You ask why we should believe in a tomorrow that has nothing to do with today while citing the percentage of people in agriculture which looks nothing like yesterday. I'd rather see an object-level argument about what exactly humans will be doing, once AI can perform all the mental and physical labor at an above human level.
So you’ve now defined a third camp in the AI debate: there are the doomers, there are the utopians (or is it effective altruists?) … but what do we label your position? (It’s one I share, by the way.)
Adam can answer for himself, but I see 4 main camps:
1. Doomers who think AI will kill us all and/or take all our jobs. This group includes many members of the Effective Altruist cult.
2. Boomers who think AI is wonderful magic, will cure cancer within 10 years, and so on.
3. Skeptics who think AI is a scam run by billionaire grifters.
4. Realists who think AI is a powerful and useful tool which will be of great benefit, but the smartest AI in the world won't eliminate the countless bottlenecks in the world of atoms, organisms, and culture.
(Obviously given how I have labeled these camps, I put myself in the fourth one.)
Sadly as a freelance creative ai has already, in very short order, replaced most of the things I’ve been hired to do over my 25 year career. Repercussions 20 years in the future I can’t imagine, but my miserable books for 2025-2026 are far from theoretical. Creatives who’ve been making logos and branding packages and producing videos and podcasts and licensing music and making animations and illustrations for ads are already displaced and wondering what the fuck to do right now. I’m told there’s no shortage of work, but there’s a huge shortage of work for everything I’ve worked for.
Such a funny, thoughtful and well written piece. The image of screeching terminators coming to do our graphic design made me laugh out loud on the subway.
But they are doing that, because (wait for it) people don't currently appear to value that as much as they pretended to, apparently.
Now when everyone gets bored of AI slop, and everything looking the same, that's when actual designers should put their daily rate up, and the value is evident.
Adam - love your posts and ideas! The mouse and the cookie is the best explanation of this I've seen. I watch it at my martial arts school every week. A parent gets the result they came for (their kid can finally focus), and within a month they want the next thing: the next belt, the leadership team. The want that gets satisfied doesn't stay satisfied, it just gets replaced by a bigger one. That's why I'd bet on the business owners who keep finding the next desire over the ones waiting for the work to run out. And I'm fully on board with the Terminators fetching the mouse's straw.
In 2021 or thereabouts Kevin Kelly told Tim Ferriss that no one would lose their job to AI. But Kelly, when pressed, couldn't make much of an argument beyond his own hunch. I think you've written the argument. And it's quite compelling. Many thanks.
Your analysis is in line with the view of AI as a normal technology. What we're very possibly looking at as robotics advance and capabilities reach a tipping point is a cheap, highly-replicable workforce that can replace any job, including any new ones that might be created. I'm in software. I don't feel particularly safe. Ask a math grad student how they feel about their future right now. Ask any number of workers from the gig economy. There's inertia to human systems and people are still wrapping their heads around how to leverage these things. 2026 is the year of agents, but they haven't been fully incorporated seamlessly into workflows yet, and their capabilities are still increasing. How exactly are they going to do anything but reduce jobs? Nothing in your analysis makes that particularly clear. Maybe you'd say it's unpredictable. I'm not sure about that. If what we're actually looking at is the ability to inject an enormous pool of skilled slave labor into the workforce, the results seem pretty predictable. And not good for human labor.
thing is the automated loom couldn't also do your taxes and whatever else you could imagine, what happens when automation keeps pace if not exceeds the rate of our desires? (at least with knowledge work)
That children's book is a brilliant analogy. I'm so glad you wrote this. I'd been wanting to put something out like this, so I'm so glad you already did us this service. Thanks!
Let’s say that you are right that the jobs we lose will be replaced… but there will be plenty of old people replaced by young people in the new jobs, and the old people will never find jobs again.
This happened in the UK in the 1980s. All the factories and the mines and the manufacturers closed down and we had 3,000,000 (14%) unemployed for several years. All those jobs grew back, but they were different jobs and it was the young people who took them. 40% of the unemployed never had a proper job again.
Next time it will be all the drivers and the shopkeepers and the office workers who lose their jobs. Maybe the software engineers will figure out how to work with their new AI co-workers, but the drivers won’t.
Great article, it is fascinating how chronic existential avoidance appears to create an infinite array of superficial value. 🌱
What we value has always been inextricably linked to what we perceive and how we interpret it. A simple but fitting example: a blind person cannot value the sight-correcting features of a well-engineered pair of glasses. For something to be valued, it has to be perceived as an amelioration of a perceived lack.
You ask why we should believe in a tomorrow that has nothing to do with today while citing the percentage of people in agriculture which looks nothing like yesterday. I'd rather see an object-level argument about what exactly humans will be doing, once AI can perform all the mental and physical labor at an above human level.
So you’ve now defined a third camp in the AI debate: there are the doomers, there are the utopians (or is it effective altruists?) … but what do we label your position? (It’s one I share, by the way.)
