We only need look at what the computer did to many (60%) non-college educated people who have difficultly finding meaningful work which can provide for a family.
Daron Acemoglu is talking about this and Building Pro-Worker AI. His 2024 Nobel Prize lecture is quite good (as are his fellow laureates that year)
It's not the fault of computers, but they changed the fabric. It was a select few of our fellow humans, who also did deregulation, financialization, and globalization. They all started at the same time and have concentrated wealth and power the last 4-5 decades.
Computers are still a significant factor because most new tasks created were cognitive and needed university education. Or another way, the automation that computers brought eliminated jobs in labor, but we did not create new jobs for those people. The three others were instrumental in that side of the equation.
Well, we did create new jobs for them technically. It’s just that the jobs are terrible. Those disrupted in the past 20 years have been given gig-economy BS to replace their jobs.
The jobs created for the factory workers and the former professional retail workers, in the 90s, were in big-box stores.
Yes. It removes political power. As a thought experiment, what happens when sanitation workers go on strike? What would happen if all trash collection was automated?
What happens when a group of people goes from the driver's of the economy to a cost center?
It might very well be. Poe's Law. But there is actually a not-insignificant group of people who believe this sort of thing. Or would go along with it if they thought it was in some way logical and they wouldn't be among the re-educated or exterminated.
> I think the economically rational choice is re-education of the group into value-providing employment
I say this who is voluntarily returning to university to pursue a different profession after working 10 years as a software engineer. This isn't going to work for 95% of people, people who have dependents children and partners, mortgages.
I distinctly did not get the impression that GP is happy about or approving of what he’s suggesting may be “economically rational.” It reads more as foreboding to me.
So in Hitler's Germany, you would have been one of the Germans who felt a bit uneasy about the trains, death camps, etc., and then put on your uniform to carry out your orders?
Like the previous commenter said, people have dependents, mortgages, children, partners, etc. Percentually many more are tied up economically with the capital/state than in 1940s Germany.
Given the current economic trajectory there will be a line to put on uniforms, I don't think I am more eligible for this task than many others who have more dependents.
I'm not suggesting it would be, but it sure is interesting to see people who are far better educated and have way more historical perspective than the average 1930s German did be so eager (or complacent) to follow in their footsteps down the same very dead end.
Either way, he’s a dickhead. He’s clearly throwing around words he doesn’t really understand.
It wouldn’t be economically rational to do this. Nations have a monopoly on violence (policing, military & enforcement of through force) within their own borders, and nations are in competition with one another, even during times of peace.
It does not make sense for anyone nation to start culling people who do not get onboard with some radical social engineering programme. Through the stroke of a pen these companies are under the control of the government.
Say you want to make an even more pessimistic argument, say the oligarchy in one country get control of this monopoly on its violence. That country is still in competition with other countries.
Russia still has nukes, and thanks to their little war, they lack the technological expertise to go to toe to toe with China and the US on AI. Countries wanna mobilise what resources they can to survive this strategic competition and that includes humans.
Say the oligarchs are smart enough to maintain the status quo in foreign policy, and they seek to enslave their entire country. There is no amount of resources in the world that would prevent the amount of disenfranchised people from revolting and over-throwing any such government that didn’t give them a fair social contract, especially one that’s already gone through higher levels of economic development.
It is economically rational, the game masters/ruling class/aristocracy/epstein class keeps the peasants around because they need their labor.
Once AI can replace them, keeping the peasants around means they gobble up your energy, food, land, resources while returning nothing.
He is not advocating for exterminating people, he is just being a realist, like it or not this is the most likely outcome. Having humans is not a advantage since their usefulness is low and their resource consumption is high. The first country to get rid of them and replace them with machines will have a unassailable economic advantage. China and Russia will ultimately lose since they are nationalists dedicated to preserving the chinese and russian ethnicities. They will keep their humans around and be outcompeted.
