The Architecture of Drift

How humans designed the machine that now designs you

Mar 15, 2026

Now accepting speeches, keynotes and assignments for 2026

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

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

Friends,

A quick note. My recent essays have been longer than usual including last week’s Flake 99 and the Agents are here. I know. Over the coming months, as my book SuperSkills moves toward its July publication, I want to unpack some of the ideas that underlie it. Not summaries of what I have in the book, but the actual thinking I did to get there. The stuff that kept me up at night while writing it. This is one of those ideas. If it hits, the book will be a good read for you. If it doesn’t, buy me a decaf skinny mocha with sprinkles on it and tell me where I lost you. And if you really want to understand drift, here’s an explainer video for you that I put together (with the help of notebooklm.)

I got snookered the other day.

For anyone outside the UK, snooker is a sport considered by some to be a sport, although arguably it’s one of the least sporty of sports, along with darts. This is not a post about snooker. If you’re really interested in the rules, please look it up on YouTube. But the concept matters. You use a cue to hit a white ball onto a coloured ball. And when you can’t see the ball you’re supposed to hit, when something is blocking the direct path and you have to bounce off the cushions just to get close, you are snookered. In life, it means you can’t do what you want to do. And usually, it’s your own fault you ended up in that position.

So. I was dropping my daughter to a friend’s place. I knew the area. I’d been there before. Not the exact house, but the neighbourhood, the general direction, the kind of route I’d take. Ten years ago I would have driven there on memory and instinct, maybe checked the A-Z for the last mile or two. Instead I did what we all do. I typed the postcode into the sat nav and followed the blue line like a good boy.

We drove straight into a road closure. A road I knew was closed. I’d seen the signs for weeks. I’d driven past the barriers twice that same month. And yet, because the sat nav told me to go that way, I went that way. I followed the screen instead of using my brain. My own brain. The one that had the information already in it.

The only alternative was to cut through what I call the fields. Single-lane country tracks where you have to pull into a verge every thirty seconds to give way to oncoming traffic. I hate driving through the fields. But I had no choice, because I’d been snookered. This was my only route unless we were going to be even later than we already were.

We were twenty minutes late. My daughter said, “It’s okay, Papa.” But I could see it. That look. Not anger. Disappointment. It was like the scene from Interstellar where a future version of Matthew McConaughey is screaming at the past version of himself not to do it, reaching through time and gravity to stop a mistake that’s already been made. That’s a bit dramatic for a Sunday morning in the sticks. I know. But I hate disappointment in other people’s faces. Especially my family.

And I sat there, engine idling, feeling profoundly stupid. The kind of stupid where you want to take off your own shoe and hit yourself in the head with it. I was thinking: this is it. This is drift. Not in the abstract. Not as a concept for a book. This. Right here. Me, a grown-ass man who writes about human agency for a living, blindly following an algorithm past information I already possessed, because the screen felt more authoritative than my own memory. And my daughter paid the price for it. Twenty minutes of her afternoon. A small thing. But small things are where drift lives.

I asked the dad for directions home. The first thing he said was: “Make sure you stay away from that road. It’s going to be closed for the next three months. There’s loads of traffic in that area.” Then he pointed at himself and said: “Trust the dad. Don’t trust the sat nav, Mr AI guy.”

Translated: trust the human. Don’t trust the algorithm.

The journey back took half the time. I didn’t get snookered. I didn’t have to drive through the fields. I wasn’t late. It felt like sinking a long ball into the top left pocket in the final of a snooker match. Which some people find exciting. I don’t. But I’ll take the metaphor. Rahim “Hurricane” Hirji. For those of you in the know.

The sat nav didn’t fail. I failed. I outsourced my judgement to a system that didn’t know what I knew. And I did it automatically, without thinking, because that’s what drift is. It’s not dramatic. It’s not a crisis. It’s twenty minutes late to a friend’s house and a detour through fields you hate because you trusted a screen more than yourself.

If that was an isolated incident, I’d laugh it off. It isn’t. It’s happening to all of us, all day long, in ways we’ve stopped noticing. The concept of drift is one of the key ideas in my book, and it’s the foundation upon which I build a new set of skills, SuperSkills, that help you see it, step out of it, and start designing instead. But more on that later. First, let me show you how deep this goes.

You’re in a restaurant. A proper one. Not fancy, but the kind of place where the menu has weight to it and the lighting does everyone a small favour. Maybe a local Italian. Maybe a slightly upmarket Lebanese place where you know the general dishes but not the specialities.

