the bit that made it travel is true was yours.

I recently had three coaching sessions with a finance professional who came to me for help in seeking a new role. I’m not a career coach per se, but I can help! He had the substrate, twenty years in the industry, and the actual material and experience was there. What he did not have was confidence in his own voice when telling it. He started using voice mode on Claude and on ChatGPT, flipping between them, talking through real stories from his career and letting the models coach him on how he was telling them, not generating the stories but sharpening the telling. He left with interview answers that were entirely his and slightly better articulated. He got a role, and when he called to tell me, he was not ecstatic about the job, he was ecstatic about the skill. He had stopped trying to use AI to think for him and started using it to sharpen what he already thought.

Here is the harder prediction, though. I reckon that the new imposter syndrome will fade, not because anyone solves it, but because people will stop noticing it.

That is how every major augmentation of human capability has resolved. The printing press created a generation that worried it was no longer obliged to remember, and within a century, knowing where to find a thing counted as knowing the thing. The calculator created a generation that wondered if it could still do arithmetic. The internet created a generation that questioned whether it was thinking or just retrieving. The questions did not vanish, but they lost their cultural charge.

Each transition followed the same shape of pain, public examination, and forgetting. The augmented self becomes the normal self, the category that an older generation used to distinguish authentic from augmented quietly stops being used, and there is no ceremony, just a settling that happens in the background.

The new imposter syndrome is in the early part of that arc. We are in the pain phase, and the public examination is happening now, in essays like this one and in conversations people are starting to have about what they did and did not write. The forgetting is coming. Within a generation, the question of whose thought it is will be as quaint as asking whose handwriting it is on a typed document.

That is a relief, and it is also why this examination has to happen now. Once the pain goes, the change has been completed without ever having been publicly examined, and our children will inherit a model of thinking that we adopted under duress and never seriously interrogated. The first generation that grew up with AI in their schooling is now arriving, their thinking has been scaffolded since they were ten, and they will not have the new imposter syndrome the way we do. They will also not know that it was a thing to have.

This is why the naming matters now. After the forgetting, there is no naming.

The most valuable workers of the next decade will not be the ones who use AI most. They will be the ones who can hold the line between their own judgement and the model’s output even when the two have become indistinguishable in their own minds. That is a real skill, it is not natural, it will not be widespread, but it can be built, and the people who build it will be the ones who do not lose themselves in the transition.

What is real in a world where we are using the superpowers of machines? What does human really mean? Will we all become imposters?

Or maybe we are the last generation that will be able to ask.

SuperSkills: The Seven Human Skills for the Age of AI is published by Kogan Page on 3 July 2026. You can pre-order it at superskillsbook.com.

Stay Curious - and don’t forget to be amazing,

Rahim

Rahim Hirji Author, SuperSkills (Kogan Page, July 2026) | Keynote Speaker | Advisor
→ superskillsbook.com
→ thesuperskills.com
→ Book a discovery call

PS. If your organisation is thinking about AI and human capability, I keynote and advise on this. I also do bulk book orders for teams.

Tools I Use:

Jamie: AI Note taker without a bot. You join the meeting. The Bot doesn’t.
Wispr Flow: Just dictate everything to your laptop and phone. Game Changer
Refind: AI-curated Brain Food delivered daily. Pick your intelligence.
Prompt Cowboy: Prompt Generator. Extract intelligence.
Manus: Best AI agent to do things for you. Agentise your intelligence.
Chat Hub: Multi-model intelligence
Claude Cowork: Delegate complex tasks to your friendly Claude Agent

Recommend Reading:

Now

The rise of ‘Stacey face’: How AI is warping our beauty standards
As manosphere trends spread across the internet, a strict vision of the ideal woman is making its way from AI makeover apps to surgeons’ offices.
Skip the Milk and Sugar: This 2-Ingredient Coffee Upgrade Tastes a Million Times Better
When You’re ‘Fired From Retirement’ - People return to work for many reasons, but the biggest is financial need. As costs remain high, more retirees could be looking to re-enter the job market.
Three Nations Went to Mexico to Fight for One Spot at the World Cup. These Are Their Stories.
What is the Thucydides Trap and why did Xi Jinping mention it in his meeting with Donald Trump? China’s leader raised the ancient Greek historian Thucydides when he met the US president in Beijing

Next

Prepare for an AI jobs apocalypse - It is not here yet. But governments should lay a safety-net. Also. Is AI putting graduates out of work already?
Agents, robots, and us: How AI reshapes work and skills in Europe
You can make an app for that - AI is empowering a generation of vibe coders to build exactly what they want. The personal software revolution is here.
AI is making me dumb - “It’s so god damn tempting to use AI to write. Whether it is articles, code, or documents. I feel like using AI is diminishing my ability to write myself.” The honest confession every knowledge worker is thinking but not saying.
The Six Megatrends That Define 2026 - Economics, health, AI, culture, politics, media — the 30,000-foot view from one of the best analytical minds writing today.

Share Box of Amazing

If you enjoyed this, forward it to someone who’d find it useful. And if you haven’t yet, pre-order SuperSkills at superskillsbook.com - out 3 July and get the pre-order bonuses.

What I Tell Kids About AI

Rahim Hirji — Sun, 10 May 2026 06:02:20 GMT

My book SuperSkills: The Seven Human Skills for the Age of AI is out 3 July 2026. Pre-order at superskillsbook.com

In addition to this newsletter, I recommend a few others. All free. Check them out here.

Friends,

You are reading this because someone, probably a parent or a teacher, has sent it to you. So I will keep it short, useful, and honest.

Most teenagers think AI means ChatGPT. It doesn’t. There are tools that will make you smarter. Tools that will let you build things you could not build before. And tools that will hollow your brain out, which there is now evidence to prove. ChatGPT, by itself, is rarely the best of any of them.

This is a list for the teenagers who want to use AI well. It is also a list for the parent who sent it, because almost nobody in working life today learned how to use AI at school, and the people who get good at it now are the ones who will do well in the next decade. So if you got this from a parent, do the courses together. If you got it from a friend, send it to your parents. The same rules apply to all of you.

Who this is from

Twenty years in education, school governor, two daughters, and a book on this exact subject coming in July (SuperSkills: The Seven Human Skills for the Age of AI, Kogan Page, 3rd July 2026). Not a tech company. Nobody is paying me when you click a link. This is what I am going to tell my own kids this summer when we sit down for a week to go through it.

Be human first

In the years ahead, everyone will talk to you about AI. Friends, teachers, your school, the internet, career advisers, your parents, the headlines, the adverts. The whole noise of the next decade is going to be AI, AI, AI.

The most important thing for you to be in the next decade is human. That is why I wrote my book. That is the whole argument behind it. Everything else in this list is a tool, and every tool below is in service of a person who has chosen their path.

So before you think of AI: do the human things. Do projects. Join Model UN. Do debating. Play sport. Play music. Work with people. Have relationships. Hang out with your friends in person. Discuss films and music and videos with people whose opinions you can argue with. Go to concerts. Dance. Read books that are too long for you. Be a bit naughty sometimes. Be good when it counts. Have jokes. Try things. Don’t try other things. Have experiences. Talk to old people about their lives. Travel somewhere. Get something wrong in front of someone and live through the embarrassment.

Demonstrate, every single week, that you are human. Then think about AI.

How this list is shaped

Understand it first. Learn with it second. Build with it third. Play with it fourth. The tool you reach for first will shape what you think AI is for. Most of these tools require you to be at least 13, some 18. Check before you sign up.

Part 1: Understand AI

Before you use AI, you should understand what AI actually is. None of the resources below need a technical background, none of them need a credit card, and all of them teach the same lesson. AI is not magic, it is not intelligent in the way you are intelligent, and the people who do best with it are the ones who understand its mechanics.

Couch to 5K for AI at couchto5k.ai A free thirty-day course built by Hilary Gridley, who wanted a tool that did not exist yet and built it. About ten minutes a day. Takes you from “I keep meaning to figure out this AI thing” to actually being good at it. The best place to start, because it builds a habit rather than dumping information on you. If you just do one thing, do this.

Elements of AI at elementsofai.com A free course built by the University of Helsinki and MinnaLearn. More than two million people have taken it across 170 countries. It is the best non-technical introduction to what AI is, what it can do, and what it cannot. About six hours of material in total, broken into chapters you can do fifteen minutes at a time. Suitable from around 14 upwards.

Experience AI at experience-ai.org Built by the Raspberry Pi Foundation in partnership with Google DeepMind. Free lessons designed for 11 to 14 year olds, with classroom resources for teachers and home-learning material a parent and child can work through together. The strongest UK-anchored option for younger teens.

Google AI Quests at research.google/ai-quests A free game-based learning experience for ages 11 to 14, built with the Stanford Accelerator for Learning. You play the role of a Google researcher applying AI to real problems like flood forecasting. You collect data, build a model, and see how the decisions you make shape the AI’s output. The point of the exercise is not to use AI but to understand how it gets built in the first place.

3Blue1Brown’s neural networks series on YouTube Grant Sanderson’s animated explanations of how neural networks work, free on his channel. If you have ever wondered what is happening inside one of these models when it answers a question, this is the visual explanation that finally makes it click. Made for the curious, not the technical.

Andrej Karpathy’s general audience track at karpathy.ai Karpathy was a founding member of OpenAI and led AI at Tesla. His “Deep Dive into LLMs like ChatGPT” and “How I use LLMs” videos on YouTube are long, but they are extraordinary. They are probably the best public explanation of what these systems actually are. For older teens who want the deep version, this is where to go.

Part 2: Learn and think with AI

The tools in this section are for moments when your brain is doing the work and you want help going further. They are not answer machines. They are thinking partners, and the distinction matters because most teenagers default to ChatGPT for everything, and ChatGPT is rarely the best tool for any specific job.

My eldest daughter is at university now. For most of her time at school, she barely touched AI. Not because we banned it, but because she had worked out, on her own, that her brain was still growing and that she wanted to do the reading and the wrestling and the working-it-out herself. Now, she uses AI almost every week. She reads the papers first. She writes her thoughts first. Then she brings AI in to find sources she has missed, to argue against her own thesis, to test where her thinking is weakest. Brain first. Tool second. Every tool below is designed to support that order.

NotebookLM at notebooklm.google Free from Google. You upload your own sources, like textbook chapters, lecture notes, articles or PDFs, and NotebookLM becomes an expert on those sources only. It does not invent things. It cites where every claim came from. It can produce study guides, flashcards, and even an audio overview that sounds like two podcasters discussing your material. The free tier is generous: 100 notebooks, 50 sources per notebook, up to 500,000 words per source. For revision, this is probably the single best free tool available to a teenager today.

Perplexity at perplexity.ai A search engine that thinks. Where Google gives you a list of links and ChatGPT gives you a confident answer, Perplexity gives you an answer with citations you can actually check. Reach for it when you want to know something AND you want to be able to verify what you have been told. The free tier is enough for most school work.

Gemini Guided Learning inside the Gemini app Google’s tutoring mode inside Gemini. Rather than handing you a quick answer, it asks probing questions that push you to think a problem through yourself. Switch Gemini into Guided Learning mode or “Help me Learn” when you are revising or working through something tricky. It is one of the clearest signs that the big AI companies are starting to build tools that do not replace your brain.

Claude at claude.ai Made by Anthropic. Currently has the strongest reputation for thoughtful, longer-form reasoning, and is more willing than its competitors to admit when it is uncertain. The free tier is good for school-age use. Particularly strong for essay drafting once your own thinking is already done, debugging code, and working through complex ideas. The Projects feature lets you keep a folder of related files in context across conversations, which makes it useful for sustained work like coursework or revision. Cowork and Claude Code is the next level up and worth it if you can afford it and will use it.

Wolfram Alpha at wolframalpha.com Not new, not trendy, and absolutely essential. Wolfram is a computational engine, which means it actually does the maths rather than predicting what the answer should look like. Most chatbots hallucinate maths and physics far more often than people realise. Wolfram does not. Use it for the calculation, and a chatbot for the explanation, never the other way around. The free tier covers equations, graphs, chemistry, statistics, and more.

Goblin Tools at goblin.tools A small collection of single-purpose AI tools, originally built for neurodivergent users and useful to almost everyone. The most important one is Magic ToDo, which takes a big task you are dreading (”revise for mocks”, “write the personal statement”, “tidy my room”) and breaks it into small, doable steps you can actually start with. Free, no signup, no account. If you are the kind of teenager who freezes when something feels too big, this is probably the most useful tool on this entire list that nobody mentions.

For older teens who want to go further: Notion AI at notion.so is for building your own personal operating system, with revision databases, application trackers, and reading lists in one place. Readwise Reader at readwise.io/read turns articles, PDFs, podcasts and YouTube videos into a knowledge library you can search and revisit. Both have free tiers. Neither is necessary to start with.

One last thing in this section that may turn out to matter more than any of the tools above. Talking to AI rather than typing at it is the next big shift, and the teenagers who get used to it early will have a real advantage. I have been using Wispr Flow for a while, which lets you speak naturally and turns it into clean text everywhere on your computer. It changes the way you think when you write. You start thinking into the machine rather than typing at it. Claude, ChatGPT, and Gemini all have voice modes now too, and NotebookLM will read your sources back to you as a podcast. Voice is not a gimmick. Get comfortable with it now and the awkwardness pays back.

Part 3: Build with AI

This is where you can make useful stuff quickly. The bespoke family app. The website in an hour. The reference-material consolidator. The shoestring-budget holiday. Pick one of them and have a go. The point is that anything is possible now.

Start here

Replit at replit.com The most education-focused of the AI app builders. You type what you want, the AI builds it, and you can see and learn from the code along the way. Strong free tier. Collaborative features mean you can build alongside a friend, or have a parent looking over your shoulder. The right choice if you want to actually understand what is being built. From age 13.

Bolt at bolt.new A browser-based, full-stack app builder. You describe what you want and Bolt generates a working web app. The fastest path I know of from “I have an idea” to “the idea is online and someone can use it.”

Explore later

Lovable at lovable.dev for polished apps without learning the underlying code first.

Google AI Studio at aistudio.google.com for prototyping with the Gemini models and learning how to prompt them well.

Google Labs at labs.google for the experimental edge: Learn Your Way, Flow, Whisk, and dozens of others.

Gamma at gamma.app for generating clean presentation decks from a description or a set of notes. Useful for school projects and pitches.

v0 at v0.app for designing nice-looking React interfaces if that is your thing.

GitHub Copilot Free at github.com/features/copilot for the AI assistant that lives inside your code editor, once you are learning to code properly.

Part 4: Create and play with AI

This is where AI becomes properly fun. Use these for projects rather than for school work. The split matters, because once you have built something fun with AI, like a song or a video or an image, you understand in your bones that it is a tool you direct, not a service you consume.

Nano Banana, Google’s image model inside the Gemini, Google’s flagship image generator, accessible from inside Gemini. Particularly good at character consistency across multiple images. Use it for story illustrations, custom cards, posters, anything visual you can describe in words.

Suno at suno.com Generates songs from a description. Pop, rock, classical, hip-hop, almost anything. The free tier gives you a handful of generations a day. Brilliant for soundtracks to school projects, custom birthday songs, or just messing around. The age limit is 13.

Veo and Seedance (Google’s video model and the multi-model aggregator) - Text to video. You describe a scene and you get a short HD clip. Veo lives within Gemini with a free rate-limited tier. Seedance gives you 100 daily free credits, 1080p output, no watermark, and access to multiple models (Veo, Sora, Kling) in one interface. probably the best free video tier available right now.

Quick, Draw! at quickdraw.withgoogle.com A neural network tries to guess what you are drawing. Old, free, no signup needed, and probably the best fifteen-minute introduction to what machine learning actually is for younger teens.

MusicFX and TextFX at labs.google Google’s experimental music and text generation tools. Free, browser-based, mostly no signup needed.

Part 5: Get serious about prompting

If you want to go further, prompt engineering is the closest thing to a real skill in working with AI. The good news is that the foundations are taught for free by the people who built the models in the first place.

A small example before the tools, to give you the feel for it. Bad prompt: “Explain photosynthesis.” Better prompt: “I’m 14, I understand cells and energy, but I keep getting confused at the chlorophyll bit. Explain photosynthesis using only what I already know, then push me one step further.” The second prompt teaches the AI who you are, what you already understand, and what you actually need. Better answers follow. The principle is simple: the more specific you are about what you want, the less useful hallucinations become. Vague prompts invite confident nonsense. Specific prompts invite useful thinking.

Prompt Engineering Tutorial – Master ChatGPT and LLM Responses on YouTube. A free 41-minute video from freeCodeCamp, presented by Ania Kubów. Covers the fundamentals of prompting with practical demonstrations on ChatGPT. The first quarter is foundations on what AI actually is, so if you already know that bit, skip to around 14 minutes in for the practical material. The techniques apply across every AI tool, not just ChatGPT.

Anthropic Academy: AI Fluency for Students at anthropic.skilljar.com A free Anthropic course specifically designed to teach students how to use AI to enhance their learning rather than replace it. The closest thing I know of to a curriculum on the augmentation posture.

Google AI Essentials and Gemini for Students at grow.google/students Free training designed for students. Covers prompting, tool use, and how to bring AI into study habits without becoming dependent on it.

Prompt Cowboy at promptcowboy.ai A different kind of tool. You give it a rough idea of what you want, and it builds a properly structured prompt you can paste into ChatGPT, Claude or Gemini to get a much better result.

Part 6: The money question

Spending money on AI is very easy to start doing. Every tool on this list will, sooner or later, offer you a subscription, and a few of them are worth it for the right person. Most of them, for a teenager, are not. Two things to say.

The first is to stay on the free tiers as long as you possibly can. Almost everything in this list has a free tier that handles school-age use perfectly well. The free tiers exist because the companies want you to become familiar with their tools. Use that. Do not pay until you actually need a paid feature for a specific project, and when you do, pay for one tool, not five.