Adam can answer for himself, but I see 4 main camps:
1. Doomers who think AI will kill us all and/or take all our jobs. This group includes many members of the Effective Altruist cult.
2. Boomers who think AI is wonderful magic, will cure cancer within 10 years, and so on.
3. Skeptics who think AI is a scam run by billionaire grifters.
4. Realists who think AI is a powerful and useful tool which will be of great benefit, but the smartest AI in the world won't eliminate the countless bottlenecks in the world of atoms, organisms, and culture.
(Obviously given how I have labeled these camps, I put myself in the fourth one.)
Well, if AI can eliminate the organisms and the culture I’d say it’s going to do OK with just the atoms
Sadly as a freelance creative ai has already, in very short order, replaced most of the things I’ve been hired to do over my 25 year career. Repercussions 20 years in the future I can’t imagine, but my miserable books for 2025-2026 are far from theoretical. Creatives who’ve been making logos and branding packages and producing videos and podcasts and licensing music and making animations and illustrations for ads are already displaced and wondering what the fuck to do right now. I’m told there’s no shortage of work, but there’s a huge shortage of work for everything I’ve worked for.
Such a funny, thoughtful and well written piece. The image of screeching terminators coming to do our graphic design made me laugh out loud on the subway.
But they are doing that, because (wait for it) people don't currently appear to value that as much as they pretended to, apparently.
Now when everyone gets bored of AI slop, and everything looking the same, that's when actual designers should put their daily rate up, and the value is evident.
But when will that be...
Adam - love your posts and ideas! The mouse and the cookie is the best explanation of this I've seen. I watch it at my martial arts school every week. A parent gets the result they came for (their kid can finally focus), and within a month they want the next thing: the next belt, the leadership team. The want that gets satisfied doesn't stay satisfied, it just gets replaced by a bigger one. That's why I'd bet on the business owners who keep finding the next desire over the ones waiting for the work to run out. And I'm fully on board with the Terminators fetching the mouse's straw.
In 2021 or thereabouts Kevin Kelly told Tim Ferriss that no one would lose their job to AI. But Kelly, when pressed, couldn't make much of an argument beyond his own hunch. I think you've written the argument. And it's quite compelling. Many thanks.
Your analysis is in line with the view of AI as a normal technology. What we're very possibly looking at as robotics advance and capabilities reach a tipping point is a cheap, highly-replicable workforce that can replace any job, including any new ones that might be created. I'm in software. I don't feel particularly safe. Ask a math grad student how they feel about their future right now. Ask any number of workers from the gig economy. There's inertia to human systems and people are still wrapping their heads around how to leverage these things. 2026 is the year of agents, but they haven't been fully incorporated seamlessly into workflows yet, and their capabilities are still increasing. How exactly are they going to do anything but reduce jobs? Nothing in your analysis makes that particularly clear. Maybe you'd say it's unpredictable. I'm not sure about that. If what we're actually looking at is the ability to inject an enormous pool of skilled slave labor into the workforce, the results seem pretty predictable. And not good for human labor.
Everyone is forgetting that this is not free.
The pushers want you to believe it is free, but it's not.
It will play out, and the world will realise, and a lot of harm will be done.
But it's not free.
Probably your best yet, Adam.
And great picture.
thing is the automated loom couldn't also do your taxes and whatever else you could imagine, what happens when automation keeps pace if not exceeds the rate of our desires? (at least with knowledge work)
it falls apart When You Give a Moose a Muffin
Truth IS in the woods. Been following Pinsof for a long time..happy to hear him mentioned here!
That children's book is a brilliant analogy. I'm so glad you wrote this. I'd been wanting to put something out like this, so I'm so glad you already did us this service. Thanks!
October 13, 2026 headline:
Kroger to lay off 3,000 corporate employees over 3 years as it leans into artificial intelligence
Great. And it will hire half of them back when it realisers the actual cost. Just watch.
Let’s say that you are right that the jobs we lose will be replaced… but there will be plenty of old people replaced by young people in the new jobs, and the old people will never find jobs again.
This happened in the UK in the 1980s. All the factories and the mines and the manufacturers closed down and we had 3,000,000 (14%) unemployed for several years. All those jobs grew back, but they were different jobs and it was the young people who took them. 40% of the unemployed never had a proper job again.
Next time it will be all the drivers and the shopkeepers and the office workers who lose their jobs. Maybe the software engineers will figure out how to work with their new AI co-workers, but the drivers won’t.
Great article, it is fascinating how chronic existential avoidance appears to create an infinite array of superficial value. 🌱
What we value has always been inextricably linked to what we perceive and how we interpret it. A simple but fitting example: a blind person cannot value the sight-correcting features of a well-engineered pair of glasses. For something to be valued, it has to be perceived as an amelioration of a perceived lack.
https://athinkerinnature.substack.com/p/the-evolution-of-value
Came to the comments to cite A Very Hungry Caterpillar (Carle 1969), discovered footnote 2 already covers it 😂
But somehow inexplicably doesn't mention The Tiger Who Came To Tea!
👀