You open the wine list. Forty options. You don’t know most of them. But there, fourth or fifth down, is a Sauvignon Blanc you recognise. It’s not the cheapest. You’d never order the cheapest, because that feels like a confession. It’s not the most expensive either. It sits in a comfortable middle ground. You point to it. You feel like you chose.

You didn’t.

Someone chose for you. Weeks before you walked in. Probably while you were watching Netflix.

There’s a £50 bottle near the top of the list. It’s not there to sell. It’s there to make the £30 bottle feel reasonable. Behavioural economists call this anchoring. Kahneman and Tversky documented it in the 1970s. The first high number you see rewires your sense of what’s normal. By the time your eyes reach the £30 bottle, your brain has already decided it’s a sensible choice. The £50 bottle gave you permission. And your Sauvignon Blanc? You reached for it because you recognised the name. That’s not taste. That’s familiarity wearing the costume of preference.

The menu engineer knows this. (That’s a real job, by the way. A whole career built around making you spend more at dinner.) They know your eyes go to the middle of the page first, then the top right, then the top left. Menu designers call those three spots the Golden Triangle. That’s where the highest-margin dishes live. The currency sign has been stripped from the prices because a Cornell study by Yang, Kimes and Sessarego found that seeing a £ symbol reminds you that you’re spending money, and that reminder makes you spend less. About 8% less, to be precise. So they just... take it off. And you don’t notice. The descriptions say “line-caught” and “slow-braised” because adjectives inflate what you think the food is worth. And here’s one that genuinely annoyed me when I first read it: one study served identical food from two buffets at different prices. The more expensive buffet was rated tastier. Same food. Same chef. Same kitchen. Different number. Different experience in the mouth. We are that easily played.

Then there’s the bread. It arrived before you ordered, warm and generous. You thought it was hospitality. A gesture. And maybe it was. But bread is also refined carbohydrate, and refined carbohydrate spikes your blood sugar, which can make you hungrier than you were before you started eating. The salt makes you thirsty. Thirsty people order drinks. Drinks carry the highest margins on the menu. And now many restaurants charge £3 or £4 for the bread, which shifts the psychology again. You’ve paid for it. You’re invested. You’re not leaving early. The bread is not just bread. It is the opening move in a game you didn’t know you were playing. And honestly? It’s a brilliant move. I’m not even angry about it. I’m impressed.

And this is what makes drift so hard to fight. It doesn’t feel like a trap. It feels like a gift. The bread feels like generosity. The Netflix autoplay feels like convenience. The AI-suggested reply feels like efficiency. The personalised playlist feels like someone who understands you. Drift, when it’s done well, feels like freedom. That’s the trick. That’s the whole trick. You’re not being imprisoned. You’re being released from the burden of having to choose. And it feels wonderful. Right up until you realise that freedom without choice isn’t freedom at all. It’s just a comfortable surrender.

And the soup of the day? I grew up thinking that was the chef’s creative inspiration, a daily experiment born of passion. It isn’t. The soup of the day was invented because restaurants needed to use up ingredients that were about to go off. Historically, the specials board is the leftovers board. Dressed up. Renamed. Sold to you as a curated experience. Gordon Ramsay has said he always asks what yesterday’s soup was. If it’s the same as today’s, you know the game. I’ve started doing this too. It’s incredible how often the answer is the same.

I want to sit with this for a moment. Because the soup of the day is drift in its purest form. You were told it was a recommendation. It was actually a business decision. You thought you were being offered something special. You were being offered something that needed to be sold before it went in the bin. And you bought it, happily, because it felt like a choice.

Here’s something I do whenever I eat out. I ask the waiter two questions. First: what are your three most popular dishes? Second: what are your three favourite dishes? Sometimes I’ll push it further: what would your mum or dad order if they came in tonight?

Sometimes they give you the same answer both times. Which tells you something. It tells you they’re on autopilot. They’re reciting the menu. They’ve drifted into the script the restaurant gave them and stopped having an opinion.

But sometimes, and this is the moment that matters, their face changes. They lean in. “Honestly? The chicken livers. They do something with sherry and thyme and it shouldn’t work but it does.” They become a person. Not a waiter. A person with taste, with experience, with a view. That shift, from default to design, from script to conviction, is the entire argument of this essay compressed into thirty seconds at a table.