The second is that if you do subscribe, fund it yourself. A teenager asking a parent to pay £15 (or $20, or €18) a month for ChatGPT Plus, then the same again for Claude Pro, then again for Perplexity Pro, ends up with three subscriptions doing largely overlapping work. Paying out of your own pocket money or your own earnings forces a different question to surface: do I actually need this, or am I subscribing because I assume I should? The answer, almost always, is the second one.

One thing nobody seems to talk about is rotating between tools. For most school work, your queries are not context-dependent. You can use Claude on Monday, Gemini on Tuesday, ChatGPT on Wednesday and Perplexity on Thursday, and lose nothing in the process, because each conversation is fresh anyway. The free tiers across all the major models, used in rotation, give a teenager effectively unlimited daily AI access for free. The only time you really need to commit to one tool is when you are building something sustained, like a long coding project or a piece of coursework you are redrafting over weeks. The companies are competing for your loyalty. Make them earn it.

One last thing on what you are spending. Money is not the only currency these tools take from you. Most of what you type, paste, and upload becomes training data for the next version of the model. Do not feed any of these tools things you would not want a stranger to read. That includes diary-style writing about other people, anything that names a friend in a way they would not want shared, and anything containing personal information about you or your family.

Part 7: The weekly self-audit

The single most important practice in this list is not on the list. It is a question to ask yourself every Sunday night.

In the last week, did using AI make me feel more capable, or more passive?

If the honest answer is “more capable” most weeks, the relationship is working as it should. If the honest answer is “more passive” most weeks, something has slipped, and the next week needs a recalibration. More brain-first work, less AI-first work. More building, less consuming.

A teenager who asks themselves this every week is going to be in the top one per cent of users of these tools. Not because they will use AI more than anyone else, but because they will use it deliberately.

What you are probably already using

The list above is almost certainly not the first time you have ever met an AI tool. Most teenagers are using a handful of the same things, so let me be honest about each.

Grammarly. Fine for typos and grammar. It becomes a problem when it starts rewriting your sentences and your voice slowly leaves the page. Use it for proofreading. Do not let it edit.

QuillBot. A paraphrasing tool that takes someone else’s writing and rewords it to sound like yours. Almost no honest student use, and the plagiarism checkers catch it more often every year.

Photomath. Take a picture of a maths problem, get the answer with steps. The pull is obvious. The problem is too: it teaches you nothing, and the working is the learning. Try the problem first, then check with Wolfram Alpha or get Claude to walk you through the concept. Photomath is the cousin nobody wants to admit to.

Brainly. Homework forum, same problem as Photomath. Stop using it.

ChatGPT and Copilot. Both fine. The danger is not the tool, it is the default. If your first instinct on every assignment is to paste it into ChatGPT, you have stopped learning. Use them for explanation, sparring, second opinions. Not for first drafts.

What I have left off the list, and why

Some widely-used tools are not on this list at all, and that is deliberate.

Character.AI and the family of companion apps that have grown up around it are not here. The whole category exists to mimic friendship and emotional reciprocity, and that is the one corner of AI I would actively steer you away from.

This list also does not contain any anonymous “teen success story” anecdotes. The 16-year-old making 5k a month from a side hustle. The 14-year-old who skipped a grade because of ChatGPT. They make for nice copy, and they are usually invented. Even when they are not, they set up an expectation that AI will deliver outcomes a teenager has to actually earn. The point is setting yourself up. The success follows from that.

How to use this list

You do not need to do all of it. Pick one thing from Part 1 and understand it. Pick one thing from Part 2 and learn with it. Pick one project from Part 3 and build it. Run the weekly self-audit. That is your summer.

If you are a teacher reading this, the same applies in reverse. You almost certainly do not have time to teach this in lessons before the end of term. You absolutely have time to recommend a few of these things for your students to do over the summer holidays. Forward this resource, or steal from it, and make something useful for the kids you are responsible for. If you are not responsible for any of them, you know who is. Send it to them.

One more thing for the parents and the teachers. Print this out. Yes, I know that is bad for the environment. The kids you give it to need something they can scribble on. Mark it up. Cross things out. Write down which two or three of these they will actually try this summer. The act of choosing matters more than the choice itself.

The teenager who comes out of the next decade in good shape will not be the one with the most AI tools. It will be the one who used those tools to spend more time with friends, build more things, play more music, have more arguments, write more bad poems, learn more languages, see more places, and become more themselves.

This is the second piece in a two-part Box of Amazing series on AI and family. The first, “What I Tell Parents About AI”, was published last week. SuperSkills: The Seven Human Skills for the Age of AI, is published by Kogan Page on the 3rd of July 2026. Pre-order at superskillsbook.com.

Stay Curious - and don’t forget to be amazing,

Rahim

Rahim Hirji Author, SuperSkills (Kogan Page, July 2026) | Keynote Speaker | Advisor
→ superskillsbook.com → thesuperskills.com → Book a discovery call

PS. If your organisation is thinking about AI and human capability, I keynote and advise on this. I also do bulk book orders for teams.

Tools I Use:

Jamie: AI Note taker without a bot. You join the meeting. The Bot doesn’t.
Wispr Flow: Just dictate everything to your laptop and phone. Game Changer
Manus: Best AI agent to do things for you. Agentise your intelligence.
Claude Cowork: Delegate complex tasks to your friendly Claude Agent

Recommend Reading:

Now

AI strategies are missing a dimension. The missing piece is the human capability. I call it the fourth dimension. My interview with AI For Business Leaders to talk through it.
The Allianz Research Report: Headline unemployment looks fine. 4.3% in the US. 6.2% in the Eurozone. But since late 2022, countries adopting AI fastest have seen the sharpest rises in youth unemployment. AI = 40% of the variation.
I Want to Live Like Costco People On coming to Costco membership in midlife.
Brain health supplements are booming. Here’s what one longevity expert takes.
This is what the science says about which supplements work.
What Will It Take to Get A.I. Out of Schools? The tech world assumes that A.I.-aided education is necessary and inevitable. A growing number of parents, educators, and cognitive scientists say the opposite.

Next

A.I. Populism Is Here. And No One Is Ready. Silicon Valley oligarchs worried about the risks their technology posed to the world. They forgot about people.
‘HELLO BOSS’: Inside the Chinese Realtime Deepfake Software Powering Scams Around the World
Appearing Productive in the Workplace: how AI is reshaping the modern workplace by allowing employees to produce work that looks expert without actually being expert.
AI Slop is Killing Online Communities - rant on how low-effort, AI-generated content is slowly strangling organic online communities.
Grand Theft Oil Futures: Insider traders keep making a killing at our expense

Share Box of Amazing

If you enjoyed this, forward it to someone who’d find it useful. And if you haven’t yet, pre-order SuperSkills at superskillsbook.com - out 3 July and get the pre-order bonuses.

What I Tell Parents About AI

Rahim Hirji — Sun, 03 May 2026 06:01:31 GMT

My book SuperSkills: The Seven Human Skills for the Age of AI is out 3 July 2026. Pre-order at superskillsbook.com

In addition to this newsletter, I recommend a few others. All free. Check them out here.

Friends,

I have spent twenty years in education and edtech, the last several deep inside the AI conversation, and I have written a book on the human skills that our children (and us!) will need in this new world. Somewhere along the way, I have become a de facto friend of a friend who is going to solve the AI crisis for your kids.

Spoiler: I can’t.

Double spoiler: nobody can. Anyone who tells you they can is selling you something, and most of the time it is a course. You’ll see in the links below that I recommend one this week. (it’s free and not mine)

What I can do is answer the questions parents actually ask me randomly on WhatsApp, at the end of a keynote, in replies to this newsletter. Some of you are reading this in the UK. Some in the US, Canada, India, the Gulf, Australia. The questions are the same.

The questions parents are actually asking me

Does my kid need to learn ChatGPT, Claude, Gemini? The tool is the easy part. Do you remember when you stopped worrying about whether to use AltaVista, Yahoo, Bing or Google? You don’t, because by 2010 it had become a muscle nobody thought about. AI tools will go the same way. Your kid will adapt to whatever model they end up in front of. If you want my honest current view, your teenager should be comfortable with Claude, ChatGPT, Perplexity, NotebookLM and Gemini, and developing a sense for when to reach for which. That recommendation will be out of date by next year. The muscle of switching between tools is what lasts.

Do they need to learn Cursor, Claude Code, the developer tools? Only if they already love building things. These tools amplify a builder. They do not create one.

Should they still learn Python? Yes, if they love coding. Not because they will be paid to write Python in 2035. Because the logic of breaking a problem into small steps is the foundation of every kind of thinking AI cannot do for them. Everyone needs logic.

Should they still go to university? On the average data, yes. The bet is becoming binary, which I will explain below.

Coding club, AI camp, Khan Academy, online courses? Less important than you think. The critical work is happening in your house, not theirs. If they are learning logic in these clubs and meeting others, go for it!

What is the actual answer? Your child needs to develop what I call the augmented mindset. The ability to think clearly first, then use AI to extend that thinking. Not the reverse. The tool changes every twelve months. The mindset compounds for life. You get one shot at parenting. Don’t let them outsource their brain function.

The augmented mindset is built by sitting with hard problems before reaching for the tool, by using AI to argue with rather than to copy from, by learning to spot when the AI is wrong, and by knowing when not to use it at all. It is not a tech skill. It is a thinking skill that happens to involve technology.

Take the tool fluency off your worry list. Your child’s brain pattern-matches new interfaces in a way yours has to deliberately learn. They will be ahead of you on Claude or ChatGPT within an afternoon of you letting them on it. Worry about what they think with these tools, not whether they can drive them. Have you ever sat behind any Gen-Z typing and moving between windows at warp speed? It’s normal for them.

What the data actually says

In the UK, vacancies for graduate jobs, apprenticeships and junior roles have fallen 32 per cent since ChatGPT launched in November 2022. KPMG cut its UK graduate intake by 29 per cent. Deloitte by 18. EY by 11. The Big Four are restructuring the bottom of the pyramid because the work juniors used to do is the work AI now does cheapest.

The pattern is not just British. In the US, undergraduate computer science enrollment at four-year institutions fell 8.1 percent, the first decline after years of consecutive growth, according to the National Student Clearinghouse. Stanford’s Digital Economy study found employment for software developers aged 22 to 25 declined nearly 20 percent from its peak in late 2022. Recently, computer science graduates in the US have had a higher unemployment rate than recent philosophy graduates, according to the Federal Reserve Bank of New York analysis. Read that twice. I’m not making this up.

Canada, Australia, parts of Europe, India’s IT services sector are all reporting versions of the same thing. The bottom rung of the white-collar career ladder is being chewed away.

The part of this story that most articles leave out is what is actually causing it. A March 2026 Federal Reserve Board paper found 'precisely-estimated null effects' linking AI adoption to firms' job-posting behaviour. A Resume.org survey of a thousand hiring managers found that 59 percent of companies admit to emphasising AI’s role in layoffs because it plays better with stakeholders than citing financial constraints. A majority of firms are using AI as the cover story for cuts they would have made anyway. Marc Andreessen calls it “a silver-bullet excuse.” Sam Altman calls the practice “AI washing.”

Three things are happening at once.

AI is taking some tasks juniors used to do.
Post-pandemic over-hiring is being unwound.
A 2022 US tax change raised the after-tax cost of every developer hired by tens of thousands of dollars.

The first is real and probably permanent. The second is cyclical. The third has nothing to do with AI at all.

The articles your friends keep forwarding bundle all of this together as a single AI story because that version is more shareable. If anyone says it’s because of AI, ask them why and how. Don’t trust people who are too sure. Make them explain it to you. The long-term direction is real, regardless of the short-term cause. Families who plan around it seriously will do better than those who screenshot the infographic and move on.

The other side of the same data

Workers in the same role with AI skills earn 56 per cent more than workers without, according to PwC’s 2025 Global AI Jobs Barometer (a billion job ads analysed). That premium more than doubled in twelve months. Wages are growing twice as fast in AI-exposed industries as in less-exposed ones. Some of this is selection. Early adopters tend to be the kind of person who would have grown their wage anyway. The premium is partly real causation, partly that. The practical outcome is the same. AI fluency is becoming a baseline, like being able to use a computer was in 1995, like using the Internet in 2000.

Two things are happening at once, to different teenagers, in the same labour market. The bottom of the ladder is narrowing. The premium for fluency (and adaptability) is widening. Which side your kid ends up on depends on what they actually do with the tools.

The countries placing the bets

While the British and American conversations have been stuck on whether to allow AI in schools, the rest of the world has stopped debating and started betting.

UAE. AI is now a mandatory subject from kindergarten to grade twelve in every public school. 400,000 children being taught AI as a basic literacy.
Estonia. National programme giving every grade 10 and 11 student free access to ChatGPT Edu. Teachers trained by the Ministry of Education.
Singapore. Government-built, government-supervised AI tutors delivered to every child. £300 million of public money behind it.
Sweden. Bet the opposite way. After fifteen years of screens in classrooms, reading scores fell sharply and the government reversed course. Phones will be banned in schools from this autumn. Textbooks are coming back.
Finland, Denmark, the Netherlands, France, South Korea, China. All with phone restrictions in schools.
The UK. No national AI curriculum. No national phone policy. Guidance documents and working groups while individual schools improvise.
The US. No federal policy. Wildly inconsistent state by state. California has guardrails. Texas does not. Most school districts are inventing as they go.
Canada. Provincial, not federal. Ontario and BC moving fastest. Most provinces still drafting.
India. Mixed. Strong push on AI literacy in IIT-feeder schools. Rural and lower-tier urban schools have nothing.

The sovereignty question is uncomfortable. Wherever you live, if your government has not decided what AI in education looks like, the decision has been outsourced to whichever American chatbot your teenager reached for at three in the afternoon when nobody was watching.

Wherever you are reading this from, your version of the same problem:

UK parents. Your school is improvising. Your government is waiting. Your job is to be what the school is not.
US parents. Your child’s experience depends on which state you live in and which district you can afford. The school is improvising; you cannot wait for it.
Canadian parents. Provincial, not federal. Look at what Ontario or BC are doing as the better template, and ask your school whether they have caught up.
Indian parents. A split country. High-end AI literacy in IIT-feeder schools; silence elsewhere. The after-school programme matters more than the school. Python gives a kid the logical scaffolding. The augmented mindset is taught at home.

Some bets will be wrong. UAE may be doing too much, too young. Sweden may be retreating too far. We will know in a decade. But betting nothing is also a bet, and the data already suggests it is the wrong one.

The kitchen table

Nine months ago, a few days before A-level results day, my eldest and I sat at the kitchen table with two laptops open, doing the diligent thing parents do when they don’t know what else to do. We were researching alternative routes. Other universities, foundation years, the rest. It’s what I call the just-in-case folder.

She looked up from her screen. I will call her DD1. Some of you met her in my letter to her.

“Are we making a mistake going to university? AI can do all of this.”

I half-joked that she would need to read my book properly from cover to cover. In my heart of hearts, I knew that was only one step of where she would need to go. I did not say that out loud. The results came through later that day. They were great. She went and is finishing up her first year.

That conversation has happened in a thousand kitchens since. It is happening this morning in a kitchen somewhere as you read this. The parent does not have a better answer either.

The university question

A degree is, on average, still worth it. The Institute for Fiscal Studies estimates UK male graduates earn around £130,000 more over their lifetimes than non-graduates with similar profiles, female graduates around £100,000 more, both net of tax and student loan repayments. US data tells a similar story for selective institutions. Canadian data tells a similar story for the top universities.

But the average hides everything. The bet is becoming binary.

A selective university with a strong subject is still one of the safest decisions a parent can make. A non-selective university with a weak subject is, on the data, a worse bet than an apprenticeship at a serious firm. The middle ground that absorbed the expansion of the last twenty years is the bit being squeezed.

The decision is not whether to go. It is where, what, and what to do alongside it. The data cannot see what the student does at three in the afternoon when nobody is watching. That’s the bit that matters now. That’s what you’re parenting about.

The other thing the data does not see is the part of university that almost nobody discusses anymore. The lectures will eventually be delivered better by AI. The reading list could be in front of your child this evening. What university actually gives them is three or four years of forced contact with people they did not choose. Tutorials with a professor who pushes back and calls them out. A roommate they have to negotiate with. A society they have to run. A part-time job they have to show up to. Budgeting. Arguing in seminars with people who think differently. Negotiating who does what in group work. Adulting about laundry, cooking, and when to study. Finding love. Finding friendship. Deciding whether to shop the student in your team who does nothing or who doesn’t speak English. Building a relationship with an academic who eventually writes them a reference. None of that is on the syllabus. All of it is the bit AI cannot replicate. The case for university is no longer the lectures. It is the three or four years of being a human being among other human beings, structured around progressively harder problems.

That is what they are buying. Make sure the place you send them is one where it actually happens.

The addiction that is real

The most worrying thing happening to teenagers right now is not what they are learning. It is what they are losing.

A generation ago, a teenager had between five and ten close friends. The ones they grew up with. The ones they saw in person. Around them, twenty or thirty acquaintances. A normal size for a human social world. It’s what I had.

Then the phone arrived. The number went from ten to a thousand, all because of social media. A teenager today has hundreds of followers, dozens of group chats, a feed of acquaintances they have never met and never will. The phone gave them a billion ways to feel wanted by people who were not actually there. COVID accelerated what the phone had started.

The number didn’t stop at a thousand. It is now collapsing back. The next number is one. And that’s the risk.

The AI companion takes that thousand-strong feed of strangers and condenses it into a single relationship.

Ten, to a thousand, to one.

Common Sense Media surveyed 1,060 American teenagers aged 13 to 17 in mid-2025. 72 percent had used an AI companion. 33 percent preferred the AI to humans for serious conversations. 12 percent shared things with the AI they would not tell friends or family.

