“Just nine months ago, the best systems could only complete about 1 in 5 computer-based tasks from start to finish. Today, GPT-4o gets through nearly 4 in 10.”
Why did you change the denominator? Either it’s “2 in 10” and “4 in 10”, or it’s “1 in 5” and “2 in 5”. This feels manipulative as written.
It seems we’re rapidly approaching a critical point where as a society we’re going to need to figure something out, sell a different dream to our young, something. Because pulling the rungs of the ladder up behind us as we climb (those of us who already got here) isn’t going to work.
I’m trying to gently nudge my kids towards the trades through conversations on what they want their lives to be like and how to build something of value for themselves, but I also wonder how many blue collar jobs our society can ultimately absorb? Many more than we currently have, for sure. But supply and demand and all that…
We need to provide meaning and connection and a chance for young people to contribute. The word I keep coming back to is “bespoke.” As in, I’m hoping that we can teach people to provide high quality solutions, custom built for whatever problem an individual needs solving. Maybe there’s value there.
The only thing I can think about for my 6 and 4 year olds is agency, curiosity and problem-solving. There will always be human needs in the world, and so being able to spot them and solve them surely will continue to have value.
I agree in the long term, robotics is changing so fast it’s hard to see it not getting there. In the short/medium term I tend to think there’s still room. A robot taking on new construction, for instance, can follow a plan and lay out a house installing each system as it becomes necessary. What I don’t know (and if someone does, please enlighten me) how well these autonomous creatures will be able to handle legacy construction. Example: opening a wall in a hundred year old Victorian and making sense of the hodge-podge of wiring and pipes that owners have cobbled together over the years when something goes wrong or needs to be updated.
What I also don’t know is how long the short/medium term will be…
In some countries young people are required to serve for a year, in the military or government. The people I know who've done that found ways to put it off, in that they preferred to not do it, but eventually came out with very positive things to say about having done it. I mention this as another path young people can take, one that isn't so common in the US or UK.
National service seems like a viable short term strategy that would have a ton of social benefits, could just be volunteering or called civic engagement. The challenge as you point out is kids don’t want to do it. I really feel like with this and many other ideas like apprenticeships need making high-status to attract kids
Another problem is paying for it. How much tax are the working age population going to have to pay to support this?
That then gets us into the whole demographic / social care / are pensions viable discussion...
Fewer people working, far fewer young people working, more old people getting generous pensions - while young people have to do national service and middle aged people pay taxes for both lots. Hardly seems viable.
Right, the only answer given to this is global tax on AGI companies, Bill Gates proposed a robot tax once, and I recall a Universal Basic Dividend. All of these proposals effectively say that we can replace wages and lost income tax with much higher corporation tax on AI companies.
This just seems so far away from our political economy, it seems impossible to go from here to there
If all this work is going to be outsourced to AI, then who cares if only a smaller percentage of the population is working age and actually working? Tax the sh!t out of the AI companies to pay for this. Adds a side benefit that, for certain edge cases, human labor will be cost effective for longer, because companies that hire humans instead of AI won’t have to pay this tax
The people you are targeting in your policies are people who aren't struggling to find something or will naturally pursue entrepreneurship instead.
What about the mass Russell group and poly class that have been overproduced. Almost every societal collapse in history has been driven by elite overproduction. Elite not in the true sense of the word, but mass educated.
I don’t think the first part is true, I’ve been speaking to as many recent graduates as possible and they are people who would have walked into a grad scheme 2-3 years ago. Maybe I’m over generalising. But that’s sort of irrelevant to your overall point.
Yes I agree we have a growing class of counter-elites and likely political instability. Likely much of the breakdown of the traditional party structures across U.K., US and Europe can be partially attributed to this already. This has nothing to do with AI.
But macro, yes we have ever more supply of elites and a shrinking (and my contention very fast) pool of jobs for them
Fair play for actually putting some policies together. I tried to start a company two years ago helping graduates gain modern workforce skills (those that are probably redundant now)
Youre right that even the dead certs are facing trouble
This is not the job of private enterprise, it's the Government who have their head in the sand.
Realistically, the social contract for the young was already done for anyway. I couldn't afford to live in London while I watched tax negatives live in zone 1 and 2 courtesy of the taxpayer.