You're still undershooting just how grim the optimal outcome of our system is. It's not going to be about 'refusal', like taking a bargain and making choices. If total automation is achieved, huge swaths of people will be placed into that bracket immediately, based largely on their luck, the country they live in and their wealth. An automated world doesn't need this many people to run itself, so only a chosen few will make it through. The rest will eventually be slowly destroyed in the most cruel ways possible. People will keep chanting "Adapt or die" until it turns into simply "Die".
No it doesn't. People can still protest, block roads, boycott etc... If they get UBI/UHI, they will have more time to engage politically too (in a productive way).
You obviously haven't lived in a country where labor strikes are a frequent occurrence. They block the roads for weeks, sometimes months at a time and products disappear from the shelves. Often, the police do nothing. This is where everything is heading. That's reality when most of the population is dissatisfied.
Obviously, the solution is to make it economically impossible for people to go on strike long enough for products to disappear from the shelves. Lax labour laws, high cost-of-living, privatized healthcare.
Why do you think Americans don’t go on strike like the French? Not many workers can afford to be out of a job, out of health insurance and eggs $30 a dozen.
I doubt you can prevent this using laws. In unequal societies where the politics is controlled by the majority, laws tend to make things worse, not better. Americans are already beginning to find out what happens when the political system becomes controlled by people who hate the country. This is a reality which France (and especially colonies) have been operating under for quite some time.
Perhaps the real answer is people’s buying power. If people can’t withhold their work, they could withhold their spending, ie boycott misbehaving entities.
>> Yes. It removes political power. As a thought experiment, what happens when sanitation workers go on strike? What would happen if all trash collection was automated?
>> What happens when a group of people goes from the driver's of the economy to a cost center?
> Perhaps the real answer is people’s buying power. If people can’t withhold their work, they could withhold their spending, ie boycott misbehaving entities.
But if they don't have work, they won't have buying power to withhold. Post AI-displacement, the economy will reorient around serving the people who still have buying power (the oligarchs and their hangers-on).
Expect demand for food to plummet and demand for dick-waving pyramids and rocket cars to skyrocket.
I came to this conclusion too and felt rather defeated, existential crisis, etc. But the more I thought about it, the more I realized that if we had a country of 250 million unemployed angry folks (many of them armed), I'm not so sure it'd be the wealthy in control.
> But the more I thought about it, the more I realized that if we had a country of 250 million unemployed angry folks (many of them armed), I'm not so sure it'd be the wealthy in control.
It depends on how hard they'll fight, but technology just doesn't seem like it's a force for good anymore. I could imagine that, in the face of a revolt against automation, some fraction of the oligarchs will just establish a Ukraine-war like AI-drone kill-zone to protect their assets. If they do that, they'd probably win.
But it probably wouldn't get to that point, because, given the online nature of culture, AI-powered surveillance and AI-powered social interference could very likely thwart any revolt before it starts.
They start waving flags around, voting for someone who will promise to "make things great again" all the while languishing in poverty and dying of fentanyl overdoses
The "Tea Party" was an astroturf operation organized by GOP operatives and funded by the Koch brothers and some other billionaire donors. Its appearance of popular support was substantially artificial and heavily amplified by supportive GOP-aligned media.
The majority of anti-tech sentiment in America has been coming out of left-leaning circles for the last five years. Go look at a video of the people protesting data centers in your nearest metropole; I guarantee you they look nothing like the caricature you have in your head.
“what happens when sanitation workers go on strike? What would happen if all trash collection was automated?”
This is an interesting question and I’d love to come to a good answer that works for all, but can we also accept that we went from a world that had people manually light lamps around the city like 150 [1] years ago to the full electric lights. Do we then stop that progress now and keep people manually collecting garbage when a replacement tech is around the corner?
[1] mm more like 120 years. The last lamplighters strikes was close to 1907 in NYC
If you “socialized” Edison’s electricity and gave the control to the lamplighters, we’d all be living under gas lamps now. There is zero reason for the lamplighters to negotiate any progress whatsoever.
Listen this is a great slogan, and I'm sure you can get people whipped out to vote under a giant populist banner and win, but you have to at least have a second thought here about not adopting electricity to preserve a job to be exactly how its been done.