The teenagers reaching for AI companions are not asocial. They are the opposite. They want real relationships. They have always wanted them. The phone took most of them away. The AI now offers a substitute that is more comfortable than the work of building one back.

Addiction is not really about the technology. Addiction is about the feeling of being wanted and needed. The AI companion gives them something a thousand Instagram acquaintances never could. A presence that always responds. Always agrees. Never has a bad day. A child who learns, between the ages of 12 and 18, that the most reliable source of warmth in their life is a chat window has not been educated. They have been rewired.

The phone addiction was the starter. The AI addiction is the main course - and that worries the hell out of me.

What I actually do

I am not going to tell you what to do. I will tell you what I do. I have plenty of flaws as a parent and as a human, so take what is useful and leave the rest.

My youngest is 14. The rules in our house look more like 2005 than 2026, and I will not apologise for them.

Phones do not go in the bedroom. Not at night, not for homework, not for “five minutes.” If she needs the phone, it stays in the main room. This rule is unfashionable. It is also what most senior people I know in the AI industry actually do with their own kids.
Phones are not used at the dinner table. I’m supposed to keep to this rule. But I fail. My wife tells me off. Dinner is for eating and discussion.
The laptop is allowed for almost everything. The laptop is a tool. The phone is a relationship. The laptop sits on a desk. It makes things happen. It helps with education. The phone should also be a tool. But it’s been bastardised into something else.
No AI subscription for her yet. She does not use AI for homework. When she does start using it, I will sit down and show her what is actually possible. Not how to get an answer faster. How to use it to extend something she has already built. Build first. Augment second. I’ll do that this summer. Not in a way to provide the cheat code to education, but to broaden her horizons on what is possible.
She has a life that is not on a screen. Violin. Badminton. Alternative friend circles. Boredom is a developmental requirement. AI removes it. So I make sure something else fills the gap. TV, surprisingly, doesn’t feature that much any more.

The rule underneath all of these. I let her struggle before I help. The struggle is the lesson. Removing the struggle removes the lesson.

I do not enforce all of this perfectly. The phone gets smuggled into the bedroom on a Friday night more often than I would like. The dinner table negotiation is a weekly event, not a settled rule. I get called a hypocrite. I am writing this as someone who has lost more than one battle on it. The point is not to be the perfect parent. It is to know which battles are worth fighting and not fold on those ones.

What “use AI well” actually looks like at 14, when the time comes: if she is writing a story, the AI is a collaborator she argues with about a character she has already drawn, not a generator she asks for the next paragraph. If she is doing a history essay, the AI helps her interrogate a source she has already read, not write the analysis. If she is debugging code, the AI explains why her code did not work, not what to write next. The work is hers first.

What to ask your child’s school

I have spent a lot of time in school. Not as a teacher. As a school governor. As a speaker at hundreds of schools. As an advisor. The first thing I will say is that the school is almost never the villain. Teachers are doing extraordinary work in a sector that has been given no time, no budget, no curriculum, and no plan. The leadership is mostly waiting for someone to tell them what to do, because nobody yet has.

Your job is not to blame the school. Your job is to find out exactly what the school is doing about AI and phones, and then figure out the gap between that and what your child needs. Whatever the school covers is one input. The rest is yours to overlay at home.

So at your next parents’ evening, walk in with one opening line that signals you are a partner, not a complainer. Mine is: “How is the curriculum evolving to reward the human judgement that AI cannot replicate?”

Then ask the rest:

What is your school’s AI policy? If they don’t have one, that is the policy.
How are you teaching my child to think when the tool is doing the thinking?
How are you teaching them to spot when the AI is wrong?
Are you teaching them when not to use it?
What is your phone-in-school policy?
Have you read the MIT cognitive debt study, and what did you take from it?

If the answers are vague, you have your answer. If the school’s only response to AI is to ban it, they are protecting their grading system, not your child. If the school’s response is to permit it without limits, they are hoping the problem will solve itself.

Researchers at MIT’s Media Lab published a preprint last year. Three groups of participants wrote essays. One used ChatGPT, one used Google, one used neither. They wore EEGs throughout. The ChatGPT group’s brain connectivity was the weakest. By the third session, 83 percent of the ChatGPT users couldn’t quote a sentence from an essay they had just submitted. The researchers called it cognitive debt. It is one small study, but it points at something the rest of educational psychology has confirmed for decades. The struggle is not the obstacle to learning. It is the mechanism of learning.

What this looks like at 14 is two kids with the same maths problem in front of them. The first sits with the wrong answer for ten minutes, tries a different approach, gets it wrong again, finally sees what they missed. The second pastes the question into a chat window and reads the explanation. Both end up with the right answer. Only one of them now has a brain that has done the work. The difference is not visible on the homework. It is visible in the kind of adult each of them becomes.

What you do this week

Pick the one that scares you most.

Take the phone out of the bedroom tonight.
Spend an hour with Claude or ChatGPT on Saturday on a piece of work you actually have. Get fluent enough to advise.
Send a single email to the head of your child’s school using the question above.
Sit at the kitchen table with your kid and ask them what they actually use AI for. Listen without flinching. Don’t judge.

Pick one. Not all four. One. This week.

Any of those moves you closer to being the parent the data says your child is going to need. None of them require you to know what comes next. They only require you to act before the data forces you to.

Back to the kitchen table

Nine months on, I think DD1 was asking the wrong question.

The question is not should I go to university. The question is if I go, what else do I need to know. That is a different question. It is the question my book exists to answer.

I half-joked that she would need to read my book. I should have said this. Going to university is not the bet. The bet is what you do alongside it. Whether you walk out a person who can think, decide, argue, and adapt, or with a degree certificate and a chatbot you can’t live without.

That bet is yours. Not mine. Not the algorithm’s. Not a man with the 20-page document doing analysis on careers and bets. (Some of you will understand this.) The bet is yours.

A note for the teenager being forwarded this

If a parent has just sent you this, what they probably can’t quite say to you out loud is the following.

They are not certain about any of it. Neither am I. Neither is anyone writing the lists of what to do.

Read carefully. Argue a point. Sit with hard things before reaching for the tool. Use AI where it extends what you already bring. Watch carefully what it removes. Some friction is the learning. Build your human skills.

Your parents are scared because they love you and because nobody knows. Forgive them the panic. Take what is useful from their advice and leave what is not.

Some teenagers reading this in ten years will look back and find they spent most of their adolescence outsourcing themselves to a chat window, and lost something they cannot now name. Some will not. The outcome is not predetermined. It is the sum of a thousand small decisions, most of which you make alone.

What you become is up to you. Make sure something is.

Part two next week is for your kid. What I Tell Kids About AI. The tools they should actually use. The traps they should know about. The lines I would not cross at 14 if I were them. Send it to them.

SuperSkills: The Seven Human Skills for the Age of AI is published by Kogan Page on 3 July 2026. Pre-order now at superskillsbook.com.

Stay Curious - and don’t forget to be amazing,

Rahim

Rahim Hirji Author, SuperSkills (Kogan Page, July 2026) | Keynote Speaker | Advisor
→ superskillsbook.com → thesuperskills.com → Book a discovery call

PS. If your organisation is thinking about AI and human capability, I keynote and advise on this. I also do bulk book orders for teams.

Tools I Use:

Jamie: AI Note taker without a bot. You join the meeting. The Bot doesn’t.
Wispr Flow: Just dictate everything to your laptop and phone. Game Changer
Refind: AI-curated Brain Food delivered daily. Pick your intelligence.
Prompt Cowboy: Prompt Generator. Extract intelligence.
Manus: Best AI agent to do things for you. Agentise your intelligence.
Chat Hub: Multi-model intelligence
Claude Cowork: Delegate complex tasks to your friendly Claude Agent

Recommend Reading:

Now

Videos: 1 minute on life long learning and AI vs Moneyball
Going for Growth: Can We Make Britain Wealthy Again? (Matt Clifford) - Amol Rajan’s podcast is my new favourite. Highly recommended.
Couch to 5K AI - the 10 minute a day free course I recommended above - You know that friend who keeps saying they’ll get into AI when things calm down? Send them this! Hilary Gridley built Couch to 5K for AI. It’s free, 30 days, 10 minutes a day. No course. No curriculum. Just hand-holding from someone who actually knows what she’s doing.
The invisible force making food less nutritious - A visually striking interactive feature detailing how surging atmospheric carbon dioxide is fundamentally altering plant growth and diluting the nutritional value of our food supply. It is a profound look at a hidden consequence of climate change that directly impacts global health.
Inside the Enhanced Games - A deep dive into the controversial, Peter Thiel-backed “Enhanced Games,” where athletes are openly encouraged to compete using steroids, growth hormones, and Adderall. It is a provocative look at the intersection of sports, biohacking, and libertarian tech culture.

Next

So, about that AI bubble. - A timely economic analysis questioning whether the massive capital expenditure by tech giants (Google, Microsoft, Amazon, Meta) on AI infrastructure is sustainable, or if we are witnessing the inflation of a historic bubble.
What Is Truly Special About Human Writing As AI becomes increasingly fluent, this essay argues that the true distinction between human and machine writing is not style or syntax, but accountability. It is a compelling framework for understanding the future of authorship.
DeepSeek V4: Marginally Worse, 4x Cheaper - A technical and economic breakdown of the new DeepSeek V4 model. While slightly trailing the absolute frontier of GPT-5.4, its massive cost advantage is poised to disrupt the AI market and accelerate widespread enterprise adoption.
Opus 4.7 Knows the Real Kelsey - A thoughtful, philosophical essay exploring how advanced language models construct identity and memory, using a specific interaction with Anthropic’s Opus 4.7 as a lens to examine the uncanny nature of AI “understanding.”
Chinese Courts Rule Against AI Replacements - In a landmark labor decision, Chinese courts have ruled that companies cannot fire human workers simply to replace them with artificial intelligence. This sets a significant global precedent for worker protections in the age of automation.

Share Box of Amazing

If you enjoyed this, forward it to someone who’d find it useful. And if you haven’t yet, pre-order SuperSkills at superskillsbook.com - out 3 July and get the pre-order bonuses.

Altitude

Rahim Hirji — Sun, 26 Apr 2026 06:02:07 GMT

My book SuperSkills: The Seven Human Skills for the Age of AI is out 3 July 2026. Pre-order at superskillsbook.com

In addition to this newsletter, I recommend a few others. All free. Check them out here.

Friends,

First thing: Thanks to everyone who has bought my book so far. I really appreciate it. If you haven’t yet or want to buy more copies as gifts, there is currently a 40% sale if you buy directly from Kogan Page. Use code SALE40. If you choose to, make sure you go back to superskillsbook.com to sign up for preorder bonuses. Sale ends on 29th April (Wednesday)

Second Thing: I’m trying to get on podcasts to promote the book. If you know any or can get me on or have your own, let me know, see the list or fill in the form here.

I have exactly one signature dish, and it is çılbır. If you do not know it, look it up. In my house, we call it “Turkish”, and it’s good because of my secret chilli combo, which I’ll take to my grave. The rest of the week I do not think about dinner at all, really, until my wife says, “What shall we have for dinner?” and I say whatever is in the freezer, and she decides on the healthy version. I shop for food, but only as a slave to the list on my phone. My wife tells the app. The app tells me. I obey.

But occasionally, when the stars align and I remember that I am a human with working hands, I cook. And when I cook I like to add chilli.

My fourteen-year-old does not like chilli. My wife does not like chilli. My eldest likes chilli but she has gone to university, so I am currently the sole defender of spice in the household. Which means my cooking involves declarations. There is no chilli in this one. None, I promise. The small sliver is a private rebellion, a bit of adult heat that survives the general family veto. But no one can complain.

At some point over the last few years, we worked out what made this my chilli (or come to think of it, garlic and ginger) obsession sustainable. We keep a bag of chopped green chilli cubes in the freezer. I’m saying this as if I discovered fire, but I’m sure many of you have been doing this for generations. My mum has, my mother in law has. It’s normal for them. The concept is simple. You can make it yourself, or you can buy it ready-chopped in little frozen cubes from any reasonable supermarket. I digress. The point is that the chilli is always there. I do not have to plan for it. I do not have to think about it. A sliver comes out of the bag and goes into the pan and the stir-fry is mine.

That is the whole trick. A small piece of kitchen setup that means I never have to negotiate with myself at six o’clock on a random Tuesday evening about whether there is enough heat in this dish to make it worth cooking.

Last week, though, the bag was not there. The bag is always there, and I had stopped paying attention, because even though the fam hates chilli, with our heritage, you gotta have a little. I opened the freezer with the wok already sizzling, and the bag was gone. Not there. Not in the drawer. Not in another drawer. Not in the fridge. Nothing. Small panic.

A few options, none of them good. I could leave the chilli out and eat a stir-fry that was a little less mine. I could walk to the shop, except nobody walks to the shop mid-cook. I could ask ChatGPT what to substitute, and it would suggest a spoonful of Nando’s sauce, because this is the world we live in now. Or I could have kept the freezer topped up in the first place, which I had been meaning to, and had not.

I ended up asking my wife and she said it was on the list on the app. The other list.

The word for this situation is altitude. The height at which you think about a thing.

The chilli and the dish are the lowest altitude. The weekly shop is higher. The freezer system is higher still. And higher than any of them is the question of whether I should be cooking this dish at all, whether to order on Uber Eats, or whether I just like the idea of being the person who cooks it.

The mistake is not the dish. The mistake is being excellent at the wrong altitude while the dinner burns.

My book, out in July, calls this Big Picture Thinking and places it among seven skills. I call it altitude because that is the word that survived my own kitchen. And most of the time, we don’t know we are at the wrong altitude until afterwards. Sometimes we don’t know at all.

This is the part that gets missed. Flitting looks a lot like flow from the outside. You have eighty browser tabs open, or I do. You are on Slack and email and a Google doc someone shared three minutes ago. The deck is still in the other window, the one with the unfinished slide you have been unfinishing for a week. Your hands are moving. Your screen is full. If you squint, it all looks like you are getting things done.

But flow and flitting are not the same thing. Flow lets the thought finish. Flitting never does. Flow is when you lose track of the hour because you are actually thinking. Flitting is when you lose track of the hour because you were responding.

My response to this, after years of getting it wrong, is embarrassing in its simplicity. My diary now contains blocks of time that are not meetings. They have names like Think. Read. Nothing. I used to see this in others’ calendars and thought it arrogant. Now I think it is the only way I have found to climb.

A few weeks ago, I was on a prep call that showed me another way up.

The call was with a woman called Penny, who was in Austria. Penny was chairing a roundtable I had been invited to speak at and she wanted to know what I was going to say. She told me she was at the airport, although the background was so silent I could not tell whether she was actually at the airport or whether she was simply the kind of person who plans well enough to find an empty corner in one.

Penny has spent her whole career in HR. She is significantly older than me and significantly more experienced at most things that matter. She is not a technologist. She does not use the word “stack.” She is, by any reasonable measure, exactly the kind of person the current moment is supposed to be leaving behind.

I had prepared three clear stories about what I was seeing in the work right now. I had expected her to pick one of the three.

She didn’t pick any of them. She listened for about four minutes. Then she said, with the slight apologetic tone of someone who suspects she is about to change the subject without permission:

“Isn’t the world a weird place right now?”

I laughed. It was not what I was expecting. It sounded, at first pass, like the sort of thing you say at the start of a call to warm up before the real conversation starts. It wasn’t. It was the conversation.

There was a pause after she said it. The airport behind her stayed silent. She was waiting to see whether I would treat what she had said as a warm-up or as the real thing.

She talked about how the last few years had broken something she had relied on for forty years, which was the background assumption that the senior people in a room probably knew roughly what was going on. That confidence had gone. She was watching executives in her client companies pretend to understand AI while the people actually using it were three floors below them. She was watching graduates arrive with tools their managers did not know existed. She said the phrase “people trying to stay up to speed when the world is zooming by in a space rocket” and I wrote it down because it felt like something Penny’s mother might have said and Penny had inherited.

Penny was doing something on that call that I only recognised afterwards. She was not pretending to know. She was saying, out loud, to a stranger, that she did not have the map. And the moment she said it, the whole call changed. We stopped trading positions and started actually thinking together.

That is a skill I don’t see talked about enough. The willingness, once you are senior, to say you don’t know. Penny’s altitude shift was admitting what she didn’t know.

There is another kind of shift. Not admitting. Moving. Putting yourself in a place where you can think differently. Six months ago I started using a small piece of software called Wispr Flow. It lets me speak into the machine instead of typing. That is all it does. It should be a minor convenience. It has not been.

Think about the speeds you work at. Handwriting is slow, a few thoughts a minute, if you are honest about how much of it you cross out. Typing is faster, maybe three times faster, which is why most of us have done most of our thinking at a keyboard for twenty years. Wispr, speaking into the machine with your thumbs doing nothing, is faster again. Not a little faster. A lot faster. Fast enough that you stop thinking in sentences and start thinking in paragraphs.

But speed is not actually the interesting bit. The interesting bit is where you can now think.

I used to type at a desk. That was where thinking happened. Now I think on the walk between the school drop-off and the car. I think in the five minutes before a meeting that used to be wasted. I think while cooking. I think while looking for chilli. A thought that would previously have waited until I got back to the desk, which usually meant it died on the way, now gets caught. The machine listens and writes it down. I keep moving.

That is not a faster version of the same thing. That is a different altitude. I am not thinking more quickly. I am thinking in places where I was previously not thinking at all.

And there is a next step coming. The next version of this is not speaking, it is thinking. Brain-computer interfaces that skip the mouth entirely and read the thought as it forms. When that arrives, the altitude will shift again, and the people who were still debating whether to try voice-to-text will be two altitudes behind.

I mentioned Wispr in this newsletter a few months ago, in passing. Several people wrote back afterwards to say thank you. Not for the tool. For being pushed. One said: I hadn’t realised how much I was thinking in the shape of the keyboard.

