Here’s an enjoyable piece of light corporate subversion. Buy a copy of Alex Pang's book, Rest. Leave it visible on your desk. Say nothing. Let the tagline — “Why you get more done when you work less” — do the talking for you. (If you're a founder, you can still troll yourself by doing the same.)

I discovered Pang five years ago, when an interest in AI still overlapped heavily with a large Asimov collection. Back in 2021, my post-pandemic burnout was colliding with an undiagnosed chronic condition to tank my energy levels to record lows. My window for happily getting hard things done had shrunk to two hours a day. The rest of the time I was wading through treacle.

Into this mess, Pang's argument felt like gospel. He writes that four hours a day of doing hard things is not a personal failure, but rather the natural default for the human brain. Our greatest thinkers prove time and again that the real way to make history is a few concentrated hours of deep work followed by a little light letter-writing and a nice walk in the woods. 

He also sheds new light on the 10,000 hours theory: the idea, popularised by Malcolm Gladwell, that you have to put in 10,000 hours of work to achieve greatness. The Beatles were said to have spent 10,000 hours performing in Hamburg clubs before breaking out as international superstars. Kasparov exceeded this threshold before becoming the youngest undisputed chess champion at 22. But Pang goes back to the original study, a 1993 survey of violin students. He discovers some key rest-related details that didn’t make it into the pop-culture statistic — like the fact that the top performers slept more than their mediocre peers. The best violinists also estimated their leisure time far more accurately than the average players.

Pang concludes that the deliberateness of both practice and rest is key here: “We’ve come to believe that world-class performance comes after 10,000 hours of practice. But that’s wrong. It comes after 10,000 hours of deliberate practice, 12,500 hours of deliberate rest, and 30,000 hours of sleep.”

I could end here. But in 2026, there’s a problem with the gospel of Pang. AI enables anyone to act like an expert after 30 minutes of prompting. No one cares about your 10,000 hours any more. And Kasparov lost to Deep Blue back in ‘97. The result? Admitting to human frailty hasn’t felt this foolhardy for a generation. Like, if you confess to an off day, Claude will be doing your job before your sniffles have even dried up.

This reaction is as counterproductive as it is instinctive. No one knows what happens next. But I am sure about this: you and I will not outrun the machine. Not with our temperamental bodies, inefficient social ties and tendency to get snacky around 3pm. You can’t run faster than a car. You will sometimes need a sick day. 

Perhaps, instead of viewing our need for rest as an embarrassing deficit in the machine age, we should view it as a strength — a uniquely human quality which AI is actually struggling to imitate.

Jack Clark, co-founder of Anthropic, is a surprisingly contrarian voice. On Derek Thompson’s Plain English podcast, he’s deep in discussion around the 50 minute mark about why AI is still missing that spark of creativity and intuition that makes humans stand out. And then he begins to sound an awful lot like Pang: “I don't think we know how to get these machines to stop working and idle… There's some essential property here of being present in the world and not working, but thinking and interacting with the world. My guess is some aspects of creativity comes from this very subtle thing that is very special about us.”

It comes as no surprise, then, that a new wave of AI labs are working on an array of alternatives to LLMs. AMI Labs are building systems that process real world data: “Real intelligence does not start in language. It starts in the world.” Ineffable Labs hope to create a ‘superlearner’ that discovers all human knowledge for itself without a shred of data being fed into it. Recursive aim to design endlessly self-improving systems, while Thinking Machines are creating AI that specialises in interactive workflows. 

They’re all fascinating answers to the same question: what does it mean to think? But none seem to have taken the idea of idleness as a core. Nor is it even clear what a machine would do with such a thing. Perhaps Clark is right and letting our brains idle is a uniquely human contribution.

If so, that’s at once reassuring and inconvenient. Yet again, the AI revolution has a knack for bringing into the light the awkward questions: When did you or I last let ourselves be a little bored? Why do I, for one, insist on wearing headphones to water the plants, even though I love listening to the evening birdsong?

A quiet rebellion against filling our brains with pings and ‘to do’s for every second of the day is growing. Ironically and inevitably, it has spawned a collection of internet trends: nothing maxxing, rawdogging boredom, the do nothing challenge.

This is our edge, and it’s uncomfortable. It’s time to linger in the uneasy space between one packet of distraction and the next. Who knows, we might even learn to like it here. I can’t say for sure that the ability to touch grass will survive as a human advantage, but it might be our best shot. And in the meantime you touched some grass. No one can take that away.

‍