I live for that moment. It tells me everything I need to know about a restaurant. And honestly, about a person. Because the same thing happens in meetings, in interviews, in conversations at parties. You can hear when someone is reciting and when they’re thinking. The reciter gives you the safe answer. The thinker gives you *their* answer. And their answer, even when it’s wrong, is always more interesting.

The person who repeats the most popular dishes is drifting. The person who tells you about the chicken livers is designing. They’ve thought about it. They’ve tasted it. They’ve formed a position. And they’re sharing it with you because they believe in it, not because it’s on the briefing sheet.

Most of us, most of the time, are giving the first answer. Not just in restaurants. In our careers. In our opinions. In our lives.

None of this is new. The menu, the bread, the soup, the shelf placement, the eye-level product in the supermarket. Humans have been designing the environments in which other humans make choices for centuries.

Kids’ cereals with cartoon mascots sit at the eye level of a child in a trolley, not an adult. Sweets live at the checkout. Milk and bread go at the back of the store, so you walk past everything else to reach the only thing you came for. Casinos have no clocks and no windows. Ikea walks you through every department on a one-way path to reach the one thing you actually need. The shortcut signs are there, but the default is the full loop. And defaults win.

In every one of these examples, a human sat down and designed the architecture of someone else’s choices. And the design serves the architect’s interests. Not yours.

But it was tolerable. It was human-scale. The restaurant owner nudges two hundred people a night. The supermarket layout gets redesigned once a year. And the architecture was at least visible. You could see the menu. You could notice the shelf placement. You could catch the tricks if you were paying attention.

Then we handed the playbook to the machines.

Walk into a McDonald’s in 2026. Not the one from your childhood, with a teenager behind the counter asking if you want fries with that. The one with a six-foot touchscreen that has already started making decisions for you before you’ve touched it.

The screen shows you a photograph of a burger so perfectly lit it could hang in a gallery. The “YES” button for adding a drink is large and bright and positioned exactly where your thumb wants to go. The “No thanks” option is smaller, greyer, tucked to the side. When you finish selecting, it asks if you’d like to make it a meal. That upsell converts over 40% of the time when the timing is right. Kiosks raised McDonald’s average order value by 20 to 30 percent. Not because people wanted more food. Because the machine never forgets to ask, never gets tired of asking, and never makes you feel judged for saying yes. Think about that for a second. A teenager behind a till might skip the upsell because they’re having a bad day or because the queue is long or because they’re a human being who gets bored. The kiosk has no bad days. The kiosk has no conscience. The kiosk just asks. Every single time.

McDonald’s paid $300 million for an AI company called Dynamic Yield. Its technology changes what the screen shows you based on the weather, the time of day, how busy the restaurant is, and what you’ve already put in your basket. Cold afternoon, it suggests coffee. Hot day, ice cream. It learns what makes people spend and then it does more of it.

This is the bridge. The restaurant owner’s instinct, the one that said put the high-margin item where the eye falls and remove friction from the spending decision, got encoded into software. The logic is identical. The scale is not.

The restaurant owner nudged two hundred people a night. The algorithm nudges two hundred million. The menu got redesigned once a season. The algorithm adjusts in real time. And the menu was visible. You could hold it, study it, argue with it. The algorithm is invisible. You never see the architecture. You only see the choice it already made for you.

That’s the shift. Humans built the system. The system now shapes the behaviour. And the human inside the behaviour no longer realises they didn’t choose it.

That’s drift. And if you’re feeling uncomfortable right now, good. That discomfort is the feeling of waking up.

I need to say something clearly here, because this is where people sometimes misread me. I believe AI is a superpower. I believe it is the most significant amplifier of human capability we have ever built. I use it. I build with it. I think about it every single day.

This is not an anti-technology argument. This is a pro-agency argument. The question is not whether you use AI. The question is whether you are using it, or whether it is using you. And that question applies just as fiercely to Netflix and Spotify and Amazon and LinkedIn and every delivery app on your phone.

Because behind every one of those products, there are people. Real people. With job titles and performance reviews and a set of metrics they are paid to move.

Product managers design the flows you move through. Growth engineers optimise the sequence of screens between your impulse and your purchase. Retention specialists build the nudges that bring you back after you’ve been gone for a day or a week. Behavioural designers study the psychology of how you make decisions and apply it to button colour, notification timing, and default settings. Engagement analysts track every tap, every scroll, every hesitation, and feed that data back into the system so it can get better at keeping you.