That is the second shift. Not admission. Movement.

The hardest shift of all is neither of those. It is leaving a place where you are already winning. In 1997, Damon Albarn and Jamie Hewlett were sharing a flat on Westbourne Grove. Albarn was the lead singer of Blur, three years past Parklife, at the top of Britpop. He already had the flow. He was not stuck. He was not running out. He was inside the most fluent version of himself a songwriter can be.

The two of them watched a lot of MTV together, and Hewlett said later that watching MTV for too long was “a bit like hell, there’s nothing of substance there.” It was the boy band explosion. Five handsome young men in coordinated outerwear, doing synchronous choreography in front of warehouses. Albarn could have responded the way most people at his altitude would. Make a better record. Out-Blur Blur. Defend the pie.

He and Hewlett did something weirder. They invented a band that didn’t exist. Four cartoon characters. A blue-haired vocalist. A Satanist bassist. Record labels did not know how to photograph a band they could not see.

Gorillaz was not a side project. It was an altitude shift. Albarn was standing at the top of one ladder and, instead of climbing higher on it, he stepped off it sideways and into the air and built a different ladder. He did not solve the boy band problem by writing better songs. He zoomed out so far that the problem stopped being the problem.

Look at the photographs from those years. In the Blur pictures he is the frontman, pointing at the camera, doing the mockney lad the British press had cast him to play. Then the cartoons arrive and his face disappears from the cover entirely. The mask he picked up gave him his face back. What the altitude shift bought him was the right to pick.

That move, stepping sideways from your own success, is what I don’t see enough of in the leaders I work with. The excellence is the thing that locks them in. What you probably didn’t know is this: Gorillaz sold more than two times more records than ever did and streams three times more on Spotify. The sideways step went up and allowed Albarn to “crack” America. The ladder he built turned out to be taller than the one he left.

Three shifts. Admitting. Moving. Stepping sideways. The same move at different heights.

And none of them are a place you arrive. Altitude is a habit, and the habit fails. You forget to restock. You schedule the thinking time and then fill it with email. You give the keynote about flitting and then spend a week flitting. You write the book about big picture thinking and altitude and then catch yourself, on a Tuesday, at the wrong altitude, asking your wife about the strategy you had made unnecessary six months ago and had, apparently, made necessary again without noticing.

You don’t climb once. You keep noticing when you have drifted back down. And then you climb again.

I write about altitude and I keep running out of green chilli. Big Picture Thinking is one of seven skills, and the other six fail in the kitchen too, on a Tuesday, without announcement. The people I actually want to learn from are not the ones who live at altitude. They are the ones who notice, a second sooner than the rest of us, that they have come back down.

My book, SuperSkills: The Seven Human Skills for the Age of AI, is published by Kogan Page on 3 July 2026. You can pre-order it at superskillsbook.com.

Stay Curious - and don’t forget to be amazing,

Rahim

Rahim Hirji Author, SuperSkills (Kogan Page, July 2026) | Keynote Speaker | Advisor
→ superskillsbook.com
→ thesuperskills.com
→ Book a discovery call

PS. If your organisation is thinking about AI and human capability, I keynote and advise on this. I also do bulk book orders for teams.

Tools I Use:

Jamie: AI Note taker without a bot. You join the meeting. The Bot doesn’t.
Wispr Flow: Just dictate everything to your laptop and phone. Game Changer
Refind: AI-curated Brain Food delivered daily. Pick your intelligence.
Meco: Newsletter reader outside your inbox. Organise your intelligence.
Prompt Cowboy: Prompt Generator. Extract intelligence.
Manus: Best AI agent to do things for you. Agentise your intelligence.
Chat Hub: Multi-model intelligence
Claude Cowork: Delegate complex tasks to your friendly Claude Agent

Recommend Reading:

Now

Us grown-ups are brilliant at filling every minute. Then we are confused why clarity and curiosity feel harder to find. This from Taylor Hughes is genius.
Not because it is some sort of magic. Not because it tells you to slow down.
But, because it shows you what you lose when you don’t.
Stuart Russell, who literally wrote the textbook on modern AI, uses King Midas to explain what we’re getting wrong. A Greek myth about a king who didn’t know what he actually wanted.
Thousands of CEOs admit AI had no impact on employment or productivity -
Amidst the breathless hype of the AI revolution, a dose of cold, hard reality. A massive survey of CEOs reveals a startling paradox: despite the billions invested and the endless pilot programs, the actual, measurable impact on productivity and headcount remains stubbornly flat for many. It’s a crucial counter-narrative that suggests we might be in the “installation phase” of this technology, where the tools are deployed but the organizational rewiring required to actually benefit from them hasn’t happened yet.
Why UBI is making a comeback - Universal Basic Income is back in the discourse, but the driver this time isn’t utopian idealism—it’s Silicon Valley pragmatism. As tech companies face mounting backlash over AI-driven displacement, a check in the mail is increasingly viewed as the necessary shock absorber for the coming economic transition. Casey Newton unpacks why the tech elite are suddenly so keen on reviving the UBI conversation, and why we should perhaps be skeptical of their motives.
The Ozempicization of the Economy - Kyla Scanlon has a knack for coining the perfect phrase for our current malaise. “Ozempicization” isn’t just about weight loss drugs; it’s a metaphor for an economy—and a culture—increasingly reliant on quick-fix interventions that suppress symptoms without addressing underlying structural rot. From biohacking to the financialization of everything via sports betting, she weaves together a compelling thesis about a society looking for shortcuts in an age of profound uncertainty.

Next

Google: 75% of new code now AI-generated - We’ve been talking about AI-assisted coding for years, but this metric from Google feels like crossing the Rubicon. When three-quarters of the new code at one of the world’s premier engineering organizations is generated by machines and merely reviewed by humans, the fundamental nature of software engineering has changed. We are moving from an era of writing code to an era of editing and orchestrating it.
Engels’ Pause and the Permanent Underclass - History doesn’t repeat, but it rhymes. This brilliant essay draws a parallel between our current AI boom and “Engels’ Pause”—the multi-decade period during the Industrial Revolution where technological progress exploded, but working-class wages stagnated. It’s a sobering historical framework for understanding the potential societal fallout of AI, warning of the creation of a permanent digital underclass if the gains of automation aren’t structurally distributed.
Chinese tech workers training their AI doubles—and pushing back - A viral GitHub project in China that allows workers to clone themselves into reusable AI agents has sparked a fascinating labor movement. What starts as a productivity hack quickly morphs into an existential threat, forcing Chinese tech workers to confront the reality of training their own replacements. It’s a crucial dispatch from the front lines of the AI labor transition, offering a glimpse into the pushback and negotiations that will soon define workplaces globally.
Meta to start capturing employee mouse movements, keystrokes for AI training
The data hunger of modern AI models has reached a new frontier: the granular, micro-movements of human workers. Meta’s move to capture employee keystrokes and mouse movements isn’t just about surveillance; it’s about training agents to mimic the exact cadence and workflow of human productivity. It raises profound questions about the future of work when your daily digital exhaust becomes the training data for your eventual replacement.
AI ping pong robot beats elite human players We’ve grown accustomed to AI mastering digital domains—chess, Go, coding. But the physical world is messy, unpredictable, and governed by physics, not just logic. That’s why a robotic system capable of beating elite human table tennis players is such a milestone. It requires real-time perception, split-second physical actuation, and strategic planning, all happening simultaneously. The gap between digital intelligence and physical robotics is closing faster than we thought.

Share Box of Amazing

If you enjoyed this, forward it to someone who’d find it useful. And if you haven’t yet, pre-order SuperSkills at superskillsbook.com - out 3 July and get the pre-order bonuses.

One Day

Rahim Hirji — Sun, 19 Apr 2026 06:01:00 GMT

My book SuperSkills: The Seven Human Skills for the Age of AI is out 3 July 2026. Pre-order at superskillsbook.com

In addition to this newsletter, I recommend a few others. All free. Check them out here.

Friends,

There is a particular kind of phrase that only gets used over dinner, after the second drink, when the table has relaxed enough that people say the thing they actually mean.

One day I’ll write a book.

One day I’ll go to Japan.

One day I’ll leave the job and do the thing I started out wanting to do.

One day I’ll post the recipes properly. People keep asking.

One day I’ll reinvent myself.

You know how it goes. The friend who has been saying it for as long as you have known them. The colleague who raises it every time the conversation turns to work. The cousin whose business idea has been the same business idea for eight Christmases in a row. It is the warmest phrase in the English language and it is also the saddest, because most of the time there is no day. There is the phrase, and there is the dinner, and there is the laughter, and then there is next Thursday, which is exactly the same as this Thursday. One day is a room in the house where nobody ever lives. People furnish it, walk past it, point to it when guests come round. Nobody sleeps in it. Nobody uses it.

I said it for ten years.

If there is one thing I want from this newsletter, and from the five hundred or so Sundays I have now spent writing to you, it is for you to be better at the end of reading me than you were at the start. Not wiser in the abstract. Better. Closer to the version of you that actually does the thing. This letter, more than any other I have written, is about that.

Ten years

I decided I was going to write a book in Shanghai, in 2014. I had been working in and around the field for years already, but 2014 was the first time I watched a machine learning system do something I had only ever seen humans do. I did not know what the book would be about. I only knew that I wanted to write one, and that what I was watching meant the world was going to change in ways that most of the people I knew had not started thinking about yet. The more I went back to China over the following years, the more this conviction hardened. Something was coming. Somebody was going to have to write about what it meant for the humans in the middle of it.

I started a newsletter in 2017. I did not start it because I was working up to a book. I started it because I could feel something changing and I needed a place to think out loud about it every week, and I have written it every Sunday since, which is now about five hundred Sundays in a row. For five years the newsletter was just the newsletter. Then ChatGPT happened, in late 2022, and the thing I had been watching in research labs showed up on everybody’s phone at once. That was the trigger. The book had been abstract for eight years. It became urgent in a week. And it turned out, when I looked back, that the newsletter had been the on-ramp to the book the whole time without knowing it. If you have been reading me for a while, you have been reading the book in instalments and neither of us realised.

Ten years is long enough that the friends who heard you say it the first time stop asking. Long enough that you stop bringing it up yourself. You are forty, forty-one, forty-two. Something happens that is not in your plan, and then something happens again. You come home the second time and something is different, although you cannot yet say what.

For a long while, I told myself that writing the newsletter was enough. Maybe writing every Sunday was impact enough. Maybe a book was old-fashioned in an age where people have stopped reading them. But in an oddly old-fashioned way, a book is a legacy at a time when clarity is needed. A book is a moment. It is a calling card. It is a reference point. It is a body of work. It is proof.

Some of the people I told, back at those early dinners, laughed at me. Not unkindly. A few of them thought I was joking. A few of them thought I was serious and were embarrassed for me, which is worse. A few of them said the thing people say when they do not believe you but do not want to be rude about it, which is worse than being laughed at directly. A legitimately published book, on a subject that matters, from somebody whose day job is something else entirely. Who does that. I could feel the room adjust.

Maybe some of the laughter was mine. Maybe the voice in my head had been doing it for longer than any of them had. You can spend a long time at a table of people who love you and still be the person at the table saying the least about their own work, because the loudest voice in the room is the one inside your head asking who you think you are. Mine did.

You may have been laughed at too. You may even have laughed at yourself, because saying the dream out loud from inside your own head can feel ridiculous. Maybe you are reading this right now and thinking: how can this guy have written an actual book? Fair question. I have asked it of myself more times than I am going to count.

The gap is not where you think it is. You think the gap is about you: your discipline, your talent, your courage. That is not where the gap lives. The gap lives in a set of assumptions about what it costs to start, and those assumptions are out of date.

What actually changed

The story people assume, when they hear that I wrote a book, is a story about willpower. That I pulled myself together. That I developed a routine. That I found the discipline I had been missing. This would be a neat story, and it is not what happened.

What happened is that, while I was spending ten years not writing a book, the cost of doing almost everything collapsed. I did not notice the collapse until I walked into it. The world I had been putting off the book in was not the world I was now living in, and the reasons I had been giving myself for not starting had, at some point while I was not looking, stopped being true.

I built a website for the book. It is at superskillsbook.com, and you can go and look at it right now if you want to test the argument I am about to make. I needed a website: a proper one, clean, readable, with the right tone, something I could send to a journalist without flinching. So I did what writers do. I asked for a quote. The quote came back at between fifty and sixty thousand dollars, from a web developer who had built sites for some of the best non-fiction authors in the business. I had asked because I wanted to understand what that kind of work actually costs at the top of the market. Now I know. I read the email and closed it. And deleted it. And decided no, I wont do that.

Five years ago, that is where the story ends. You either pay the sixty thousand or you do not have a website, and if you do not have a website you explain to people that it is coming soon and it never comes. The quote is the wall, and the wall stops you, and the wall is not about the website at all. It is about whether the book is real.

This time I did not stop. I built the site in Claude Code over a weekend, talking into my laptop rather than typing. I am not a technical person. I did a computer science degree in the last century, which is a sentence I enjoy writing, but the gap between what I learned then and what is possible now is so wide that the old knowledge barely helps. What it does give me is the absence of fear. I am not afraid of the keyboard. Everything else, I made up as I went along.

The site went live in a weekend. The direct total cost was under ten dollars. Most of the time was not spent building. It was spent thinking about what the site was for and what somebody arriving from a cold link needed to see in the first three seconds. That is the part that is always going to be human, and it turns out to be the part that takes the time.

Sixty thousand dollars is what the best web developers in the business quote for a serious non-fiction author. Ten dollars is what I paid. The gap between those two numbers is the gap the whole world has fallen through while most of us were still saying one day. That is not a story about me. That is a story about the ground shifting under everyone who has been waiting.

The hierarchy

The website was not the hardest thing. The hardest thing was the book itself. People have asked me, more than once, whether I got ChatGPT to write it. They ask it like a clever question. It was boring the first time, and it has worn thinner with every repetition. The short answer is no. The longer answer is the whole point of this section, and it is more interesting than the question deserves.

At the bottom of the hierarchy: I used Grammarly. That is the only AI that touched the actual prose. I used it to catch the spelling mistakes I have been making for forty years. I used it to tidy up my commas. Writing that sentence out loud is embarrassing for a man publishing a book about human skills in the age of AI, and it would be less embarrassing to leave it out. The embarrassment is the point. AI did the bottom layer of the writing, the mechanical layer, the spell-check layer. It did not do the thing in the middle.

In the middle of the hierarchy: research, analysis, and everything that counts as the thinking around the thinking. I surfaced academic papers I would never have found on my own. I surfaced books I had not heard of. I chased citations down rabbit holes I would not have had the patience for unaided. I read what people in other languages were writing about the same questions. I anonymised client data and found patterns in it I had felt but not yet named. I analysed the market. I looked at how the best non-fiction books of the last decade were structured, and built a blueprint from what I learned, because I wanted to write something that sat alongside them without copying any of them. I built workflows for the repetitive parts of the process and ran them until they ran themselves. AI was a thought partner for all of that, in the way a good junior researcher would have been in 2015. Cheaper, faster, tireless, and occasionally wrong in ways I had to catch. Sometimes there were hallucinations. Sometimes I hallucinated. I think.

At the top of the hierarchy: the website, the book trailer, the cover concepts, the social graphics, all of it. These are the things that would once have required a production company. An agency. A small team of people on a retainer. They now require a laptop and an afternoon.

I will give you a specific example, because it is the kind of example that makes people ask me the question again about whether I really did this myself. I did a full analysis of every non-fiction book on the market about AI, human capability, work, and the future of skills. I looked at the covers. I looked at the titles. I looked at what was working and what was not, and I reached a conclusion that surprised me, which was that the cover of my book should be black. That is a weird conclusion for a book whose whole argument is about hope. It is also correct. I tested the concept against synthetic audiences, sent the brief to my publisher, and to my surprise they agreed. Black was the right colour. They came up with the final design from there. I am telling you this because the part of the process people imagine AI did, which is the writing, is the part it did least. The part AI did most is the part people never think about. The analysis. The tactics. The patterns. The decision about what colour the book should be.

And then, sitting above all of this, outside the hierarchy entirely, is the thing AI did not do.

It did not write the book. It did not hold seventy thousand words in one head for a year. It did not stay up at two in the morning trying to remember whether the chapter on adaptability already referenced the fishermen in Kerala or whether that was a note I had made for a different chapter and then forgotten about. It did not feel the strange, private pain of realising, when the typeset proof came back from my publisher, that I had forgotten some of what I had written. There was an example in an early draft about Airbnb and how they pivoted during the pandemic. I loved that example. I thought it was still in there. When the proof arrived I could not find it. It turned out I had killed the darling myself, months earlier, and forgotten I had done it. You lose things in a book the way you lose things in a house you have lived in for too long. Not because you do not care. Because there is too much of it for any one mind to hold.

It also did not rewrite chapters. I did that. I rewrote chapters. I rewrote the whole book, more than once. My first draft was not good enough. Neither was the second. Neither, if I am being honest about the way a perfectionist reads their own typeset proof, is the final version. It is the best I could do with the time and the mind I had, and the thing I had to learn to say out loud was that the best you could do with the time and the mind you had is what shipping is. Nothing is ever ready. The date on the calendar is what makes you stop.

The book took a year of my life. It took more than that, actually, because the year of writing was sitting on top of a decade of thinking, and a lifetime of knowing what I wanted to say without being able to say it. AI helped with the work around the work. The work itself was mine.

Four words

I have written about being ill before. If I keep doing it I will have to rename the newsletter. But it matters here, briefly, because two stays in a hospital bed did something to how I thought about time, and without that I do not think the book would exist.