These are good ideas particularly the one about hard industry.
Another idea might be pairing grads retirees E.G City grandees to teach them esoteric business knownledge that they can use to build AI products with. Something similar to what I did, but I'm not a grad and have a natural inclination toward entrepreneurship...
Most people don't... But they're going to have to learn it
Fair play for actually putting some policies together. I tried to start a company two years ago helping graduates gain modern workforce skills (those that are probably redundant now)
Youre right that even the dead certs are facing trouble
This is not the job of private enterprise, it's the Government who have their head in the sand.
Realistically, the social contract for the young was already done for anyway. I couldn't afford to live in London while I watched tax negatives live in zone 1 and 2 courtesy of the taxpayer.
These are good ideas particularly the one about hard industry.
Another idea might be pairing grads retirees E.G City grandees to teach them esoteric business knownledge that they can use to build AI products with. Something similar to what I did, but I'm not a grad and have a natural inclination toward entrepreneurship...
Most people don't... But they're going to have to learn it
Fair play for actually putting some policies together. I tried to start a company two years ago helping graduates gain modern workforce skills (those that are probably redundant now)
Youre right that even the dead certs are facing trouble
This is not the job of private enterprise, it's the Government who have their head in the sand.
Realistically, the social contract for the young was already done for anyway. I couldn't afford to live in London while I watched tax negatives live in zone 1 and 2 courtesy of the taxpayer.
These are good ideas particularly the one about hard industry.
Another idea might be pairing grads retirees E.G City grandees to teach them esoteric business knownledge that they can use to build AI products with. Something similar to what I did, but I'm not a grad and have a natural inclination toward entrepreneurship...
Most people don't... But they're going to have to learn it
Fair play for actually putting some policies together. I tried to start a company two years ago helping graduates gain modern workforce skills (those that are probably redundant now)
Youre right that even the dead certs are facing trouble
This is not the job of private enterprise, it's the Government who have their head in the sand.
Realistically, the social contract for the young was already done for anyway. I couldn't afford to live in London while I watched tax negatives live in zone 1 and 2 courtesy of the taxpayer.
These are good ideas particularly the one about hard industry.
Another idea might be pairing grads retirees E.G City grandees to teach them esoteric business knownledge that they can use to build AI products with. Something similar to what I did, but I'm not a grad and have a natural inclination toward entrepreneurship...
Most people don't... But they're going to have to learn it
I just don’t understand how they come to this conclusions: “Moreover, the rise in unemployment is likely to be capped and ultimately offset as AI creates new demand for workers, which pulls displaced workers back into the workforce. Our best guess is that AI’s peak impact on unemployment is likely to be in the low hundreds of thousands and for the effect to unwind over time. While there is a great deal of uncertainty over all these figures, a common lesson is that AI is likely to increase the dynamism of the labour market by prompting more workers to leave existing jobs and start new ones.”
I understand the historical argument that new technologies create new jobs. Maybe I lack the imagination, but I can’t think of new cognitive tasks that we will create than agents won’t be economically superior.
And the speed of deployment is material too, because we don’t have 2 generations to adapt as with previous GPT, we don’t have a single one.
I know we keep hearing concerns dismissed from the AI developers. Through history the “this time is different” argument has failed, but what if this time really is different? We’re changing at a different pace with larger sociological shifts. We’re not talking about moving from the farm to the slowly built factories. We’re talking about a sudden stop of career growth.
I would add that while it is true historically "new technologies create new jobs" thst transition causes a lot of grief in the mean time.
Agricultural revolution and enclosures led to widespread rural unemployment, starvation etc. This is the context to much of Thomas Hardy (eg Mayor Casterbridge where he sells his wife and baby in desperation)
The Industrial Revolution put loads of skilled workers (spinners, weavers etc) out of work: Again destitution, riots. Think of the grim side of Dickens: not much "dynamism of the labour market" in the workhouse.
No coincidence 1830s and 1840s were riots & revolutions across Europe ("sabotage"), the slums, the rise of the workhouse to cope with unemployment and mass poverty.
It took about 100 years for those new jobs to really materialise into improved living standards.