These people are not villains. I want to be emphatic about that. I’ve worked in technology for over twenty years. I’ve personally been in projects and OKRs where engagement was the sole and primary goal of that quarter. Sometimes I’ve been the one directing it. The people are talented, thoughtful, and solving genuinely hard problems. But the problems they are solving are not your problems. They are solving their company’s problems. And their company’s problem, stripped of the mission statements and the brand values and the quarterly earnings calls, comes down to one word: attrition.

The user who cancels. The subscriber who unsubscribes. The person who closes the app and doesn’t come back. Attrition is the thing every digital business fears. And so every digital business, whether it’s a streaming service or an AI company or an ecommerce platform, hires brilliant people to prevent it.

They make it one click to subscribe and five clicks to cancel. They auto-renew your subscription and bet on your inertia. They make the “Accept All Cookies” button large and bright and the “Manage Preferences” button small and grey and buried three screens deep. They design the experience so well, or make you so dependent on it, that leaving feels harder than staying.

These are not accidents. They are not bugs. They are features. They are the product of skilled people solving the problem they were hired to solve. And the problem they were hired to solve is how to keep you inside the system.

Which means the problem, at the end of the chain, is you. Your attention. Your time. Your money. Your inability to close the tab. That’s what they’re optimising for. That’s what a room full of smart people with whiteboards and dashboards and PhDs in behavioural psychology are working on right now, at this very moment, while you read this sentence.

And here’s the thing I keep coming back to. The algorithm doesn’t want you to be unhappy. An unhappy user leaves. The algorithm doesn’t even want you to be satisfied. A satisfied user is done. The algorithm wants you to be *undecided*. Hovering. Scrolling but not selecting. Watching but not deciding to watch. Browsing but not committing. That state of almost-choosing, that permanent gentle indecision, is where engagement lives. The moment you make a firm decision, you might decide to leave. So the system is designed to keep you interested enough to stay but never so resolved that you stop. Earlier I said drift feels like freedom. This is why. The freedom you’re feeling is engineered indecision. And engineered indecision is the most profitable psychological state ever invented.

Once you see this, you see it everywhere. Here is what drift looks like in a normal day. See how many you recognise before you start getting annoyed with me.

Your morning. You open your phone and immediately react to whatever’s at the top of your inbox. You didn’t rank those emails by importance. Google did, based on who you usually reply to and how fast. The email from the old friend who writes once a year is buried. An algorithm decided whose voice matters most in your morning. And you let it. Every single day.

Your route. You take the first suggestion Google Maps offers. Every time. You’ve lived in your city for years and there are whole neighbourhoods you’ve never driven through because the algorithm never sent you there. You’ve lost the ability to be lost. And with it, the ability to discover. When was the last time you took a wrong turn on purpose?

Your evenings. Netflix autoplays the next episode before you’ve decided if you want to watch it. That five-second countdown is a masterclass in design. It exists to prevent a moment of conscious choice. You didn’t decide to watch four hours of television. You decided to watch one episode and then failed to decide to stop. There’s a profound difference between those two things, and Netflix knows it.

Your music. Spotify’s Discover Weekly feels like a gift. Someone made you a playlist. But it’s a feedback loop that narrows your taste while making you feel like it’s expanding it. You will never hear the equivalent of a friend saying “listen to this, you’ll hate it at first, but trust me.” The algorithm doesn’t have friends. It has data. And data doesn’t take risks.

Your friendships. You respond to the people who message you, not the people you miss. Your phone surfaces whoever texts most often. The friends who don’t text much slowly fade. Not because you chose to lose them. Because the algorithm deprioritised them and you didn’t notice. I think about this one a lot. Some of the most important people in my life are terrible texters. If I let my phone decide who matters, I’d lose them.

Your judgement. You check reviews before choosing a restaurant. You google symptoms before listening to your body. You ask ChatGPT before forming your own opinion. The sequence has flipped. It used to be experience first, then validation. Now it’s validation first, then experience. We’ve become a civilisation that asks permission from strangers before trusting ourselves.

Your consent. The “Accept All” button is large and requires one click. “Manage Preferences” requires five clicks through nested menus. That is not neutral design. That is manufactured consent. And almost nobody clicks through to manage. They know that. They designed it that way.