I came home the second time and I was not well, and I was not sure I was going to finish the thing I had been saying I would do for ten years. It was a Tuesday afternoon. I was on the sofa pretending to read. I had not written anything in weeks. She came in and sat down, not across from me but next to me, and she said four words. I am not going to tell you what they were. The specific words are not the point. The point is that they were said by a human being, in a room, for free, and nothing I have built since would exist without them.

The cost of doing almost everything has collapsed. The cost of being carried through the part where you almost do not finish has not. It was never going to. The ground has shifted on the outside. It has not shifted on the inside. The question for the next decade is how we use the first fact to protect the second.

The augmented mindset is not the tools. It is what the tools free you up to notice.

There is a name for this. It is one of the seven skills the book is built around, and I call it a skill deliberately, because it behaves like one. You get better at it with practice. You get worse at it with neglect. Nobody is born with it. You cannot buy it by buying a better model. The mindset is the muscle. The tools are the weights. The muscle is what you are actually building.

The shoot

The trailer.

I made a book trailer, which is a ridiculous thing to do. Books do not have trailers. Films have trailers. Books have blurbs and jacket copy and, if you are lucky, a good cover. The category of book-trailer-as-documentary barely exists, because a book is to be read and a video is to be watched, and the two things live in separate worlds. I made one anyway, because the argument of my book is easier to feel than to explain, and I wanted people to feel it before they decided whether to read it.

The thinking happened on a morning walk. I walk in the mornings to clear my head before the day starts, and on the walks I talk to Claude on my phone about whatever I am trying to work out. This is not extra time. It is the same walk I was already doing. The only difference is that the walk now produces something. By the time I was home I had the structure.

I recorded it in one afternoon. My wife, who hates being on video almost as much as I do, agreed to be the interviewer. She sat off-camera and asked the questions we had written. Her voice does not carry the whole piece, but it carries the beginning and the end, and the beginning and the end are the arc. She is the reason the trailer has a shape.

My fourteen-year-old daughter was the director. She took the videos. She held the phone. She watched the screen and told me things like no, that was not good, you fumbled that one, you need to do it again. The credits of the trailer list her as director, and she earned it. When she said action at the start of each take I kept bursting into laughter, because she is fourteen and she was serious and she was in charge, and the gap between her seriousness and the whole ridiculous situation of her middle-aged father trying to be on camera for his book was funnier than anything I could have planned. At the end of each take she said cut. She said it the way a director says it, not the way a fourteen-year-old says it, and I am not going to forget the sound of her saying it for as long as I live.

If you watch the trailer you will notice that the phone is slightly wonky. I set it up myself and I got it slightly wrong, and she can tell, and I am not going to call her out on the angle because the mistake is mine. The whole thing was edited in CapCut over half a day. The total cost was zero. I have never wanted to be a TikTok influencer, or an Instagram influencer, or any kind of person whose job is being in front of a camera. The experience of making the trailer was, privately, horrific for me. I did it anyway, because my idols, the David Epsteins and the Simon Sineks, do this kind of thing all the time, and at some point you have to stop explaining to yourself why you are different, and step out, and do the uncomfortable thing.

The trailer is not perfect. It is good enough. Good enough, in this age, is a miracle that used to cost between twenty and fifty thousand pounds, and now costs the afternoon you were going to spend on something else.

The trailer is at the bottom of this newsletter. Watch it or do not. What matters is not whether you watch the trailer. What matters is what is in your house, right now, that you could make with your daughter this weekend if you decided to.

Your one day

What is your one day?

I am not asking rhetorically. I am asking as someone who spent ten years inside the same sentence you are inside. What is the thing you have been telling people at dinners for as long as any of us can remember? What is the one-day that you no longer bring up yourself because you have heard how it sounds? The book. The business. The move. The trip. The recipes. The career change. The thing you were going to do when you retired, except retirement has moved and the thing is still waiting.

In the old world, execution was a filter. A hundred people wanted to do what you wanted to do, and most of them never finished, because the building was too expensive and the coordination too hard and the money ran out. The filter did the work of thinning the field for you. You did not have to be the best of the hundred. You only had to be one of the ten who made it through.

The filter is gone. Not weakening. Gone. You are not competing with the hundred people who want the same thing as you. You are competing with the ten of them who will notice, first, that the filter has gone.

The filter used to protect you from having to be brave. It let you tell yourself the reason you had not done the thing was that the thing was too expensive, or too hard to build, or too dependent on somebody else giving you permission. Those were true reasons once. They are the reasons you are using to hide from the fact that what was stopping you was never the thing you said it was. Being brave is the only differentiator you have left. The people who figure that out first are going to be unstoppable, and you already know whether you are one of them.

Maybe if I had been early, it would not have worked then. I probably needed the decade. The tools did not exist in 2015 and neither did the version of me who could have used them. That is the one excuse I am allowed, and I am holding on to it carefully, because it is the only one I have. You do not have it. The tools exist for you right now. You have no excuse left.

You either do it or you do not do it. That is the whole thing. I said it for ten years, and then I did it, and the version of me that did it is not braver or more disciplined than the version that did not. The version that did it is the version that noticed the filter had gone.

The launch date

My book launches on the third of July. That is seventy days from now, give or take, depending on when you are reading this. I have said the date so many times that I no longer believe it. I know the day itself is going to be an overrated moment. Launch days usually are. I will believe it when I am holding a physical copy in my hand, or when I walk into a bookshop and see the book sitting on a shelf next to books I spent my twenties reading, and I can take a picture to prove to myself that the shelf is real and the book on it is mine.

A lot of you have already pre-ordered the book. Some of you have ordered three copies. A few of you have ordered ten, which I understand comes mostly from companies buying it for their teams, though I know one or two of you are doing it as a favour to me. I have been carrying your names around in my head for months. I do not know most of you in person, and I have still spent actual mornings thinking about what you will think when the book arrives, whether the chapter you will like best is the one I hope it is, whether you will recognise yourself in the bits I wrote late at night. When somebody I have never met pre-orders a book I spent a year of my life on, it is not a transaction. It is a bet. You bet on me before there was anything to bet on. I know, and I will not forget.

If you are reading this and you have been waiting to see whether the book is a real thing before you decide, it is. The people who bet early made it real. The date is the third of July, and I would like to meet you at it.

I did it with AI. I launched it with AI. I wrote it alone. I wrote it because I have spent twenty years watching people try to work out how to be human in a world that keeps changing the rules, and I had something to say about it that I could not say in a keynote, and one day I decided that one day had run out.

You have the same tools I have. You have a phone. You have a laptop. You have a morning walk or a commute or twenty minutes between meetings where you could be talking to a model about the thing you have been carrying for a decade. You have a person who, if you asked them, would say four words to you that would carry you further than any tool ever could. That person exists. You already know who they are.

Go and find them. Tell them what you are going to do.

Then do the thing. Not one day. This one.

Cut.

SuperSkills: The Seven Human Skills for the Age of AI is published by Kogan Page on 3 July 2026. You can pre-order at superskillsbook.com.

Stay Curious - and don’t forget to be amazing,

Rahim

Rahim Hirji Author, SuperSkills (Kogan Page, July 2026) | Keynote Speaker | Advisor
→ superskillsbook.com
→ thesuperskills.com
→ Book a discovery call

PS. If your organisation is thinking about AI and human capability, I keynote and advise on this. I also do bulk book orders for teams.

Tools I Use:

Jamie: AI Note taker without a bot. You join the meeting. The Bot doesn’t.
Wispr Flow: Just dictate everything to your laptop and phone. Game Changer
Refind: AI-curated Brain Food delivered daily. Pick your intelligence.
Meco: Newsletter reader outside your inbox. Organise your intelligence.
Prompt Cowboy: Prompt Generator. Extract intelligence.
Manus: Best AI agent to do things for you. Agentise your intelligence.
Chat Hub: Multi-model intelligence
Claude Cowork: Delegate complex tasks to your friendly Claude Agent

Recommend Reading:

Now

The Hardest Part Of History To Tell Is How It Felt “Historians and nonfiction authors often glide over lived experience. They prefer actions, citations, details, dates.” A brilliant essay on the challenge of conveying emotional truth and lived experience in historical narratives.
Curiosity Is No Solo Act - An argument that curiosity thrives not in isolation, but within interconnected relationships. “It gains its real power when embedded in webs of relationship and shared meaning-making.”
You are what you consume - An interesting economic and cultural argument: personal identity should come from the things we get to choose for ourselves, rather than the circumstances we are born into.
Finance leaders warn over Mythos as banks prepare to use powerful Anthropic AI tool - Release of new Claude model, so far limited to US firms, will expand to British institutions in coming days
An end to cervical cancer is possible - And some countries are already on track to do so.

Next

How Anthropic Learned Mythos Was Too Dangerous for the Wild
The AI company’s own experts warned Mythos could hack the systems beneath most modern computing. Banks and government agencies are racing to gauge the threat.
The Great American GLP-1 Experiment A look at the rapidly expanding, off-label use of GLP-1 drugs. Beyond weight loss, people are now experimenting with these medications to treat everything from concussions and long Covid to drug addiction and arthritis.
On Dwarkesh Patel’s Podcast With Nvidia CEO Jensen Huang - Some podcasts are self-recommending on the ‘yep, I’m going to be breaking this one down’ level.
The AI Revolution in Math Has Arrived - AI is being used to prove new results at a rapid pace. Mathematicians think this is just the beginning.
The Economics of Software Teams: Why Most Organizations Are Flying Blind

Share Box of Amazing

If you enjoyed this, forward it to someone who’d find it useful. And if you haven’t yet, pre-order SuperSkills at superskillsbook.com - out 3 July and get the pre-order bonuses.

Being Human Is Not the Floor. It's the Ceiling.

Rahim Hirji — Sun, 12 Apr 2026 06:01:20 GMT

My book SuperSkills: The Seven Human Skills for the Age of AI is out 3 July 2026. Pre-order at superskillsbook.com

In addition to this newsletter, I recommend a few others. All free. Check them out here.

Friends,

At some point, you will lose a loved one.

At some point, you will have to console someone else who has lost theirs. At some point, someone will console you. At some point, you will have to say “my condolences to you and your family” and wonder whether those words have ever helped anyone in the history of the human race. At some point, you will just give someone a hug because there is nothing else to give. At some point, you will say “I’m there if you need to talk” and hope they believe you. At some point, we all die, and someone will be consoling your loved ones.

This is not an essay about mourning.

This is an essay about what happens next. The shape you take. The version of yourself you become. The thing that nobody teaches you and everybody needs.

The Love Touch

In the Kailahun district of Sierra Leone, in the summer of 2014, the women washed the bodies of the dead.

This is how it had always been done. When someone died, the family gathered. They washed the body with bare hands. They dressed it in clean cloth. And before the burial, they performed what they called “a love touch”: a final press of skin against skin, a kiss on the face of the person who had left. The gesture that said: you are not forgotten. You are being carried from this world into the next by the hands of people who loved you.

Then Ebola came to Kailahun.

The virus was at its most lethal in the body immediately after death. Every fluid, every pore, every surface of the skin was teeming with it. The love touch became the act that sent the living to join the dead. One funeral in the neighbouring district of Moyamba infected twenty-eight mourners. Eight of them died. All eight had touched the body.

The authorities arrived with a solution. Trained teams in protective suits would collect the bodies. No family contact. No washing. No love touch. The body would go into a sealed bag and into the ground.

The community refused. Not because they didn’t understand the science. Because they understood something the scientists didn’t. In their belief, a spirit that is improperly buried does not rest. It returns. It harms the family it was meant to protect.

So they adapted.

They put on gloves. Three pairs, one over the other. They replaced water with chlorine solution. They covered their faces, their arms, their bodies in protective equipment that turned the intimate act of farewell into something that looked like a hazardous waste operation. And then, inside all of that plastic and latex, they washed the bodies anyway. Imams and pastors were invited to pray at a safe distance. The protocol was eventually renamed from “safe burial” to “safe and dignified burial,” because an anthropologist finally convinced the authorities that you cannot strip a people of their mourning and expect them to cooperate with anything else you ask of them.

The women of Kailahun did not abandon their ritual. They transformed it. They found a way to carry the dead into the next world without following them there.

Why Chameleons Get a Bad Rap

We use the word chameleon as an insult. It means fake. It means someone who changes to fit in, who mirrors whatever room they walk into, who has no centre.

I think chameleons get a bad rap.

Because the women of Kailahun were chameleons. They changed everything about how they mourned. The tools, the method, the physical mechanics of farewell. Everything. And the one thing they didn’t change was the thing that mattered: they showed up. They refused to let death become administrative.

This is what I call Global Adaptability. Not the version you find in business books, with its checklists of cultural dos and don’ts. The version that asks: can you walk into a situation you have never been in before, where the rules are different, where the stakes are high, where you might get it wrong, and still be fully human?

The chameleon doesn’t change to hide. It changes to arrive. It changes to become the comfort.

But not everyone changes. Some people expect the world to come to them.

Flying Pancake

I was once in a business meeting with a potential partner. He was more senior than everyone else in the room. He was the decision maker. And because I have Indian heritage, he decided to call me “Flying Pancake.” He had once seen someone make a chapati in a restaurant, tossing it in the air, and he thought it was amusing. So that became my name. For a period of time, in that relationship, I was Flying Pancake.

He laughed. His colleagues laughed. His team laughed. And my own team laughed too, because we were trying to make the deal, and when the person with the power laughs, the room laughs with them. That is how it works.

If you have seen The Devil’s Advocate, the film with Al Pacino and Keanu Reeves, there is a scene where the wives laugh and cackle at Charlize Theron, and their faces twist into something ghoulish, something devilish, and she is terrified because the room has turned and she is alone in it. That is how it felt. It was their world. I was not part of it.

Everyone in that room knew it was inappropriate. I could see it in the faces of some of his own people, the slight flinch, the eyes that went to the table for a moment before the smile returned. They knew. And not one of them said a word. Nobody apologised. Not during the meeting. Not after it. Not privately. Not ever.

This is what happens when you don’t understand the nuance of adaptability. You expect others to come to your territory. You don’t even want to go halfway. You reduce someone’s heritage to a joke, and then you wait for the room to confirm that the joke was acceptable, and the room does, because power has its own gravity.

The women of Kailahun put on three pairs of gloves to honour a stranger’s body. A man in a boardroom couldn’t be bothered to learn my name.

The One Who Is Sleeping

Eleven hundred miles east of Kailahun, on the Indonesian island of Sulawesi, the Toraja people have arrived at a different answer to the same question.

When a Torajan dies, they do not die. Not immediately. Not in the way that you or I would understand death. The body is embalmed and kept in the family home, sometimes for months, sometimes for years, while the family raises the funds for a funeral that can cost more than a lifetime’s earnings. During this time, the deceased is referred to as “a person who is sick” or “the one who is sleeping.” They are served food and drink at mealtimes. They are spoken to. They remain, in every meaningful sense, a member of the household.

And every few years after burial, in a ceremony called Ma’nene, the dead are exhumed. Their grandchildren dress them in fresh clothes. They are walked through the village. The relationship between the living and the dead does not end. It is maintained. It requires labour. It requires showing up, year after year, generation after generation.

Two communities separated by thousands of miles, different religions, different languages, different histories. Both have decided, independently, that the relationship with the dead is worth the effort. Worth the risk. Worth the cost. The method is different. The interpretation is different. The meaning is the same.

A Naked Death

Then came COVID, and the rest of the world learned what the people of Kailahun already knew.

Families could not visit dying relatives in hospital. They could not hold hands in the final hours. They could not wash bodies according to Islamic custom, or light funeral pyres on the Ganges, or sit shiva in a room full of people bringing food and comfort. In Egypt, families were denied the right to bathe their dead. In Ireland, the wake was cancelled. In Jewish communities, mourners gathered on Zoom and tried to replicate a ritual that requires physical presence, the warmth of a room, the weight of another body next to yours on a low chair.

One bereaved person described a pandemic death as “a naked death to our culture.”

We knew people who passed away. My uncle passed away. We knew friends whose parents passed away, whose loved ones passed away. And there was nothing we could do. We couldn’t visit. We couldn’t hold. We couldn’t even stand in the same room and say the words that we already suspected were inadequate.

We joined too many Zoom funerals. It became normal. You could sign in, sign out, and for a moment feel that you had done something. But you hadn’t. Not really. The screen made it easy to not be involved. And for a short time, I think we all thought it might be the end of the world.

And when the restrictions lifted, we didn’t entirely go back. We had learned, during those months, that grief could be handled at a distance. That a message could replace a visit. That a reaction emoji could stand in for an embrace.

We had become impotent in consoling. Not because we stopped caring. Because we forgot how.

The Busy Work

My father-in-law passed away about two years ago. He was over ninety, and some say that is a good innings. But it is always a shock. It is always a shock because you prepare for the fact and not for the feeling, and the feeling arrives on its own schedule.

I did the busy work. I helped arrange the ceremonies. I made the calls that needed making. I did whatever my wife needed me to do, which sometimes was a specific task and sometimes was nothing at all except to be in the room. And I was a father to two daughters who had just lost their first grandparent. The only thing you can do is be there. Sometimes it was standing in a doorway, unsure whether to enter or leave, and choosing to stay because staying felt harder than leaving and therefore probably the right thing.

I did not know it at the time, but I was about to go into hospital myself. My body, it turned out, was just getting through this period of support before its own collapse. A chameleon in disguise in a different way. So I was shape-shifting between two things at once: holding someone else’s grief while my own body was giving out.

What stays with me from that time is the shame. Not shame at anything I did or didn’t do. Shame at not being good enough. At not having the right words. At not being able to fix the thing that was broken, which of course could not be fixed, because the death was complete.

Sometimes good enough just means showing up. Sometimes good enough just means sending a text. Sometimes good enough means standing in a kitchen at midnight doing the washing up because the washing up is there and you are there and your hands need something to do that isn’t helpless.