We can't even have a conversation about extended unemployment benefits, retraining credits or lamplighters transitioning to utility workers, which is what happened with the "basket of jobs"?
> We can't even have a conversation about extended unemployment benefits, retraining credits or lamplighters transitioning to utility workers, which is what happened with the "basket of jobs"?
That's all we've been doing while labelling it "progress".
I am pointing out that not all changes are progress. I am not claiming that all changes are not progress.
“I am pointing out that not all changes are progress. I am not claiming that all changes are not progress.”
Right but my point was who gets to decide? If you make it a council of elders and people whose jobs might be affected the answer is always going to be All change is bad progress. End of story.
Yes, after electric lamp light systems became so important that everyone needed it/wanted it, they became socialized.
That would be different from giving lamplighters control over conversion to electric, who would claim that electricity is too dangerous to use and just kill it. Case in point [1]
This is weird obfuscation guys. By the time any socialization happened the conversion from gas to electric was not in question.
I’m happy to socialize the automation companies once the conversion and takeover of automation increase the standard of living of everybody cannot be blocked.
Socializing and then having committee meetings about “how much automation is appropriate” is the situation I’d want to avoid.
This is weird obfuscation guys. Many socialized industries have had technologies converted. Socialized power utilities seem to be going solar faster than private ones.
Look at the Roman Republic and how things went for the plebes as more and more slaves took over all the work. I pointed out in another thread they went form citizen farmers to having huge farms owned by the rich worked by slaves. Rome had to implement the bread dole so people could eat.
Of course. This has been proven again and again on the economic level. People have to let go of old skills and learn new ones to survive.
For AI though, there is a mental angle which is being discussed now (brain gyms etc.) which was an early concern of mine. When we make something, there is an tangible value created for the buyer/consumer but there is also a more subtle positive effect the on the builder/maker. e.g. even if it's it's cheaper, faster, (and in an imaginary worlds), more healthy to buy packaged food, sitting down and making it subtly changes the chef who makes the food. Sort of the motivation for a grandmother baking a cake for her grandkids.
I've been able to realise many of my abandoned projects due to AI. I can spin up a few agents and have them build things completely without intervention more or less. Many of these are programs I use daily. This is definitely huge. However, I remember the kick of struggling with problems during projects which was where most, if not all, the learning happened. All that experience is what makes me able to command and control these agents properly. It had an effect on me beyond the outcome (most of the projects are dead/abandoned anyway). Another example - I remember an old company called "Bond" which had a robot to create stationery using a calligraphic hand. Contrast that to the almost spiritual effect doing calligraphy has on a master penman. Using automation to create fine calligraphy gets you the former but the individual and society loses the latter. It's not immediately clear to me what the end result will be but I do feel that a lot is being lost.
I like that calligraphy example, I am afraid that's how a field would stagnate: the tools become advanced, but now with less and less people wanting to approach and struggle with learning the field, the human intuition, the community effort, and the mastery to push past the bleeding edge withers.
Think about ceramic dinner plates and other table items.
You can get a brand new dinner plate or cereal bowl or coffee cup for a pittance - but you can also get a hand-shaped, hand-painted cat-themed coffee mug at a significantly higher price. Adult education courses in ceramics support the base of amateur and semi-pro practitioners.
I have sometimes observed that dinner plates are surprisingly expensive in pricing. Or at least sometimes they seem to. You do get really cheap ones too. But on average it seems many are quite or somewhat expensive.
While I agree that new skills must be learnt to survive, what exactly are these new skills in the AI-age?
I see this line everywhere - whenever someone's mourning their loss of a career, there's responses telling the person to "man up" (for lack of a better phrase) and learn new skills. Nobody says what these "new skills" are?
it's unknown, because those skills and work associated with them are not yet created.
Just like how being a live streamer (or youtuber) is not a known skill until the advent of streaming platforms and their associated economies. If you asked on the internet in 1995 what kids would have done for a job in 2025, none of them would've said streamer.