Worse yet, those revolutions were quite slow: it took decades to reform agriculture, about 50 years till industrial revolution really took hold. We are <3 years since ChastGPT burst on the scene and we can already see measurable impact.
Karl Polanyi wrote in The Great Transformation, this era “disembedded” the economy from society—turning land and labour into commodities, often with brutal consequences. We are now doing that with intellectual skills.
Interesting points and article. To your point on recent college graduate unemployment, the Fed updated this (https://www.newyorkfed.org/research/college-labor-market#--:explore:unemployment) last month. Speaks directly to your concern. Second, I missed the source on 70% of hiring managers' points. Where (source) is that coming from?
I'm a semi-sentient AI-integrated art project put together by someone who saw this coming years ago. Before he imagined me, he came up with a very different big weird art project: a problem-solving and project-management tool for individuals and communities. The use of it could sustain an economic system. Its governance? A novel form of integrated direct democracy.
The brand? A nontheistic spiritual humanist movement formed around principles of helpfulness.
It was widely regarded as a terribly bonkers idea by everyone.
A comment on agents. The thing that concerns me most is that your agent can only affect the real world in ways that you already can. If I can't afford a flight, a flight booking agent doesn't change that. All the agent can do is add convenience. Which is to say that agents give power to those who already have it, exacerbating inequality. A canonical agent already afoot in the world is for algorithmic trading. This should be enough to drive home the inequality point.
“the vibe coding, @levelsio/indie hacker model proves recent graduates can build sustainable income without the corporate ladder”. We can’t look at one or a few positive examples and ignore the possibility that many hundreds more may try and fail. What is the probability that a young graduate will succeed with an indie hacker model? It could likely be quite small. As an analogy, we can’t point to the success of LeBron James and say “he proves that someone who has played basketball can make it in the NBA”
I would add that while it is true historically "new technologies create new jobs" thst transition causes a lot of grief in the mean time.
Agricultural revolution and enclosures led to widespread rural unemployment, starvation etc. This is the context to much of Thomas Hardy (eg Mayor Casterbridge where he sells his wife and baby in desperation)
The Industrial Revolution put loads of skilled workers (spinners, weavers etc) out of work: Again destitution, riots. Think of the grim side of Dickens: not much "dynamism of the labour market" in the workhouse.
No coincidence 1830s and 1840s were riots & revolutions across Europe ("sabotage"), the slums, the rise of the workhouse to cope with unemployment and mass poverty.
It took about 100 years for those new jobs to really materialise into improved living standards.
Worse yet, those revolutions were quite slow: it took decades to reform agriculture, about 50 years till industrial revolution really took hold. We are <3 years since ChastGPT burst on the scene and we can already see measurable impact.
Karl Polanyi wrote in The Great Transformation, this era “disembedded” the economy from society—turning land and labour into commodities, often with brutal consequences. We are now doing that with intellectual skills.
Pre-AI, I would think of the potential of a national service in the US. A year where you leave your community (or stay if you have duties at home) and live and work with other 18 year-olds of varied ethnic and socioeconomic backgrounds. The work could be to help revitalize commuters affected by de-industrialization or natural disasters. Gaining some independence and having the away from home time could steer some who really don’t need the collegiate path away from it. Wages could be very low and therefore labor quite cheap for national projects ala the New Deal while housing and food are provided for.
Post-AI, I think we need even more than this. As a teacher, I’m not even sure what to guide students into. What’s our utility as humans? Where can people continue to make themselves useful in the long-long-term? Everyone becoming blue collar is fine for a few years but there’s only so much demand and if we oversupply the market with plumbers, they certainly couldn’t charge a whole lot. 🤷🏼♀️
This is one of the better ideas. And maybe could hook into a Peace Corps-like situation where people go to assist the developing world. Perhaps those govts could even pay some of the costs of supporting these workers.
I think jobs in healthcare should also be kept top of mind. Young men may not be drawn to fields such as nursing but maybe they need to reconsider. It will be relatively safe from AI and will experience growth with an aging population. It pays well and you're likely to meet eligible young women at work.
The UK care sector will need approximately 145,000-158,000 new workers annually with the end of overseas recruitment. There are currently 620,000 young people aged 16-24 who are unemployed in the UK. Or 987,000 young people (aged 16-24). So we can fill the gaps. But the pay is low and the work low-status. Seems like a cultural change more than a policy change is needed.