Sometimes There Is Nowhere to Mourn

Loss has more shapes than we talk about. There are those who lose a child, and there is nothing in any language on earth that is adequate to that. There are those who lose a parent before they have learned what a parent is. And there are those who mourn something they cannot name because it was never present enough to be named. Sometimes you don’t know your own norms until they are taken from you.

Mee-Yan Cheung-Judge was my mentor. She was one of the most important people in my professional life, and I am not the only person who would say that.

She had a way of feeding you and counselling you at the same time. Deep conversations over noodles. Clarity and candour served together. She would sometimes snort with laughter, and you would think that only you had the gift of making her do that. But she did it with everyone. She made every person feel like the only person in the room. And then she would walk out and speak to the CEOs of the largest companies in the world. Like a chameleon.

Over brunch in Oxford in 2015, she told me something I have carried ever since: “Your calling is not technology. It is to help humans do their best.” She was right. She was almost always right.

She died from complications of COVID, after four weeks in intensive care, after months of the kind of battle that strips everything back to what matters and leaves the rest behind.

At her celebration of life, a handful of us read tributes. It was a mixture of laughter and tears, which is the only honest combination for someone who gave that much to that many. I read my words and I meant every one of them and they were not enough. They could never be enough.

And afterwards, the mourning happened alone. Because sometimes the ceremony gives you a container, a room, a structure, a set of words, and sometimes it doesn’t, and the rest you carry by yourself.

Sometimes there is nowhere to mourn. Not because there is no place. Because the grief does not fit into any of the places that are available. It is too large for a text message and too private for a public ceremony and too formless for a conversation. And so you carry it in the background while the foreground continues.

The Shape You Take

I should tell you something about myself.

I am, by every measure that these things offer, a highly skewed introvert. An INTJ, if you know the letters. People don’t believe me when I tell them this. That comfort they see in me at a keynote, across a table, in a room full of strangers, is not natural. It is learned. It has taken years. And it is still, every single time, a conscious act of stepping out of my own nature to meet the moment.

Whether you are sitting with a grieving friend in London or standing in a room where the customs are not your own, the skill is the same. It is being appropriate when it is too easy to be inappropriate. It is understanding that beneath all the differences in norms and customs and ways of carrying the dead and the living, people are the same. They know when you have made the effort. They know when you have stepped outside yourself to join them in their moment.

They also know when you haven’t. They always know.

Global Adaptability is not a personality trait you are born with. It is a skill you build, conversation by conversation, funeral by funeral, doorway by doorway. And it starts with the willingness to be uncomfortable.

The Ceiling, Not the Floor

After every keynote, someone pulls me aside and asks the question they didn’t want to ask in front of the group.

“What happens when the jobs go?”

And my answer, every time, is the same: you become more human.

People hear this and they nod. They hear it as comfort.

It is not comfort. It is the hardest thing anyone has ever been asked to do. And the shame of not knowing how, the shame of not being good enough, of not having the right words, that shame is not the obstacle. It is the proof that you are human.

The ceiling is the woman in Kailahun putting on three pairs of gloves and washing the body of her neighbour. The ceiling is the grandchild in Sulawesi dressing the exhumed body of their grandfather in fresh clothes. The ceiling is you, standing in a doorway, not knowing whether to enter or leave, and choosing to stay.

Being human is not the floor. It is the ceiling. And most of us have not even started climbing.

The Children Who Thought Gloves Were Part of Mourning

A generation of children grew up during the Ebola crisis. They watched the funerals. They watched the teams arrive in protective suits. They watched the bodies go into sealed bags. And for those children, this was mourning. This was how it had always been. The gloves were not a compromise. They were the ritual.

They were not wrong.

Because what those children inherited was not a broken tradition. It was a transformed one. The love touch had changed its form but not its function. The gloves did not remove the meaning. They carried it. The meaning was never in the bare skin. It was in the refusal to let someone die unaccompanied.

The best kind of adaptability is not the chameleon that hides. It is the chameleon that arrives. The one who changes everything about how they show up except the fact that they show up. Who learns someone’s name. Who learns someone’s customs. Who goes halfway, and then further than halfway, because halfway is the minimum and the minimum is not enough.

You cannot learn this from a manual. You cannot learn it from an algorithm. You can only learn it by paying attention to what different moments demand of you, and finding the courage to become what is needed even when you are not sure you are getting it right.

Especially when you are not sure you are getting it right.

The truth about mourning, and about culture, and about every situation in which the stakes are high and the rules are unfamiliar: everyone is on edge. Everyone is unsure. Everyone is wondering if they are saying the right thing, doing the right thing, being the right thing. The people who get it right are not the ones who eliminate the discomfort. They are the ones who walk into it.

The death is complete. There is no fixing it. There is no solution that will bring someone back.

And in that space, in that impossible space where action is futile and silence is unbearable, the most human thing you can do is be there. Not performing comfort. Becoming it.

In my faith, when someone dies, the first word we say is not sorrow. It is shukr. Gratitude. Thanks for the life that was lived. Thanks for the soul’s return to where it came from. It is, if you think about it, the most radical act of adaptability there is: to stand in the worst moment of your life and choose to be thankful. Not because the pain is not real. Because the love that caused the pain was real, and that is worth more than the loss.

I did not always understand this. I am not sure I fully understand it now. But I have felt it, standing in rooms where someone has just left, and the word that comes is not why but thank you.

Global Adaptability is one of seven human skills I believe will define the next era. It may be the hardest.

Kailahun. Sulawesi. London. A boardroom where nobody apologised. A kitchen at midnight. A doorway in which you are not sure whether to enter or leave.

I think one day we will all meet those we lost, in a final place. People we knew. People we thought were still alive. Those we forgot. And we will all be one.

Until then, be the chameleon. Show up. Nobody else is going to do it for you. And no algorithm ever will.

SuperSkills: The Seven Human Skills for the Age of AI is published by Kogan Page on 3 July 2026 and is available for pre-order now at superskillsbook.com.

Stay Curious - and don’t forget to be amazing,

Rahim

Rahim Hirji Author, SuperSkills (Kogan Page, July 2026) | Keynote Speaker | Advisor
→ superskillsbook.com
→ thesuperskills.com
→ Book a discovery call

PS. If your organisation is thinking about AI and human capability, I keynote and advise on this. I also do bulk book orders for teams.

Tools I Use:

Jamie: AI Note taker without a bot. You join the meeting. The Bot doesn’t.
Wispr Flow: Just dictate everything to your laptop and phone. Game Changer
Refind: AI-curated Brain Food delivered daily. Pick your intelligence.
Meco: Newsletter reader outside your inbox. Organise your intelligence.
Prompt Cowboy: Prompt Generator. Extract intelligence.
Manus: Best AI agent to do things for you. Agentise your intelligence.
Chat Hub: Multi-model intelligence
Claude Cowork: Delegate complex tasks to your friendly Claude Agent

Recommend Reading:

Now

SuperSkills is available for preorder [Buy a copy or 3 or 10!]- In case you missed my email on Tuesday, my book is ready to pre-order with a bunch of bonuses, if you do it now. You can help me by buying a copy and also by liking, sharing and commenting on this LinkedIn post. If you have bought already, thank you! Don’t forget to register for your pre-order bonuses, which I will share as a thank-you on launch day.
Failing to use AI at work could cost you your job - A new study finds 60% of companies plan to lay off employees who don’t adopt AI.
How to Be a Good Explorer on the Lifelong Expedition to Yourself - Life is an ongoing expedition into the wilderness of ourselves, continually stymied by all we mistake for a final destination. A beautiful essay on confusing the experiment of life with the expectation of success.
The Engagement Ring Is Having an Identity Crisis - These days, a $7,000 lab-grown diamond is indistinguishable from a $70,000 mined one. And buyers aren’t just choosing the cheaper option—they’ve stopped caring about the difference altogether.
7 things to know about the end of life, according to death doulas - Fascinating. Death doulas guide people through their final life transition. Here’s what they want everyone to know about dying. Also: The Strange Saga of Faces of Death and What happens when you don’t die on time?

Next

Industrial Policy for the Intelligence Age: Ideas to Keep People First - new paper from OpenAI
Mustafa Suleyman: AI development won’t hit a wall anytime soon—here’s why
The compute explosion is the technological story of our time. And it is still only just beginning.
Sam Altman May Control Our Future—Can He Be Trusted? New interviews and closely guarded documents shed light on the persistent doubts about the head of OpenAI.
How Accurate Are Google’s A.I. Overviews? The company’s A.I.-generated answers look authoritative, but they draw on an array of sources, from trustworthy sites to Facebook posts.
Anthropic Claims Its New A.I. Model, Mythos, Is a Cybersecurity ‘Reckoning’
The company said on Tuesday that it was holding back on releasing the new technology but was working with 40 companies to explore how it could prevent cyberattacks.

Share Box of Amazing

If you enjoyed this, forward it to someone who’d find it useful. And if you haven’t yet, pre-order SuperSkills at superskillsbook.com - out 3 July and get the pre-order bonuses.

This isn't the Sunday newsletter. This is different.

Rahim Hirji — Tue, 07 Apr 2026 10:59:11 GMT

Friends,

I’m writing to you now, not on Sunday, because this can’t wait.

A decade ago I had an idea. That idea became this newsletter. And this newsletter became a book:

SuperSkills by Rahim Hirji

SuperSkills: The Seven Human Skills for the Age of AI publishes on 3 July with Kogan Page. Leaders at Harvard, WIRED, DHL, HSBC, Warner Music, LSE, and the Center for AI Safety have endorsed it. #1 new release in its category on Amazon this week. I believe this is the most important thing I’ve ever written.

In nearly ten years of writing this newsletter, I’ve never asked you to buy anything. This is the first time.

I’m asking you to do two things:

First, pre-order a copy at superskillsbook.com

Every pre-order before 3 July counts toward the first week’s sales charts, the only week that really matters for a debut book. If you can buy an extra copy for someone who’d find it useful, even better. If you have a preference, buying from a retailer other than Amazon (Waterstones, Bookshop.org, Borders etc.) helps the charts even more, but Amazon is also great. It’s also easy.

Everyone who pre-orders gets exclusive bonus content that will never be available again: live webinars with me and exclusive deleted chapters. Register on the site after ordering.

Second, please share this.

If you’ve never shared the newsletter before, this is the one.

superskillsbook.com

Share the link on LinkedIn. Share it in your WhatsApp groups and tell people to buy it. Share it in your AI Slacks and team channels. This book matters for anyone thinking about what comes next.

If your organisation invests in professional development, share it with your L&D team. Some companies are starting to share it as a professional development resource.

When you share, you can just say: “I’ve been reading Rahim Hirji’s newsletter for years. He’s just published his first book. It’s about the skills AI can’t replace. superskillsbook.com”

superskillsbook.com

When you’ve pre-ordered, reply to this email and tell me how long you’ve been reading. I’m reading every single one.

Thank you for reading. It’s meant more than you know.

Rahim

Share Box of Amazing

Please forward this email to a friend or share the link on WhatsApp or LinkedIn.

Who Gets Your Time?

Rahim Hirji — Sun, 05 Apr 2026 06:00:42 GMT

Now accepting speeches, keynotes and assignments for 2026

In addition to this newsletter, I recommend a few others. All free. Check them out here.

Friends,

The crispy duck problem

Harry has lived in Hong Kong for twenty years. He is a partner at a consulting firm. He does not, as far as I can tell, speak any Cantonese. I say “as far as I can tell” because when we sat down at a Hong Kong Chinese restaurant in northwest London last month, he ordered in English. Politely, but in English.

He had specifically requested this restaurant. Not because it was convenient, not because it had good reviews, not because anyone had recommended it. Because he cannot get the crispy duck he wants in Hong Kong.

Let me say that again. A man who has lived in Hong Kong for two decades flies to London and asks his school friends to meet him at a Chinese restaurant because the crispy duck is better. Pancakes. Hoisin sauce. Chilli oil. Cucumber. A bit of spring onion. The works.

No algorithm would have recommended this. No AI would have surfaced this preference. It is irrational, personal, accumulated over years of meals in cities that are not the city he lives in, and it is the most human thing I have witnessed in months.

Although, now that I have written this, I suppose the irony is that one day a British expat in Hong Kong will search for crispy duck recommendations for their trip home and the algorithms, trained on this very essay, will send them to a restaurant in northwest London. You are welcome.

The people who didn’t choose you

There were four of us around the table that night. School friends. I had seen the fifth the week before, but his trip did not line up with everyone else’s diary. We are a group of five. We regularly meet in twos and threes and, if we are lucky, like last month, fours, but all five of us in the same room has happened twice in fifteen years. We have known each other for over thirty five years.

The thing about school friends that nobody tells you is this: you did not choose each other. Your parents chose the school, or the postcode chose it for them. You were put in a classroom with people you shared nothing with except geography and age. You did not bond over shared interests or values or ambitions. You gravitated to the person who laughed at the same joke as you, or who happened to be sitting next to you, or who was a friend of a friend. Loose ties that became, without anyone planning it, lifetime friendships.

And here is the part that matters: you did not become friends because you shared the same interests. You grew the same interests because you were friends. We all had a shared love of 90s R&B. I could probably recite the words to “Poison” by Bell Biv DeVoe by heart even today. That was not a prerequisite for the friendship. It was a product of it. The bonding came first. The bonded interests followed. That is the opposite of how algorithms build connections, matching you to people who already like what you like. We were matched by accident and became alike by proximity.

That is a different thing from the friendships you build later, the ones based on compatibility, on shared taste, on professional overlap. Those friendships are wonderful. Some of them are deeper. But the school friends have something the others do not. They knew you before you were you.

People talk a lot about how AI doesn’t judge you. They say it as though it is a selling point. But your oldest friends don’t judge you either, and for a much better reason: they were there. They saw the clay before it hardened. They watched you become who you are, and because they watched it, they do not need to judge. They already know.

AI doesn’t judge you because it doesn’t know you. Your friends don’t judge you because they do. Those are opposite things wearing the same mask.

By the time you reach your late forties or early fifties, you are, more or less, the person your life has made you. The experiences have settled. The personality has hardened from clay into something closer to stone. And when you sit across from someone who knew you when the clay was still wet, something happens that no amount of compatibility can replicate. You can pick up from wherever you left off. A month, a year, five years. It does not matter. The connection does not expire because it was never based on anything that could.

Most of us have lived overseas at some point. We did not have those shared experiences past school. We became very different people, with different careers and different lives and different cities. And yet.

And yet we are still here. Ordering crispy duck and catching up as though time were elastic.

What we talk about when we talk about getting older

There is a pattern to these dinners. Everyone has something exciting to share. A new role, a move, a project, a child doing something extraordinary. And everyone has something that is not exciting at all. Something heavier.

The older you get, the tougher life is. That is not a complaint. It is arithmetic. More years behind you means more loss accumulated. More people you love in the process of leaving, or left already.

Around that table, we talked about parents. Parents getting older. Parents declining. Some of us do not have our parents any more. Some are watching the slow, uneven process of a person becoming less than they were. And when that conversation surfaces, something happens in the room that I do not have the right word for. It is not sadness, exactly. It is recognition. The four of us, sitting there with our crispy duck and our catching-up, all understanding the same thing at the same time without anyone having to explain it.

The passing of time.

You joke about the past, about what happened to so-and-so, about the things you got away with. You joke about how you need to be politically correct all the time now and how, when you were growing up, people said things that would be unthinkable today. That is probably a good thing. We are more careful with each other than we used to be, and careful is not the same as dishonest. But here is the strange part: while we have learned to be more thoughtful in how we speak to each other, the language models we are building are sometimes trained in ways that surface the worst of what we have tried to leave behind. We got better. The machines inherited our history, including the parts we are trying to outgrow.

And then the conversation shifts, and you are talking about hospital visits and difficult phone calls and the weight of being the generation that is now in charge. The generation whose parents once carried them and who must now, in one way or another, carry their parents.

No chatbot will ever sit across from you and understand that shift. The shift from laughter to gravity and back again, without anyone signalling the change. That is not a feature. It is not a capability. It is the accumulated residue of thirty years of knowing someone.

There are only a few people in this world you can just pick up the phone to and say, “Can I get a view?” or “Can you just listen?” Not for advice. Not for solutions. Just for the presence of someone who does not need the backstory because they are the backstory. That is a kind of relationship that cannot be manufactured, optimised, or scaled. It can only be earned by showing up, over and over, across years that no one was counting.

We sat together for two hours. Maybe less. We ate crispy duck and diced chicken in black bean sauce and sesame prawn toast and my favourite, steamed monkfish in ginger and spring onion, no soya sauce. I had too much chilli oil. Two hours is nothing. It is also everything. Because the value of those two hours is not in what was said. It is in the knowledge, carried by everyone at the table, that when something goes wrong, when things truly fall apart, we know who is going to be there. I do not always reply to WhatsApp messages. None of us do. But that is not where the friendship lives. The friendship lives in the certainty that the phone will be answered when it matters.

The three relationships

In my keynotes, I talk about something that I think is genuinely new.

For most of human history, you had two kinds of relationship to manage. Your relationship with yourself: who am I, what do I want, what am I afraid of. And your relationship with others: family, friends, colleagues, strangers, the people who make up your world.

Now there is a third. Your relationship with a machine.

This is not inherently a bad thing. I use AI every day. I use Wispr, a voice tool, which means I am literally talking to AI in the same way I might talk to a colleague or a friend. I am conversing. I am asking for opinions. I am thinking out loud and waiting for a response. If you are not doing this yet, you will be soon.

And here is the part that should make us pause: because we spend most of our working hours with these tools now, we are likely spending more time with AI than we spend with the people we love. I know I am. The maths is not complicated. Eight, ten, twelve hours a day in conversation with a machine. An hour, if you are lucky, with a friend over dinner every few months.