I should be fair to routine. Making your coffee the same way every morning is not drift. Taking the same walking route because you love it is not drift. Having a favourite restaurant and always ordering the same thing is not drift. Those are preferences. They’re yours.

The difference is one question. Whose autopilot are you on?

When you chose the routine, consciously, because it serves your life, that’s design. When someone else designed the routine and you fell into it without noticing, that’s drift. And when that someone else is a system built to serve their interests while you think it’s serving yours, that is algorithmic drift.

In my book, I map this as a matrix. Two axes. On one side: are you drifting, or are you designing? On the other: how aware are you of which one you’re doing?

That gives you four quadrants. The Designers: high awareness, actively building their own path. The Programmed: high capability but no awareness that they’re on someone else’s rails. The Stuck: they know something’s wrong but can’t move. And The Sleepwalkers: low awareness, full drift. Going through the motions of a life that was designed for them by someone else’s algorithm.

Most people, when they’re honest, sit somewhere in the bottom half.

The goal is not to be in the top right quadrant at all times. That’s exhausting and probably impossible. The goal is to know which quadrant you’re in. Because once you can see it, you can move.

So. Two steps. In that order.

Step one: see it. You can’t redesign what you can’t see. The first move is not action. It is attention. Noticing that you took the first route. Noticing that you watched three episodes, not one. Noticing that you haven’t called that friend in four months. Noticing that your entire morning was spent responding to other people’s priorities.

This is not about guilt. It is about opening your eyes. The sleepwalker’s first job is to wake up.

Step two: add friction. Not efficiency. Friction. The opposite of what every platform wants from you. You pause the autoplay. You take the second route. You close the delivery app and open the fridge. You write the email yourself instead of letting the AI suggest one. You form an opinion before checking what everyone else thinks.

Every product you use has been engineered to remove friction. Because friction is the moment where you might make a conscious choice. And a conscious choice is where their control over you ends.

Design is the act of putting friction back where it matters.

Here is where I want to leave you. Not with theory. With questions. Five of them. The kind you can ask yourself today, tomorrow, on Monday morning when the inbox is screaming and the phone is buzzing and the algorithm is already two steps ahead of you.

One. What is the first thing you did this morning, and did you choose it, or did your phone choose it for you?

Two. What opinion did you form this week that wasn’t influenced by what you saw on a screen?

Three. When was the last time you did something for the first time?

Four. If you turned off every recommendation engine for a month, every algorithm, every suggested-for-you feed, what would you actually choose to read, watch, eat, and listen to?

Five. Who is the last person you contacted not because they messaged you, but because you thought of them?

If any of those questions made you pause, good. That pause is visibility. That pause is the first step from drift to design.

The sat nav moment was awkward fro me. Not because it mattered. Nobody died. My daughter said it was okay, and she meant it. But I can still see her face. That disappointment. And I know it was there because of me. Not because of the algorithm. Because of me choosing not to think. ME. NOT. CHOOSING. TO. THINK!

A dad in a doorway gave me better directions in ten seconds than the algorithm managed in twenty minutes, because he knew the roads, he’d driven them that morning, and he used his brain.

Trust the dad. Don’t trust the sat nav, Mr AI guy. He didn’t use a better algorithm. He used his brain. That’s the whole point. It’s not bad advice for a Sunday morning. It’s not bad advice for a life.

The second step, the skills you need to stay awake, to make design a practice and not just a fleeting moment of awareness, is what the book is about. SuperSkills is not an argument against AI or technology. I think AI is extraordinary. I think it changes everything. But I also think it is the most sophisticated drift machine ever created, and the skills required to use it without being used by it are the most important skills any of us will ever learn. Those are the seven skills in the book. And if you’ve read this far, you already know why they matter.

SuperSkills: The Seven Human Skills for the Age of AI is published by Kogan Page on 3 July 2026.

If this piece made you pause, send it to someone who needs to see the architecture. The best way to wake a sleepwalker is to show them where they’re walking.

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

This new emoji is all of us in 2026 - Get ready to see this bug-eyed little guy everywhere.
Wired headphone sales are exploding. What’s with the Bluetooth backlash?
Robots, cameras, and lots of data about garbage: Inside the recycling industry’s new bet. Like just about every other industry right now, recycling is betting that AI can change the calculus
How Gambling Ate the World In less than a decade, betting apps swallowed sports. And now they’re doing the same to the news
The yoghurt delivery women combatting loneliness in Japan

Next