Now accepting keynotes for Q3/Q4 2025
The SuperSkills Era: Thriving in the Age of AI - I’m delivering 60-minute executive sessions for senior teams based on the research and frameworks from my upcoming book. Available across the UK, Europe, and globally in person or virtually. To book for your board, corporate team, or keynote event, fill in this form.
In addition to this newsletter, I recommend some other great ones. All free. Check them out here.
Friends,
Earlier this year, in a small one-on-one demonstration I ran, a London accountant clapped when ChatGPT drafted an audit note he couldn't explain, then fell silent. In that pause, you can hear three forces reshaping work: the Exponential Gap, the Human Agency Crisis, and the Reverse Singularity.
Recent expert surveys put a 50% chance of human-level AI around 2047. Timelines vary, but the shift has started. Depending on who you listen to, many think it’s within five years. Arguing timelines misses the point. Most people are sleepwalking through this new phase.
For those encountering the term AGI for the first time, Artificial general intelligence (AGI), sometimes called human‑level intelligence AI, is a type of artificial intelligence that would match or surpass human capabilities across virtually all cognitive tasks. We’re not there yet with GPT-5, but we’re going in that direction. (see recommended links below)
The Exponential Gap
Human intuition rises in gentle slopes while machine capability explodes in stair-steps. Training compute has grown about 4–5x per year since 2010. Doubling a penny daily fills a stadium by day 30. On day 29, it's half empty. That's how exponential feels until it's too late.
What unsettles me is that we're not just bad at predicting exponential change, we're wildly overconfident in our predictions. I've spent years writing about technology, yet I still catch myself underestimating the pace. I’m really feeling it now - and I don’t feel as if I have slowed down.
Today, copywriters, radiologists, and tax analysts face the same exponential blindness. But what astonishes me is: no one is preparing for their next career. People are updating their LinkedIn skills section whilst AI rewrites entire job categories.
If your five-year plan still fits on a spreadsheet, it's already obsolete.
Goldman Sachs estimates up to 300 million jobs worldwide, about one job in ten globally, could face automation pressure this decade. Advanced AI chips depend on a few hundred EUV scanners, all made by ASML. ASML, a Dutch company you've never heard of, makes the only machines capable of printing advanced AI chips, shipping roughly 60 per year against runaway global demand.
Speed alone won't kill you; speed without the steering wheel will.
The Human Agency Crisis
Agency means the capacity to set goals and act without unseen algorithmic nudging. My guess is that many knowledge workers already use AI daily. When algorithms increasingly guide our working decisions, we're not just losing control. We're surrendering parts of our identity.
The counterintuitive reality is that most people fighting AI displacement are targeting the wrong threat. The risk isn't robots taking jobs. It's humans becoming robotic.
Microsoft's 2024 Work Trend Index found 75% of knowledge workers use AI, often without management awareness. C-suites still treat GPT like Clippy on steroids, unaware it's quietly drafting their severance letters.
When Hitesh lost his role as a tax analyst overnight, he felt stripped of identity and purpose. Yet his redundancy opened a path to work with Alzheimer's patients, turning professional loss into deeply meaningful renewal. Hitesh hadn't been replaced. He'd been rehumanised. He became AI Displaced: a highly skilled professional seeking new purpose in an AI-dominant economy. He doesn’t make as much money, but he’s more secure in his role - and when I spoke to him as part of my research for my book, he seemed resigned to this new life.
The lucky ones will find new meaning. Others may struggle to find new purpose.
What terrifies me is that we are unknowingly creating a two-tier system. Those with an augmented mindset who learn to work fluidly with AI will thrive. Those doing non-replicable work (the therapist, the physiotherapist, the hairdresser) will survive. Everyone in between? They're edging towards redundancy.
Job displacement numbers tell a brutal story. The largest economic shocks in living memory, the 2008 financial crisis and COVID-19, together displaced around 136 million workers. AI automation could affect more than twice that number in a single decade.