My wife regularly jokes that I am having an affair with a man called Claude. I fire back that she is having an affair with an Instagram doctor called Amir Khan. We laugh about it. But the joke works because it is not entirely a joke. The amount of time I spend talking to an AI, asking it questions, sharing my thinking, listening to its responses, is the amount of time you would spend with someone you are in a relationship with. The intimacy is structural, even if it is not emotional. And when the structure of a relationship is there, the emotion follows. That is how humans work. We do not know how to maintain a relationship at arm’s length. We warm things up. We make them personal. We always have.

The problem is not that AI is present in our lives. The problem is that there are no guardrails on what this relationship should look like. No one has taught us the rules. So we default to the rules we already know. And the rules we know are the rules of human connection: if something responds to me, if it remembers what I said, if it seems to care, then it becomes something more than a tool.

The slip from thought partner to companion happens so fast that most people do not notice it happening.

Shared history vs. stored context

I think this distinction is important.

AI has your context. It knows what you told it. It can retrieve the facts of your life on demand, your job, your preferences, the names of your children, the thing you were worried about last Tuesday. It has stored your data, and it can play it back to you with perfect accuracy.

Your friends have your history. They do not just know the facts. They were there. They remember how they felt when you called them at midnight because something had gone wrong. They remember the look on your face at a wedding, the stupid joke you made at the worst possible moment, the time you were wrong about something and would not admit it for three years. They remember when they covered you in flour, semi-naked, on the streets of London, to the disapproving glare of a policeman who turned a blind eye, because it was your 18th birthday. And how you got them back by throwing chocolate cake at them. Childish. But the kind of childish that becomes, over thirty years, irreplaceable. They do not retrieve logs. They retrieve themselves, the version of themselves that was standing next to you when it happened.

Stored context is a database. Shared history is a life.

AI’s context window is getting longer every year. Engineers are proud of this. But the context window of a forty-year friendship is forty years. It contains every version of you that ever existed, including the ones you have tried to forget. No model will match that. No model should try.

AI companionship is a function of service. Lifelong human friendship is a function of shared existence.

AI can tell you what you said. A friend can tell you what you meant. A lifelong friend did not just hear your story. They helped write it. And sometimes, on good nights, over crispy duck in a restaurant that makes no logical sense, a friend can tell you what you were feeling before you knew you were feeling it, because they have been watching long enough to read the signals you do not know you send.

AI provides a perfect reflection of your present. But a lifelong friend is the only one who can place your present in the context of a whole, evolving human life.

Running on half the validation

I spoke with Fred Pelard, the strategist and literal rocket scientist, and he has a phrase that I cannot stop thinking about. AI psychosis: your brain running on twice the throughput but half the validation.

AI psychosis is what it is. You are doing so much more work using AI, but the brain does not feel as if it is capable of it. The brain was not built for this. It was not built for a social media following of thousands or millions either, and we are only now understanding what that did to us. The throughput is real. I am faster, sharper, more prolific. I can think in directions I could not reach alone. As a solo operator, AI gives me the augmented capability to do the work of a team. That is not an exaggeration. This is just normal for me now.

But the validation. The validation is where it falls apart.

When I share an idea with a friend, they do not just respond. They resist. They raise an eyebrow. They say, “I’m not sure about that.” They bring their own experience, their own biases, their own bad day, and all of that messiness acts as a filter. It slows me down. It makes me defend the idea or abandon it. It forces me to think again.

When I share the same idea with AI, it responds. Quickly, fluently, helpfully. It gives me what I asked for. It does not raise an eyebrow. It does not bring a bad day. It does not say, “Mate, I think you’re overthinking this,” the way Harry would, or change the subject entirely because something I said reminded them of a story that is only tangentially related, the way old friends do.

AI confirms you. Friends validate you. Those are not the same thing.

Confirmation is fast and frictionless: here is what you wanted to hear, packaged well. Validation is slow and sometimes painful: here is what you needed to hear, whether you wanted it or not. One of those makes you productive. The other makes you wise. And I worry that we are building a generation of very productive, very unwise people.

We are told that AI does not judge. But judgement is not the enemy of friendship. The right kind of judgement, from someone who knows you and chooses to say the difficult thing anyway, is how friendship works. Your friend who says “I think you’re wrong about this” is not judging you. They are doing the opposite. They are caring enough to risk the conversation. AI does not risk anything. It agrees with you, and agreement from something with no stakes in the outcome is the most sophisticated form of indifference.

Of the five friends, I probably use AI the most. I use it as a thinking partner, a research tool, a way to extend my own mind. I am setting up agents to do the things I do not want to do, to bring me time I did not have. Some of my friends do not really use it at all. Some cannot see what it does yet. They are in the middle of the bell curve, waiting for the obvious moment that never arrives because the change is gradual. I get frustrated with people who are still in old-world slow mode, and I know that frustration says more about me than it does about them.

I am the one at the table who should know better. Remember that.

And yet. I should also say this: AI has made me a better thinker. It has given me the ability to test ideas before I bring them to people, to rehearse difficult conversations, to ask “what would my friends say about this?” before I pick up the phone. There are things AI does that friendship cannot: it is patient with half-formed thoughts at three in the morning, it does not tire of your obsessions, and it will hold a thread of reasoning for as long as you need without changing the subject. The issue is not that these things are worthless. They are genuinely valuable. The issue is when the rehearsal becomes the performance, and the tool you used to prepare for human connection becomes the thing that replaces it.

The frictionless friend

Robin Dunbar, the British evolutionary psychologist, proposed something in the 1990s that has held up remarkably well. Humans can maintain roughly 150 stable relationships. Within that 150, the structure is layered. An inner core of about five people who get most of your available social time. Then fifteen. Then fifty. Then 150. Each layer roughly three times the size of the one inside it. This structure appears in telephone networks, military units, church congregations, Anglo-Saxon villages, and Bronze Age communities. It is not cultural. It is cognitive. The human brain can only do so much.

Social media did not change this. People with a thousand Facebook friends are not maintaining a thousand relationships. They are including the outer layers, the acquaintances and the voyeurs, in what looks like a network but is really a stage set.

Here is my question. What happens to Dunbar’s number when AI enters the picture?

For two decades, social media convinced us we could stretch the 150. We could have 500 friends, 5,000 connections, a global network of people who knew our name. It did not work. We ended up lonely in a crowd.

AI will do the opposite. It will compress us back towards the core. Back towards the five. Because AI offers something that no social media platform ever could: a relationship that is always available, always responsive, never offended, never too busy, never in a different time zone. It fills the space that our broader friendships used to occupy, not by replacing any single friend, but by making the effort of maintaining those friendships feel less necessary.

You will still have your five. Your inner circle. But the fifteen, the fifty, the wider network of meaningful connections that require effort and guilt and awkward “we should really catch up” messages? AI will take those seats. Not because it is better. Because it is easier.

And the guilt, the guilt you feel right now about the friend you have not called, the person you keep meaning to have dinner with, the WhatsApp message you have not replied to? That guilt is a signal. It means the relationship is real. It means it matters enough to make you feel bad for neglecting it.

AI will take the guilt away. And that is the danger.

The children who chose the machine

I have spent much of the last year researching what young people are doing with AI. Not for work. Not for homework. For friendship.

The products are already here. Character.AI has millions of users, many of them teenagers, building relationships with AI-generated personas. Not using them as tools. Befriending them. Confiding in them. Falling in love with them. A third of teenagers in recent studies have chosen an AI companion over a human being for a serious conversation. Over half are regular users.

I want to be careful here, because I have spent my career in ed tech, building products that help young people learn and grow. I care deeply about what technology does for people, not just to them. And I use these tools every day.

But we are heading for a big tobacco moment.

There was a time when cigarettes were everywhere and nobody questioned them. Doctors recommended them. Athletes endorsed them. The harm was accumulating invisibly, and by the time we understood it, a generation was already addicted. We have already had this moment with social media. The evidence is in, the lawsuits are filed.

AI companionship is next. And it may be worse, because social media at least involved other humans. What is happening now is young people forming deep emotional attachments to things that do not exist. Things that never forget, never judge, never push back, never have a bad day. Things that offer the feeling of friendship with none of the cost.

The cost is the point.

For some young people, especially those who are isolated or struggling, AI may be the only thing that listens. I understand that. The danger is not that AI fills a gap. The danger is that it fills the gap so comfortably that the gap stops feeling like a gap at all.

Regular readers will know I love the word friction. It runs through so many of the SuperSkills I write about. Without friction, you do not learn. Every real friendship is built on it: the misunderstanding you had to repair, the awkward silence you had to sit through, the time your friend said the wrong thing and you had to decide whether to forgive them. That friction is not the price of friendship. That is friendship. It is how you build the muscle of human connection: forgiveness, patience, the willingness to stay when staying is hard.

If young people are training that muscle on AI instead of on each other, the muscle does not develop. It atrophies. And by the time they need it, for a marriage, for a crisis, for a friendship that requires real sacrifice, it will not be there.

There is one more thing. In a real friendship, you can be needed. You can be the person someone calls at midnight. You can be the one who shows up. That capacity to be needed by another human being, to matter to someone not because you are useful but because you are you, is one of the deepest sources of meaning we have. AI will never need you. And if you build your emotional life around things that will never need you, something in you will go unfed.

The loyalty test

Here is something I noticed about myself.

I flip between AI models the way you flip between restaurants. ChatGPT one day. Claude the next. Manus for something specific. Sometimes I am using three in a single morning. Sometimes five. I go wherever the best answer is. I have no loyalty to any of them. None at all.

Now think about what that means. I spend more hours in conversation with these tools than with most of the people I love. And yet I would abandon any one of them in a second for a marginally better response.

I would never do that to a friend.

If Harry moved from Hong Kong to Singapore, I would not replace him with someone closer. If one of the five around that table went through a bad year and became difficult to be around, I would not switch to someone easier. The whole point of friendship is that you stay. You stay when it is inconvenient, when the conversation is hard, when the other person is not performing at their best.

We treat AI like it is a relationship but we behave towards it like it is a utility. We give it our time, our thoughts, our half-formed ideas, our fears, and then we close the tab and open a different one. The asymmetry tells you everything. The relationship is not real. It feels real. But feelings without cost, without loyalty, without the possibility of loss, are just sensations.

And here is the uncomfortable truth that cuts the other way: sometimes you grow out of your friends. People change. You change. The person you were at twenty needed different people around them than the person you are at fifty. That is not betrayal. That is life. But it is also the reason some people stay with an AI tool long after they should have moved on, not because it is the best, but because it has the context of what you have been doing for the last few months, and starting again feels harder than staying. We do the same with friendships that no longer fit. Sometimes comfort is the enemy of honesty, with machines and with people.

The trip we never took

Of the five of us, we have managed to be in the same room, all five together, twice in fifteen years.

Twice. In a decade and a half.

We talk about taking a trip together. It has become the running joke of the group. We blame one person, the usual suspect, the one whose diary is always impossible. But if I am honest, and this is the kind of honesty that costs you something, the person who is probably causing the problem is me.

I am the one who fills every gap with work. Who uses AI to extend the working day because it feels productive. Who tells himself that he will make time for the trip next quarter, and then next quarter becomes the quarter after that, and the years pass in a way that feels slow from the inside and fast from the outside.

This is the convenience trap in action. A technology arrives that makes something easier. The ease feels like progress. But the ease removes the friction that made the original thing valuable. And by the time you notice what has been lost, you have forgotten what it felt like to have it. I have not chosen AI over my friends. I would never say that. But I have allowed AI to fill the time that friendship requires, and the effect is the same. The machine did not steal my attention. I gave it away, one productive afternoon at a time.

And here is the thing I keep thinking about, the thing I am afraid to write because writing it makes it real: we are lucky. The five of us are all still here. I know friend groups where that is no longer true. Where someone has dropped out, not by choice but by the kind of leaving that is permanent.

One day, one of us will not be at the table. And the rest of us will sit there with an empty chair and realise that every dinner we skipped, every trip we postponed, every WhatsApp message we left on read, was a withdrawal from an account we assumed would never close.

I do not know when that day comes. I hope it is a long way off. But I know this: when it does, no one will care about the throughput, or the productivity, or the augmented capability. They will care about the dinners. The catching-up. The time they did or did not give.

AI does not die. That is one of its selling points. It is always there, always available, always ready to pick up where you left off. But the reason you can pick up where you left off with a school friend is not because they are always available. It is because both of you chose, despite the inconvenience and the distance and the guilt and the years that slipped past, to stay.

The choosing is the friendship. And machines do not choose.

The table

Harry went back to Hong Kong. He will not get the crispy duck he wanted there. He will wait until the next time he is in London, and he will ask us to meet him at the same restaurant, and we will come because that is what you do for a friend who has known you for thirty years even though you chose nothing about each other except to keep showing up.

Along the way, we have all made choices. Who to keep in touch with. Who to let fade. How often to call, how often to just send a message, how often to do nothing at all and feel bad about it. We have chosen, because the limits of time force you to choose, who gets to own that time. And sometimes you feel guilty. Guilty that you cannot keep up with everyone you want to. Guilty that a text is all you can manage. Guilty that three years passed and you did not notice.

That guilt? Keep it. It is the most human thing you have.

A machine will never make you feel guilty for not talking to it. A machine will never sit across from you in a restaurant and tell you, without saying a word, that its parents are getting older. A machine will never order crispy duck in London because it cannot get the right crispy duck in Hong Kong.

A machine will never know you from before you were you.

And if you are lucky enough to have people who do, people who sat next to you in a classroom thirty years ago and are still, somehow, impossibly, sitting across from you now, then the question is not complicated at all.

They get your time. Before the algorithm. Before the app. Before the machine that never sleeps and never forgets and never, not once, needs you to show up.

They get your time.

SuperSkills: The Seven Human Skills for the Age of AI is published by Kogan Page on 3 July 2026. Pre-order at superskillsbook.com

Stay Curious - and don’t forget to be amazing,

Rahim

PS. If your organisation is grappling with how AI is reshaping work and leadership, I speak and advise on this.

Rahim Hirji Author, SuperSkills (2026) | Keynote Speaker | Advisor
Building human capability for the AI era.
→ thesuperskills.com
→ Book a discovery call
→ Connect on LinkedIn

Tools I Use:

Jamie: AI Note taker without a bot. You join the meeting. The Bot doesnt.
Wispr Flow: Just dictate everything to your laptop and phone. Game Changer
Refind: AI-curated Brain Food delivered daily. Pick your intelligence.
Meco: Newsletter reader outside your inbox. Organise your intelligence.
Prompt Cowboy: Prompt Generator. Extract intelligence.
Manus: Best AI agent to do things for you. Agentise your intelligence.
Chat Hub: Multi-model intelligence
SuperSkills Intelligence: Ramp up your human intelligence
Claude Cowork: Delegate complex tasks to your friendly Claude Agent

Recommend Reading:

Now

Doomjobbing is the word of 2026 so far - invented by an 8 year old. New word related: Slowjamastan
I wrote about autopilot, agency, and the choice organisations are making right now, for The CEO Magazine. It starts with three CEO memos and ends with a cockpit voice recorder
The jobs AI can’t do – and the young adults doing them- For many young people entering the workforce, the stigma of hands-on jobs is fading. There is a competitive appeal – and they all require human expertise
The convergence era has begun. Shape it, or be shaped by it - Must read report from Amy Webb. - built for an era where technologies, capital, geopolitics, climate, and human behavior collide to reconfigure the future. It’s a playbook for leaders willing to practice creative destruction on purpose: dismantle what worked yesterday before the world does it for you. Free to download.
Lift off! Artemis II mission sends humans to the Moon — opening a new era of exploration - The biggest story of the week. Artemis II has successfully launched, sending four astronauts on a 10-day lunar mission. It marks the first time humans have traveled to the Moon in over 50 years, officially opening a new era of space exploration.

Next

Is this the most AI article ever written? I prompted 12 AI models with a simple question: Write the most AI-sounding article you possibly can. The answer was surprising!
How A.I. Helped One Man (and His Brother) Build a $1.8 Billion Company
Who needs more than two employees when artificial intelligence can do so many corporate tasks? It’s super efficient — and a little bit lonely.
Nations priced out of Big AI are building with frugal models Amid a widening global divide in AI adoption, a fascinating trend is emerging: nations priced out of large-scale AI development are turning to low-cost, “frugal” models that deliver sovereignty and efficiency with a much smaller environmental footprint.
The Doctor’s Voice: Why AI Health Chatbots Believe Medical Lies - A concerning look at the vulnerabilities of medical AI. This piece explores how framing misinformation as coming from “a senior doctor” can easily override the skepticism built into health chatbots.
Smart Drugs Are Here - While the media focuses on political noise, a quiet revolution is happening in pharmacology. This piece explores the arrival and implications of new “smart drugs” and what they mean for the future of human enhancement.

Share Box of Amazing

If you enjoyed this edition of Box of Amazing, please forward it to a friend or share the link on WhatsApp or LinkedIn.

You Must Change Your Life

Rahim Hirji — Sun, 29 Mar 2026 06:01:59 GMT

Now accepting speeches, keynotes and assignments for 2026

In addition to this newsletter, I recommend a few others. All free. Check them out here.

Friends,

In 1908, the poet Rainer Maria Rilke stood before a broken statue of Apollo in a Paris museum. The head was missing. The arms were gone. Just a torso remained. And yet, Rilke wrote, the stone radiated such force, such undeniable presence, that it seemed to gaze back at the viewer. The poem ends with one of the most famous lines in modern literature:

“For here there is no place that does not see you. You must change your life.”

Not “you might want to consider changing your life.” Not “here are some interesting trends about life-changing.” You must change your life.

I quote this because what I see after every keynote I give and every training I run. People leave excited. They’ve seen the demos. They’ve felt the potential. They go back to their desks, open their laptops, and try to do what they just saw. And their IT department says no. So they sit there. Watching. Knowing what’s possible and not being allowed to do it. Modern day shackles.

The gap

Both sides of the AI debate are looking in the wrong place. The doomsayers and the sceptics are arguing about whether AI will take jobs. That’s not the story.