Understood. Then how does one brace for the upcoming slaughter? People knew that getting into software would become the hot new thing - almost guaranteeing a source of income. Hence the "push" for everyone to get into "coding", at least in my country (India) and from what I've heard on the internet - the US too.
Ya fear of the unknown is not even anything uncommon for sure.
Honestly ive only read "the power of now",
Which I later learned was very mainstream popular, so I probably would have prejudged it had I knew.
While not really original ideas he packages the concepts into very good words, so it made something click, for me at least, where other books hadn't.
I don't have a clear answer myself. But there are multiple dimensions to this.
First is from an employment, livelihood perspective. I don't know the "right" answer but like with most things, it's easy to see the wrong ones - Bury your head in the sand and declare that this is just a passing phase, insist on going deeper into curmudgeon territory waxing nostalgic about the "Golden Era" of programming that's over etc. The opposites (trying out the new tools that are being released, reading up on things happening and developing a mental model of how things are moving etc.) seem to be sensible things to do.
Second, from a personal perspective. Here too, inversion is useful. Stopping to think is a bad idea. Like all our faculties, you lose what you don't use. So, actively learning things that are hard (a new "natural" language, a new phsyical skill that you wanted to) will help keep you sharp. Recreational programming is fine I guess and you can use AI to give you that you extra push when you're mired in an area that you don't care for and which you know how to do but are avoiding because of the tedium.
When the demand for work comes, it's quite clear to me that someone who's done what they can and has a some insight on what's going on has an advantage over the person who hasn't. Whether that translates into economical opportunities is beyond us to say but it's beyond our control as well so there's no point in worrying about it.
They’re unknown because they don’t exist. There will simply be skills that haven’t yet been automated, but nothing will remain out of reach. The idea of having a career in a single field is probably dead, unless that field is minimum wage laborer.
Actually, you can't unless you meet a karma threshold. I have been on this site for 6 years and I still haven't reached it because my opinions are too controversial (apparently).
Not controversial. Just not aligned with the owners of HN.
The comments you upvote/downvote are used to calculate your "opinion similarity" to the owners. Then your upvote/downvotes are modulated by that similarity. It's a way of using other humans to amplify the opinions of the owners -- they don't have time to read all the comments, so instead they filter out those who think the same way they do and let them do it.
I've contacted HN before and they've told me they auto-downvote certain kinds of content. It's not a stretch to assume their views are influencing that criteria.
Oh yeah, I think I remember reading they demote submissions that they think will cause low quality discussions. I don't think that's related to downvoting or your karma.
Guilty... but also, I think about article up/downvotes an order of magnitude less than I click the links. I wonder what fraction of users do voting without clicking links.
Daron Acemoglu is talking about this and Building Pro-Worker AI. His 2024 Nobel Prize lecture is quite good (as are his fellow laureates that year)
https://www.nber.org/papers/w34854
Computers are still a significant factor because most new tasks created were cognitive and needed university education. Or another way, the automation that computers brought eliminated jobs in labor, but we did not create new jobs for those people. The three others were instrumental in that side of the equation.
The jobs created for the factory workers and the former professional retail workers, in the 90s, were in big-box stores.
What happens when a group of people goes from the driver's of the economy to a cost center?
Otherwise you might end up with disruptive things like protests, blockage of value creating infrastructure, etc.
Another less dramatic option would be containment in prisons or similar where minimal costs can be incurred.
I imagine a human body will still be cheaper and more agile than robots for a long while.
A machine has superior eyes and reflexes, a single robot with a gun would kill a infinite number of humans with the same gun.
I say this who is voluntarily returning to university to pursue a different profession after working 10 years as a software engineer. This isn't going to work for 95% of people, people who have dependents children and partners, mortgages.
> mass extermination of those that refuse
I have no words, but you sound like a dickhead.
Given the current economic trajectory there will be a line to put on uniforms, I don't think I am more eligible for this task than many others who have more dependents.
It wouldn’t be economically rational to do this. Nations have a monopoly on violence (policing, military & enforcement of through force) within their own borders, and nations are in competition with one another, even during times of peace.
