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,
In his 2024 commencement speech at Dartmouth, Roger Federer shared a statistic that startled the audience.
Over his career, he won nearly 80% of his matches. But he won only just over half the points. One of the greatest tennis players in history lost almost half of every game he ever played. His lesson wasn’t about tennis. It was about what mastery actually requires: “When you’re playing a point, it has to be the most important thing in the world. But when it’s behind you, it’s behind you.”
Champions aren’t people who avoid failure. They’re people who’ve failed so many times they’ve learned how to metabolise it. They’ve built the reflexes, the recovery, the resilience. Not by winning, but by losing, over and over, and coming back anyway.
I keep thinking about this in the context of AI.
The myth of effortless competence
You know those people who seem to effortlessly handle whatever gets thrown at them? I’m not one of them! They are the ones who appear to improvise their way through complex situations, making it look easy. You know them right?
They’re not winging it. They’re drawing on thousands of hours of feeling the pain.
Every confident improvisation sits atop a mountain of prior failure. The executive who makes a good call under pressure has made a hundred bad calls and learned from the wreckage. The writer who produces clean prose on deadline has written countless terrible drafts and discovered, slowly, what works. The negotiator who reads the room accurately has misread dozens of rooms and paid the price.
What looks like intuition is actually pattern recognition, and pattern recognition requires patterns. Which means reps. Which means getting it wrong, a lot, for a long time.
This is the part that AI lets you skip.
The drift you don’t notice
Yesterday, you could do mental arithmetic without thinking about it. Now you’re pulling out your phone calculator to work out what a 0.25 percentage point rate change means for your mortgage.
Last year, you drafted emails from scratch. Now you’re “polishing” AI suggestions, which increasingly means accepting them with minor tweaks.
A few years ago, you’d sit with a problem, feel stuck, push through the discomfort, and eventually find your way to a solution that was genuinely yours. Now, the moment you feel resistance, there’s a frictionless path available: ask Claude, get an answer, move on. Maybe compare that answer with what ChatGPT thinks. Then move on.
Each individual instance is trivial. The calculator saves thirty seconds. The email assistant handles the boring stuff. The AI consultation gets you unstuck faster.
I notice it in myself. I can feel which parts of my thinking I’ve started to outsource.
But the accumulated effect is something else. You’re not just saving time. You’re skipping reps. And the reps were where the learning happened.
What’s happening isn’t laziness. It’s something more subtle. I think of it as capability debt: every skipped rep saves time now, and costs judgement later.
This is what I call drift: the gradual, almost imperceptible outsourcing of cognitive effort until one day you realise you’re not sure what you actually know versus what you know how to look up, or ask an AI.
The opposite of drift is design: making deliberate choices about when to struggle and when to shortcut, based on a clear-eyed view of what each choice costs you.
What you lose when you skip the struggle
Here’s a finding that’s become famous in neuroscience. I’ve used it before.
London’s black cab drivers must pass “The Knowledge”, an exam requiring them to memorise 25,000 streets and thousands of points of interest. It takes three to four years of intensive study. When researchers scanned these drivers’ brains, they found their hippocampi, the brain region responsible for spatial memory, were measurably larger than average. The original study by Woollett and Maguire tracked drivers over time and showed the hippocampus actually grew as they acquired The Knowledge.
Their brains had physically adapted to the demand.
Now consider the opposite trajectory. A 2020 study in Scientific Reports tracked GPS usage and spatial memory over time. Heavy GPS users showed worse navigation ability, and their spatial memory declined faster than those who navigated without assistance. The GPS wasn’t just a convenience. It was weakening a capability.
That’s what makes this dangerous. The loss happens quietly, while everything still appears to be working.
Use it or lose it isn’t a metaphor. It’s neuroscience.
The risk isn’t that we’ll become bad at our jobs. It’s that we’ll become bad at becoming good. Every time you let AI handle something you could have struggled with, you skip a cognitive rep. One skipped rep doesn’t matter. A thousand skipped reps starts to show. Ten thousand, and you’ve become a different kind of professional. One who’s optimised for efficiency but hollowed out something harder to name.
The strategic thinking problem
Here’s where this gets serious for leaders.
Strategic thinking, the ability to see patterns, weigh trade-offs, make judgement calls under uncertainty, doesn’t come from reading about strategy. It comes from years of working in the details and slowly developing the instinct for what matters.
The executive who can look at a complex situation and see the essential shape of it? They built that capability by being deep in the weeds, repeatedly, across many contexts. The pattern recognition that enables big-picture thinking is assembled from thousands of granular experiences: deals that fell apart and why, projects that succeeded unexpectedly, decisions that looked right but weren’t.
You can’t shortcut your way to strategic judgement. You have to earn it through exposure, through feeling the texture of problems, not just their summaries.
There’s another problem people don’t like to name. Some of this capability is time-bound. The easiest period to build judgement is when the cost of being wrong is low and the opportunity to repeat mistakes is high. If those years are spent optimising for polish instead of exposure, the window doesn’t reopen cleanly later.