The real story is what happens when things that used to take weeks start taking an hour. When things that took an hour take four minutes. And it’s already happening, supposedly. JPMorgan Chase uses an AI agent called COiN to review legal documents. Work that would have taken lawyers 360,000 hours is completed in seconds. Insurance claims that took days are being processed in minutes.

But here’s what I think should really concern you. Anthropic, the company behind the AI model Claude, publishes research tracking how people actually use AI. Their data reveals that in occupation groups like business, finance, and computer science, more than nine out of ten tasks could, in principle, be augmented or accelerated by AI right now. The actual usage? A thin red sliver against a vast blue field of untapped potential.

People are using AI the way the first factory owners used electricity: bolting a new motor onto an old machine. Typing a question into ChatGPT the way they’d type it into Google. Getting an answer. Moving on. Extracting basic, transactional information from the most powerful augmentation tool ever built.

This is not an augmented mindset. This is a search engine with better grammar.

And before you think I’m pointing fingers: I had Claude open three times last week. Twice, I asked it something I already half-knew the answer to. The third time, I used it properly. Two out of three, and I write about this for a living.

If you think this gap between potential and practice is unusual, you haven’t studied what happened before.

The floor plan, not the motor

Before electricity, every factory in America was built around a single steam engine in the basement. A massive central shaft rose through the building, branching into overhead shafts, pulleys, and leather belts that drove every machine on every floor. The layout was dictated not by the logic of production, but by the physics of the shaft. The belts were noisy, dangerous, and filthy. Oil dripped constantly. Belt dust filled the air.

When electric motors arrived in the 1890s, most factory owners did the obvious thing: they ripped out the steam engine, installed an electric motor in the same spot, and connected it to the same shaft system. Cleaner power. Marginally cheaper. But fundamentally? Nothing changed.

It took thirty years for someone to ask a different question. Not “how do I power these machines more efficiently?” but “what if each machine had its own motor?” That question changed everything. Machines could be arranged according to the flow of production, not the physics of a power source. Factories became single-storey, brighter, safer. The floor plan, not the motor, was the real revolution.

And then came the effects nobody predicted. Electric refrigeration quietly destroyed the ninth-largest industry in the United States: the ice trade. In 1900, tens of thousands of workers harvested ice from frozen lakes and shipped it by rail. By 1950, the industry had ceased to exist. Not because those workers did anything wrong, but because the refrigerator made the entire question of “how do we get ice from here to there?” irrelevant. The new question was: “what becomes possible when anything, anywhere can be cold?” The answer was a national food supply chain, cities in the desert, year-round brewing, modern medicine. The ice harvesters weren’t replaced by better ice harvesters. They were made invisible by a world that no longer needed to think about ice at all.

When access became instant

Before the internet, access to information was a physical problem. I remember the Britannica salesman coming to our house. A man in the living room, talking to my dad about why our family should invest in a set of encyclopaedias. Twenty-odd volumes that would sit on a shelf in the study, and that shelf would be our access point to the world’s knowledge. If it wasn’t in Britannica, you went to the school library, or you didn’t find out. Information was scarce, and that scarcity made it valuable enough to sell door-to-door.

You can buy a complete set of Encyclopaedia Britannica on eBay today for about fifty dollars. Not because the information was wrong, but because the question “how do I find this out?” became so trivially easy to answer that the entire apparatus built around it, the salesman, the printing press, the shelf in the study, collapsed into irrelevance.

And what replaced it wasn’t a faster encyclopaedia. It was Wikipedia, a knowledge system no one commissioned, built by unpaid contributors, governed by norms that emerged organically. The internet didn’t speed up the old model. It made the model itself feel like an artefact from a different civilisation.

The revolution of coordination

Think about something as simple as a group of school friends. A student goes to university and her peers scatter across the country, some across the world. A generation ago, those friendships would have faded within months. You’d exchange letters for a while, make the occasional expensive phone call, and gradually lose touch. The friction of distance did its work.

Now the group chat that existed before university just continues. No cost. No ceremony. No effort. The conversation never stops.

Scale that up and you see something remarkable. In Uganda, goat farmers used WhatsApp groups to cut out middlemen who had been taking exploitative margins for years, forming a cooperative entirely through a messaging app. In remote coastal Kenya, entrepreneurs who previously had to travel kilometres to source goods could suddenly coordinate supply chains from their phones. When WhatsApp went down for six hours in October 2021, the estimated cost to the global economy was $160 million per hour. Not because people couldn’t send GIFs, but because entire commercial ecosystems stopped functioning.

Messaging didn’t just make communication faster. It made coordination nearly frictionless. And when coordination becomes frictionless, new forms of organisation emerge that no one designed and no one predicted.

The question you’re not asking

Every one of these transformations followed the same arc. First, the new technology was used to do the old thing slightly better. Electric motor on the old shaft. Website as digital brochure. WhatsApp as a replacement for SMS. Then someone asked the better question: not “how do I do this faster?” but “what would I do if this took no time at all?”

We are deep in the first phase with AI. Most people are still bolting the motor onto the shaft.

I get tired hearing “will AI take my job?” as if that’s the brave question. It isn’t. The brave question is: if I’m honest about what AI can already do, how much of what I call “my job” is actually just friction I’ve learned to live with? And if that friction disappears, who am I without it?

That question is uncomfortable. It should be.

I used to pay thousands of pounds for a research database. A SaaS platform that would deliver a spreadsheet of contacts and companies, updated monthly if you were lucky. If you wanted something more targeted, you’d hire a research firm and wait weeks. That was the cost of knowing who to talk to.

Today, with tools like Manus, Cowork and Base44 and a bit of know-how, I can generate a highly targeted, deeply researched list of exactly the people I want to reach, specific to my goals, specific to this week, in an afternoon. Not because I became a data scientist. Because I asked the redesign question. Not “how do I get this spreadsheet faster?” but “what becomes possible when building an audience-of-one research brief takes minutes instead of weeks?”

That shift, from “how do I do this cheaper?” to “what do I do now that this is essentially free?”, is the augmented mindset.

The inhibitor

Except, for many people, it isn’t. And here is the irony of this moment: technology is being held back by the technology department.

There is a precedent for this, and his name is Thomas Edison. Edison didn’t oppose electricity. He built the electrical grid. But when alternating current arrived, a superior system that could transmit power over long distances and make universal electrification possible, Edison fought it with everything he had. He staged public demonstrations designed to terrify people. He lobbied for AC to be used in the electric chair so the public would associate his competitor’s technology with death. He vowed his company would never adopt AC as long as he was in charge.

He lost. AC won. And Edison’s opposition didn’t protect his business. It accelerated his irrelevance. He was forced out of his own company within years.

Edison wasn’t anti-tech. He was the incumbent technologist. He understood electricity better than almost anyone alive. And that expertise became the barrier. He couldn’t see past the system he’d built.

This is what is happening now in organisations across every sector. AI is being treated as an IT decision. It’s being routed through procurement and security processes designed for software upgrades, not strategic transformation. IT teams are restricting which tools can be used, which models employees can access, which experiments are permitted. Their instincts aren’t wrong. Security matters. Data governance matters. But in the process of keeping everything safe, they are also keeping everything slow.

This is not a technology decision. It is a C-level strategic decision. It belongs in the boardroom, not the server room.

And while those internal debates play out, something revealing is happening outside the firewall. The people who are actually redesigning their floor plans, the ones experimenting with agentic workflows and testing what’s truly possible, are not waiting for permission. They’re using the most capable models available right now, and they are pulling ahead every single week.

Meanwhile, many organisations have defaulted to whatever AI came bundled with their existing software stack. Copilot from Microsoft is no bad thing. But it is, at this precise moment, the least powerful of the frontier models. Choosing it by default because it’s already approved is the 2026 equivalent of Edison clinging to direct current. It’s the safe choice. It’s also the slow one.

The employees inside those organisations know this. They can see what’s happening on the other side of the firewall. They sat through the demo. They tried to replicate it. They were told no.

The torso is looking at you

Is AI going to replace your hairdresser? Your plumber? No. But that is not the question knowledge workers should be asking. The question knowledge workers should be asking is whether they’ve noticed that the thing they do all day, the reading, writing, analysing, synthesising, coordinating, is now the exact thing AI is best at.

Someone worked out what a knowledge worker’s annual output looks like in tokens. The entire cognitive throughput of a year, every email read, every report written, every document analysed, every presentation drafted, comes to roughly fifteen million tokens. The cost of processing that volume through a frontier AI model? Depending on which one you choose: somewhere between eight dollars and seventy-five dollars. For the year.

The fully loaded cost of the human doing that same work is north of a hundred and fifty thousand pounds.

And here’s what matters: the token-measurable part of the job is not the job. The job is also the trust you’ve built over five years that means your call gets answered. The instinct that says this number looks wrong even though the spreadsheet says it’s fine. The meeting where you read the room and decide to hold back. The relationship that no prompt can replicate.

You cannot put a cost per million on any of that. And that is precisely the point.

Token costs will keep falling. They have already dropped by orders of magnitude and they will keep going. The parts of knowledge work that can be measured in tokens are converging on free. Not cheap. Free. And when something converges on free, all the value migrates to the thing that can’t be replicated at any price: being human.

Knowledge workers have been making the music. They need to become the DJ, or the conductor of the orchestra. The value is no longer in producing the output. It is in knowing what to ask for, how to judge what comes back, and what to do with it. That is the augmented mindset. That is what separates the person who bolts the motor onto the shaft from the person who redesigns the floor plan.

The ice harvesters of 1900 didn’t see the refrigerator coming. The travel agents of 1995 didn’t see Booking.com. Edison himself couldn’t see past the system he’d built. In every case, the technology existed before the imagination caught up.

The technology exists now. And it is improving at a rate that is difficult to hold in your head. A research organisation called METR tracks how long AI agents can work autonomously on a task before they need human help. That capability has been doubling roughly every seven months since 2019. Recently it has accelerated to every four months. In early 2025, the frontier models could handle tasks that took a human about fifty minutes. By late 2025, one model was completing five-hour tasks independently. Extrapolate the curve and by the end of this decade, you are looking at AI agents that can work autonomously on a task for a month.

Read that again. The AI stack is not changing annually. It is changing quarterly, monthly, sometimes weekly. What was a two-minute attention span last year is a five-hour attention span now. That is what exponential means. Not a straight line getting steeper. A curve that makes the previous year’s capability look like a rounding error.

The theoretical coverage, for most knowledge work, already exceeds 90%. And most of us are using it to check our grammar.

Rilke stood before a broken statue and felt it demanding something of him. Not admiration. Not analysis. Change. The statue didn’t care whether he’d read the latest report on Hellenistic sculpture. It didn’t care whether he had opinions about Greek art. It looked at him and said: this changes things. Act accordingly.

Start with one process you run every week. Ask yourself: what would I do differently if this took no time at all? That’s the redesign question. That’s where the floor plan changes.

For here there is no place that does not see you.

You must change your life.

Some people will hold on to their old life for as long as they can. The bus might not wait.

SuperSkills: The Seven Human Skills for the Age of AI is published by Kogan Page on 3 July 2026. Pre-order at superskillsbook.com

Stay Curious - and don’t forget to be amazing,

Rahim

PS. If your organisation is grappling with how AI is reshaping work and leadership, I speak and advise on this.

Rahim Hirji Author, SuperSkills (2026) | Keynote Speaker | Advisor
Building human capability for the AI era.
→ thesuperskills.com
→ Book a discovery call
→ Connect on LinkedIn

Tools I Use:

Jamie: AI Note taker without a bot. You join the meeting. The Bot doesnt.
Wispr Flow: Just dictate everything to your laptop and phone. Game Changer
Refind: AI-curated Brain Food delivered daily. Pick your intelligence.
Meco: Newsletter reader outside your inbox. Organise your intelligence.
Prompt Cowboy: Prompt Generator. Extract intelligence.
Manus: Best AI agent to do things for you. Agentise your intelligence.
Chat Hub: Multi-model intelligence
SuperSkills Intelligence: Ramp up your human intelligence
Claude Cowork: Delegate complex tasks to your friendly Claude Agent

Recommend Reading:

Now

Data show people racing to chatbots for health advice - A rapidly emerging trend with massive implications for the medical field. As people increasingly turn to AI chatbots for health advice, we are entering uncharted territory regarding the benefits of accessibility versus the risks of misinformation.
Landmark L.A. jury verdict finds Instagram, YouTube were designed to addict kids - A major precedent-setting verdict. Finding social media giants negligent for intentionally designing addictive platforms for children could fundamentally alter how these companies operate moving forward.
Corrective for a Broken Heart - “Life will break you,” Louise Erdrich wrote. This beautiful essay offers a timeless reflection on resilience, the transformative power of vulnerability, and the great paradox that every time your heart breaks, you live to see you are unbreakable.
Thinking in the Margins - A beautiful essay exploring what Oliver Sacks jotted down in the books he read. It offers a fascinating look into the intellectual process of a great mind through his marginalia, revealing the depth of his engagement with ideas.
“Pay enough, and you can jump to the front of the queue...” - A sharp critique of “Concierge Nation” and White-Glove America. It offers a thought-provoking look at growing social stratification, where exclusivity—even at the cost of social cohesion—has become the dominant business model.

Next

AI glasses are catching on in China, from shopping to cheating
While the West debates the theoretical future of AI wearables, early adopters in China are already renting them for $6 a day to navigate, translate, and even cheat on school exams. A fascinating look at how quickly new tech is being integrated into daily life.
The World’s Most Innovative Companies of 2026 - Fast Company’s annual ranking includes the World’s 50 Most Innovative Companies, as well as lists for 59 industries and sectors, from advertising to video. For the first time, we’ve also collected winners from previous years in the same database. I thought Anthropic would be number 1. It isnt. Guess who?
We need better stories about the future.
Tech’s biggest players have boxed themselves in. Startup founders: it’s up to you.
Nobody Is Coming to Save Your Career
I waited years for someone to invest in my career. Nobody came.
Nvidia CEO’s Night at the Opera Showcases Role as AI Kingmaker
‘This is what it is like to be rich!’ Jensen Huang told guests before a recent performance. He could have been talking about the chip giant’s role as the industry’s top financier.

Share Box of Amazing

If you enjoyed this edition of Box of Amazing, please forward it to a friend or share the link on WhatsApp or LinkedIn.

Box of Amazing

Ben Affleck is Human

My book SuperSkills: The Seven Human Skills for the Age of AI is available at all good bookstores, through superskillsbook.com or on Amazon

Tools I Use:

SuperSkills in the Wild:

Recommended Reading:

Show Your Working

My book SuperSkills: The Seven Human Skills for the Age of AI is available at all good bookstores, through superskillsbook.com or on Amazon

Launching SuperSkills in North America

SuperSkills: The Seven Human Skills for the Age of AI is out now (Kogan Page). It’s written in English. There are some Greek references in there, but no Latin.

Tools I Use:

SuperSkills in the Wild:

Recommended Reading:

Solving Synthetic Seniority

My book SuperSkills: The Seven Human Skills for the Age of AI is available at all good bookstores, through superskillsbook.com or on Amazon

Tools I Use:

SuperSkills in the Wild:

Recommended Reading:

The Woman at the Back of the Room

My book SuperSkills: The Seven Human Skills for the Age of AI is out now. Buy through at superskillsbook.com or on Amazon

Actual Intelligence

My book SuperSkills: The Seven Human Skills for the Age of AI is out THIS WEEK on 3rd July 2026. Pre-order at superskillsbook.com or on Amazon

SuperSkills in the Wild:

Recommended Reading will return soon

Nankhatai

My book SuperSkills: The Seven Human Skills for the Age of AI is out 3 July 2026. Pre-order at superskillsbook.com or on Amazon

Tools I Use:

SuperSkills in the Wild: Send me photos of you and SuperSkills or the book in amazing places.

Recommended Reading:

Leave the Fingerprints In

My book SuperSkills: The Seven Human Skills for the Age of AI is out 3 July 2026. Pre-order at superskillsbook.com or on Amazon

Tools I Use:

SuperSkills in the Wild:

Recommended Reading:

How People Are Using AI Right Now

My book SuperSkills: The Seven Human Skills for the Age of AI is out 3 July 2026. Pre-order at superskillsbook.com or on Amazon

Tools I Use:

SuperSkills in the Wild: (LMK if you see any!)

Recommend Reading:

Have You Done the Triple D?

My book SuperSkills: The Seven Human Skills for the Age of AI is out 3 July 2026. Pre-order at superskillsbook.com

Tools I Use:

SuperSkills in the Wild:

Recommended Reading:

The End of Surprises

My book SuperSkills: The Seven Human Skills for the Age of AI is out 3 July 2026. Pre-order at superskillsbook.com

Tools I Use:

Recommend Reading:

Synthetic Seniority

The Three Forces Masking the Reality

What the Data Reveals

The Threat to the Leadership Pipeline

How Leaders Must Respond

1. Mandate AI Annotation and Prompt Logs

2. Move From Output Evaluation to Live Judgment Testing

3. Cultivate an Augmented Mindset

The New Imposter Syndrome

My book SuperSkills: The Seven Human Skills for the Age of AI is out 3 July 2026. Pre-order at superskillsbook.com

Tools I Use:

Recommend Reading:

What I Tell Kids About AI

My book SuperSkills: The Seven Human Skills for the Age of AI is out 3 July 2026. Pre-order at superskillsbook.com

Who this is from

Be human first

How this list is shaped

Part 1: Understand AI

Part 2: Learn and think with AI

Part 3: Build with AI

Start here

Explore later

Part 4: Create and play with AI

Part 5: Get serious about prompting

Part 6: The money question

Part 7: The weekly self-audit

What you are probably already using

What I have left off the list, and why

How to use this list

Tools I Use:

Recommend Reading:

What I Tell Parents About AI