It does not make sense for anyone nation to start culling people who do not get onboard with some radical social engineering programme. Through the stroke of a pen these companies are under the control of the government.
Say you want to make an even more pessimistic argument, say the oligarchy in one country get control of this monopoly on its violence. That country is still in competition with other countries.
Russia still has nukes, and thanks to their little war, they lack the technological expertise to go to toe to toe with China and the US on AI. Countries wanna mobilise what resources they can to survive this strategic competition and that includes humans.
Say the oligarchs are smart enough to maintain the status quo in foreign policy, and they seek to enslave their entire country. There is no amount of resources in the world that would prevent the amount of disenfranchised people from revolting and over-throwing any such government that didn’t give them a fair social contract, especially one that’s already gone through higher levels of economic development.
Once AI can replace them, keeping the peasants around means they gobble up your energy, food, land, resources while returning nothing.
He is not advocating for exterminating people, he is just being a realist, like it or not this is the most likely outcome. Having humans is not a advantage since their usefulness is low and their resource consumption is high. The first country to get rid of them and replace them with machines will have a unassailable economic advantage. China and Russia will ultimately lose since they are nationalists dedicated to preserving the chinese and russian ethnicities. They will keep their humans around and be outcompeted.
It already looks like this is underway if you look at the state of health care and how little is being done about it.
Why do you think Americans don’t go on strike like the French? Not many workers can afford to be out of a job, out of health insurance and eggs $30 a dozen.
>> What happens when a group of people goes from the driver's of the economy to a cost center?
> Perhaps the real answer is people’s buying power. If people can’t withhold their work, they could withhold their spending, ie boycott misbehaving entities.
But if they don't have work, they won't have buying power to withhold. Post AI-displacement, the economy will reorient around serving the people who still have buying power (the oligarchs and their hangers-on).
Expect demand for food to plummet and demand for dick-waving pyramids and rocket cars to skyrocket.
It depends on how hard they'll fight, but technology just doesn't seem like it's a force for good anymore. I could imagine that, in the face of a revolt against automation, some fraction of the oligarchs will just establish a Ukraine-war like AI-drone kill-zone to protect their assets. If they do that, they'd probably win.
But it probably wouldn't get to that point, because, given the online nature of culture, AI-powered surveillance and AI-powered social interference could very likely thwart any revolt before it starts.
What kind of "left" version are you imagining?
The Open Society Foundation.
This is an interesting question and I’d love to come to a good answer that works for all, but can we also accept that we went from a world that had people manually light lamps around the city like 150 [1] years ago to the full electric lights. Do we then stop that progress now and keep people manually collecting garbage when a replacement tech is around the corner?
[1] mm more like 120 years. The last lamplighters strikes was close to 1907 in NYC
No we take the tech and socialize its control
If you “socialized” Edison’s electricity and gave the control to the lamplighters, we’d all be living under gas lamps now. There is zero reason for the lamplighters to negotiate any progress whatsoever.
Not all change is progress. Just because you can mass-remove populations of people from the economy does not mean that doing so is "progress".
We can't even have a conversation about extended unemployment benefits, retraining credits or lamplighters transitioning to utility workers, which is what happened with the "basket of jobs"?
That's all we've been doing while labelling it "progress".
I am pointing out that not all changes are progress. I am not claiming that all changes are not progress.
Right but my point was who gets to decide? If you make it a council of elders and people whose jobs might be affected the answer is always going to be All change is bad progress. End of story.
Who gets veto powers?
That would be different from giving lamplighters control over conversion to electric, who would claim that electricity is too dangerous to use and just kill it. Case in point [1]
[1]: https://www.cambridgeday.com/2026/10/26/rally-against-waymo-...
I’m happy to socialize the automation companies once the conversion and takeover of automation increase the standard of living of everybody cannot be blocked.
Socializing and then having committee meetings about “how much automation is appropriate” is the situation I’d want to avoid.
None that threatened jobs. "Switching to solar" didn't require people to transition to different jobs.
That's the crux of my point, ignore it, is obfuscation.
People got very very angry at me for pointing this out.