Your own Agency Quick-Check:
Did an algorithm choose the last movie you watched?
When did you last ignore GPS?
When were you last bored on purpose?
Recently, I realised algorithms had quietly hijacked my own curiosity. Every other book I'd read in the past six months was a suggestion. I had become a passenger in my own intellectual journey. The creeping realisation felt like waking up in someone else's dream. For someone researching human agency, this was particularly galling and embarrassing. I tell you this because I am writing a book about it, and so anyone can fall into this trap. I even took pictures of the inflight entertainment on my holiday flight and asked Chat-GPT to order them for me rather than let curiosity do the work.
SuperSkills consistently outlast technological disruption, whatever the era: curiosity, empathetic communication and the augmented mindset are three of them.
In a BCG/HBS field trial, GPT-4 improved output quality by approximately 40%, speed by around 25%, and tasks completed by about 12% for suitable work. For context, typical year-on-year productivity growth in advanced economies hovers around 1–2%. AI is delivering an order-of-magnitude jump on specific tasks. That’s what having an augmented mindset is all about. The Augmented Mindset is the ability to work in a constant, fluent feedback loop with intelligent tools without letting them erode your originality, ethics, or sense of purpose. Think of it as cognitive “co-piloting”: you treat AI as an extension of your perception and reasoning, never as the pilot in command.
The uncomfortable truth? Most people aren't developing an augmented mindset or any of the age-tested SuperSkills. They're just hoping their jobs will somehow survive.
The Reverse Singularity
The Reverse Singularity is humanity's choice to set hard limits before AI outruns governance.
This is humanity's pivot from reactive panic to deliberate choice about our relationship with artificial intelligence. This curve is not destiny until we refuse to grab the wheel.
What I find most troubling is this: we're debating AI safety whilst becoming dependent on AI. The real question isn't whether AI will become too powerful. It's whether we'll become too weak to resist. And power is already concentrating in the hands of the few who control these systems.
I fear that mass displacement creates a world where the augmented few prosper whilst everyone else scrambles for the remaining human work, caring for bodies, fixing things with their hands, providing the emotional labour machines can't replicate.
This isn’t inevitable. Three actions can change the trajectory:
AI Fire Drill: Reclaim one AI-made decision each day for a week. Your proof point: small daily overrides reset attention and reveal where automation drives your defaults.
Build your Augmented Mindset: Treat your desk as a workbench, human craft plus AI power. Your proof point: guided human-AI pairing lifts quality and speed on suitable tasks.
Build Your Network: Gather AI Displaced peers and stay human-first. Your proof point: active peer groups spread tactics and create opportunity flow during role shifts.
By 2030, many corporations may rely more on algorithms than human workers. What's clear is that millions of AI Displaced will urgently need new purposes, new skills, and fresh ways to add value. The question isn't whether this displacement will happen. It's whether you'll be ready.
Choose Your Path
We're living through the moment that will determine whether we achieve the Reverse Singularity. Every choice between convenience and connection, efficiency and empathy, optimisation and meaning, decides whether we set limits or surrender control.
We remain the last generation that can freely choose humanity's place alongside AI. Your next deliberate choice decides whether AI is a tool or a master because today the choice isn't theoretical. It's personal.
Companies are obsessing over cost cuts. Employees are looking for the next training course to add a credential on LinkedIn. Meanwhile, entire industries are about to be rewritten by people thinking exponentially, whilst everyone else thinks linearly.
Don't be everyone else.
Algorithms optimise relentlessly. Humans must choose deliberately.
Stay Curious - and don’t forget to be amazing,
Here are my recommendations for this week:
One of the best tools to provide excellent reading and articles for your week is Refind. It’s a great tool for keeping ahead with “brain food” relevant to you and providing serendipity for some excellent articles that you may have missed. You can dip in and sign up for weekly, daily or something in between - what’s guaranteed is that the algorithm sends you only the best articles in your chosen area. It’s also free. Highly recommended. Sign up.