AI is very good at summaries. It can give you the shape of a situation, the key considerations, even a recommended course of action. What it can’t give you is the hard-won instinct for when the summary is missing something, or when the recommendation is subtly wrong, or when the real issue is the one nobody’s named yet. That instinct is built on encounters with difficulty. You only develop it by getting things wrong and learning why.
If junior professionals spend their formative years having AI handle the detail work, they may never develop the foundation that strategic thinking requires. They’ll arrive at senior roles with polished outputs and shallow roots. They’ll be fluent in the language of strategy without having earned the intuition for it.
And eventually, they won’t know the difference.
Feel the pain first
So what do you actually do with this?
The principle is simple: feel the pain first, then automate intelligently.
Before you hand something to AI, make sure you’ve struggled with it enough to actually understand what you’re handing over. Not because suffering is virtuous, but because the struggle is where capability gets built.
One rule I’m experimenting with: I don’t ask AI for help on anything important until I’ve written my own answer down first. Even if it’s wrong. Even if it’s rough. The act of committing to a position before asking for assistance is what keeps me in the game. Use old school pen and paper, even if your handwriting is worse than mine.
The second principle is just as important: be honest about who made the decision.
Did you form a view and then stress-test it with AI? Or did you outsource t2he thinking and then adopt whatever came back?
These aren’t the same thing. And over time, the difference compounds.
What will matter when efficiency is free
The AI age will reward people who can actually think, not just people who are efficient.
Efficiency is easy now. Anyone can produce polished outputs quickly. That’s table stakes. The scarce thing, increasingly, will be the capacity for independent judgement: the ability to know when the AI is wrong, when the situation is unprecedented, when the real problem is the one no one’s asking about.
That capacity isn’t built by optimising for convenience. It’s built by struggling, failing, learning, and doing it again. By spending years being bad at things until you’re good at them.
Federer didn’t win because he avoided losing. He won because he learned how to recover faster than everyone else.
The reps you skip are the reps that would have made you.
Stay Curious - and don’t forget to be amazing,
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.
Tools I Use:
Jamie: AI Note taker without a bot
Refind: AI-curated Brain Food delivered daily
Meco: Newsletter reader outside your inbox
Prompt Cowboy: Prompt Generator
Manus: Best AI agent to do things for you
Recommend Reading:
Now
This is what The Brookings Institution found when they asked 500 people across 50 countries about AI in education - one of the most comprehensive global assessments of AI’s risks & benefits in children’s education based on first-hand accounts from 500+ teachers, students, parents and experts across 50 countries, positioned as a global early warning system for AI in education.
Thematic Predications. One of my buddies made my curation of the best lists of the best reports of the best predictions of 2026 super useful. Check out the predictions by themes. It reminds me a bit of Inception - a dream within a dream within a dream...
I saw the future of retail, and it’s all AI - From window shopping and browsing to reviews and recommendations, retailers and tech companies envision a future filled with artificial intelligence, whether shoppers want it or not.
CES 2026 – “The Future Is Here” - CES is a trade show and not a broad consumer event. It just so happens that it attracts so many in the industry that it has often been covered as a broad interest news event.
Welcome Back to the Office. You Won’t Get Anything Done - Return to office mandates aren’t about output. They’re about asserting control
Next
2026: This is AGI - Sequoia is calling it. Kinda
How to write a good spec for AI agents - Aim for a clear spec covering just enough nuance (this may include structure, style, testing, boundaries) to guide the AI without overwhelming it. Break large tasks into smaller ones vs. keeping everything in one large prompt. Plan first in read-only mode, then execute and iterate continuously.
The AI revolution is here. Will the economy survive the transition? - a debate
The Mythology Of Conscious AI - Why consciousness is more likely a property of life than of computation and why creating conscious, or even conscious-seeming AI, is a bad idea.
When Will They Take Our Jobs? - And once they take our jobs, will we be able to find new ones? Will AI take those too?
If you enjoyed this edition of Box of Amazing, please forward it to a friend or share the link on WhatsApp or LinkedIn.
A particularly good essay, Nahim - and one that makes me anticipate your upcoming book. As I read this article, I kept thinking, "this 'drift' vs 'design' concept would make a great book", and now I see that's exactly (part) of what you'll be addressing in the book. Good!
I think being bad at things has become mostly invisible today which is why we hardly notice that AI has made it all the more invisible. Everybody on TV skips the parts where they suck at stuff and Rocky just montages his way to awesomeness. But, to your point, the part where you suck is the part where you are also getting good. I work with so many people who can create strategies but are not strategic, proposal writers with nothing worth proposing. Unfortunately, we also have very little tolerance for suckiness in others since we have this distorted belief that people should be innately good at things like math and playing instruments. I am with you. For everyone's sake, let your suck flag fly and learn to play violin in public. Otherwise, people might stop learning to play violin at all.