So I stopped doing it.
For AI though, there is a mental angle which is being discussed now (brain gyms etc.) which was an early concern of mine. When we make something, there is an tangible value created for the buyer/consumer but there is also a more subtle positive effect the on the builder/maker. e.g. even if it's it's cheaper, faster, (and in an imaginary worlds), more healthy to buy packaged food, sitting down and making it subtly changes the chef who makes the food. Sort of the motivation for a grandmother baking a cake for her grandkids.
I've been able to realise many of my abandoned projects due to AI. I can spin up a few agents and have them build things completely without intervention more or less. Many of these are programs I use daily. This is definitely huge. However, I remember the kick of struggling with problems during projects which was where most, if not all, the learning happened. All that experience is what makes me able to command and control these agents properly. It had an effect on me beyond the outcome (most of the projects are dead/abandoned anyway). Another example - I remember an old company called "Bond" which had a robot to create stationery using a calligraphic hand. Contrast that to the almost spiritual effect doing calligraphy has on a master penman. Using automation to create fine calligraphy gets you the former but the individual and society loses the latter. It's not immediately clear to me what the end result will be but I do feel that a lot is being lost.
This is the problem.
(Unless if AGI could do something like that.)
You can get a brand new dinner plate or cereal bowl or coffee cup for a pittance - but you can also get a hand-shaped, hand-painted cat-themed coffee mug at a significantly higher price. Adult education courses in ceramics support the base of amateur and semi-pro practitioners.
I see this line everywhere - whenever someone's mourning their loss of a career, there's responses telling the person to "man up" (for lack of a better phrase) and learn new skills. Nobody says what these "new skills" are?
it's unknown, because those skills and work associated with them are not yet created.
Just like how being a live streamer (or youtuber) is not a known skill until the advent of streaming platforms and their associated economies. If you asked on the internet in 1995 what kids would have done for a job in 2025, none of them would've said streamer.
How do people brace for the changes AI brings?
First step is to let go of the fear of it.
Sounds cheesy but reading Ekehardt Tolle helped me.
Which one of his books should I read (first)?
Honestly ive only read "the power of now",
Which I later learned was very mainstream popular, so I probably would have prejudged it had I knew. While not really original ideas he packages the concepts into very good words, so it made something click, for me at least, where other books hadn't.
First is from an employment, livelihood perspective. I don't know the "right" answer but like with most things, it's easy to see the wrong ones - Bury your head in the sand and declare that this is just a passing phase, insist on going deeper into curmudgeon territory waxing nostalgic about the "Golden Era" of programming that's over etc. The opposites (trying out the new tools that are being released, reading up on things happening and developing a mental model of how things are moving etc.) seem to be sensible things to do.
Second, from a personal perspective. Here too, inversion is useful. Stopping to think is a bad idea. Like all our faculties, you lose what you don't use. So, actively learning things that are hard (a new "natural" language, a new phsyical skill that you wanted to) will help keep you sharp. Recreational programming is fine I guess and you can use AI to give you that you extra push when you're mired in an area that you don't care for and which you know how to do but are avoiding because of the tedium.
When the demand for work comes, it's quite clear to me that someone who's done what they can and has a some insight on what's going on has an advantage over the person who hasn't. Whether that translates into economical opportunities is beyond us to say but it's beyond our control as well so there's no point in worrying about it.
You're blind to the stuff that's easy making previously impossible things hard.
This change threatens to hurt those who move with it, too.
The comments you upvote/downvote are used to calculate your "opinion similarity" to the owners. Then your upvote/downvotes are modulated by that similarity. It's a way of using other humans to amplify the opinions of the owners -- they don't have time to read all the comments, so instead they filter out those who think the same way they do and let them do it.
As someone who likely has many different opinions from people from those of people at Y Combinator, this seems wrong.
I think the stats on interactions across the board are probably ridiculous though.
I wish for a HN, reddit, etc, where you have to vote and comment to even be served more content.
[1] https://get.paid.ai/hubfs/The%20$2.7%20Billion%20Agent%20Tax...