Now
“I remember the first time I logged onto Prestel” - this is my short note on the first time the internet felt alive to me. Old systems teach you what real novelty feels like. The takeaway is simple: train your gut to recognise that feeling again and you’ll cut through noise faster than any trend report. Do you remember when you had to type http:// and weren’t sure if it was forward or back slash and which one was which?
It’s obvious if you look hard enough - A quick lens I use for pattern spotting. Most signals sit in plain view once you decide what you’re actually tracking. The move is about being more Moneyball!
McKinsey and the consulting giants are reshaping around AI. This WSJ read shows delivery changing from slide volume to outcomes. The practical move is to productise one workflow this quarter and price it on throughput, not hours.
What screen time does to children's brains is more complicated than it seems - I’m back from holiday, but while there we tried to use our phones as a tool and not consume too much. At our hotel, a family next to us spent their whole breakfast on their phones watching video channels (TT,IG,YT etc). I know because I snooped as I got up to go to the buffet. They then walked off all while looking at their phones. At one point the dad told the kids to look at his video. This risks being communication for a new age.
I still love Google, but I am increasingly using perplexity - this guide to using Perplexity in your day-to-day is worth twenty minutes. If you lead teams, turn the best three prompts into a shared playbook and set a target to save five hours per person per week.
Next
OpenAI claims GPT-5 model boosts ChatGPT to 'PhD level' - GPT-5 is here - OpenAI has rolled out GPT-5. It’s sharper on code, stronger in reasoning, and better at that “pause and think” behaviour people have been waiting for. The API is live, so there’s no reason to wait for the hype to settle, run it against three real workloads this week, see where it beats your current setup, and lock a 90-day plan off the back of it. The copycat race and pricing shifts will follow fast. A massive step up for AI. I don’t like to use the word game changer, but this could be it.
DeepMind’s Genie 3 world model: A credible step toward agents that learn inside simulated worlds. This matters for robotics, logistics and edge devices.
Apple Answers - Apple’s “Answers” team is quietly building what sounds like a lean, privacy-first answer engine baked into iOS. Imagine asking a question and getting a grounded answer straight from your device, no browser tab, no search results page. If your audience lives on iPhones, start shaping your content for answer engines, concise, structured, and easy to cite, because the default behaviour of hundreds of millions of people is about to shift. Also Apple’s brain controlled iphone
The Rise of Silicon Valley’s Techno-Religion - Cade Metz takes us inside Lighthaven, block-long complex in Berkeley that’s become the spiritual and organisational hub for the Rationalist and Effective Altruism movements. These aren’t fringe gatherings anymore. Their ideas on AI risk and “maximising good” have already shaped the thinking (and sometimes the careers) of leaders at DeepMind, OpenAI, Anthropic and beyond. The piece captures the myth-making, the networking, the funding flows, and the very real influence this ideology now has over the direction of AI. If you want to understand where some of Silicon Valley’s biggest bets, and strangest moral calculations, are coming from, start here.
Silicon Valley Wants to Optimise Your Children’s Genes - This long, sparky exchange between Ross Douthat and Noor Siddiqui, founder of Orchid, is part science lesson, part moral sparring match. Orchid offers whole-genome sequencing of embryos before implantation, promising parents the ability to choose the “healthiest” child based on risks for more than a thousand conditions. The conversation moves from technical validation to the cultural and ethical fault lines: who gets access, what happens to unused embryos, and whether this severs one of the oldest links in human life, sex and procreation. If you want a single read that captures the collision of frontier biotech, personal choice, and societal unease, this is it. Related: Chinese AI Researchers Just Put a Monkey's Brain on a Computer
Free your newsletters from the inbox: Meco is a distraction-free space for reading newsletters outside the inbox. The app has features designed to supercharge your learnings from your favourite writers. Become a more productive reader and cut out the noise with Meco - try the app today
If you enjoyed this edition of Box of Amazing, please forward this to a friend. If you quote me, or use anything from this newsletter, please link back to it.
If you are interested in advertising in this newsletter, please see this link