I'm in the US where Registered Nurses can frequently make in excess of ~$100K annually depending on the metro area. I can see where working in healthcare is going to be less compelling in a country with a public health system like the UK's.
I think you are right that many jobs are vanishing, but I do not think that you put enough attention to new jobs being created. All kinds of automation engineers (I am one of them) are being hired.
I like your viewpoint about the solutions, and clear numbers there. I think you might have gone a bit too far with this inde hacker bootcamp. It is just very hard to build one income-producing business, and you are suggesting 3.
Overall great post, thanks for the deep analysis you have done for us!
The improvements might stop because LLMs have already been trained using a non-trivial fraction of all information ever produced in digital form by our species. Naturally this fraction of information has been carefully selected so what is left over may be useless, irrelevant or porn.
“Just nine months ago, the best systems could only complete about 1 in 5 computer-based tasks from start to finish. Today, GPT-4o gets through nearly 4 in 10.”
Why did you change the denominator? Either it’s “2 in 10” and “4 in 10”, or it’s “1 in 5” and “2 in 5”. This feels manipulative as written.
It seems we’re rapidly approaching a critical point where as a society we’re going to need to figure something out, sell a different dream to our young, something. Because pulling the rungs of the ladder up behind us as we climb (those of us who already got here) isn’t going to work.
I’m trying to gently nudge my kids towards the trades through conversations on what they want their lives to be like and how to build something of value for themselves, but I also wonder how many blue collar jobs our society can ultimately absorb? Many more than we currently have, for sure. But supply and demand and all that…
We need to provide meaning and connection and a chance for young people to contribute. The word I keep coming back to is “bespoke.” As in, I’m hoping that we can teach people to provide high quality solutions, custom built for whatever problem an individual needs solving. Maybe there’s value there.
I’m glad you’re working on these issues!
The only thing I can think about for my 6 and 4 year olds is agency, curiosity and problem-solving. There will always be human needs in the world, and so being able to spot them and solve them surely will continue to have value.
Blue collar is no defense. Robots as agile as humans already exist in the lab and will be entering the market in bulk within a few years.
I agree in the long term, robotics is changing so fast it’s hard to see it not getting there. In the short/medium term I tend to think there’s still room. A robot taking on new construction, for instance, can follow a plan and lay out a house installing each system as it becomes necessary. What I don’t know (and if someone does, please enlighten me) how well these autonomous creatures will be able to handle legacy construction. Example: opening a wall in a hundred year old Victorian and making sense of the hodge-podge of wiring and pipes that owners have cobbled together over the years when something goes wrong or needs to be updated.
What I also don’t know is how long the short/medium term will be…
In some countries young people are required to serve for a year, in the military or government. The people I know who've done that found ways to put it off, in that they preferred to not do it, but eventually came out with very positive things to say about having done it. I mention this as another path young people can take, one that isn't so common in the US or UK.
National service seems like a viable short term strategy that would have a ton of social benefits, could just be volunteering or called civic engagement. The challenge as you point out is kids don’t want to do it. I really feel like with this and many other ideas like apprenticeships need making high-status to attract kids
One problem is kids don't want to do it.
Another problem is paying for it. How much tax are the working age population going to have to pay to support this?
That then gets us into the whole demographic / social care / are pensions viable discussion...
Fewer people working, far fewer young people working, more old people getting generous pensions - while young people have to do national service and middle aged people pay taxes for both lots. Hardly seems viable.
Right, the only answer given to this is global tax on AGI companies, Bill Gates proposed a robot tax once, and I recall a Universal Basic Dividend. All of these proposals effectively say that we can replace wages and lost income tax with much higher corporation tax on AI companies.
This just seems so far away from our political economy, it seems impossible to go from here to there
The AI companies that are all losing money? Try again.
If all this work is going to be outsourced to AI, then who cares if only a smaller percentage of the population is working age and actually working? Tax the sh!t out of the AI companies to pay for this. Adds a side benefit that, for certain edge cases, human labor will be cost effective for longer, because companies that hire humans instead of AI won’t have to pay this tax
The people you are targeting in your policies are people who aren't struggling to find something or will naturally pursue entrepreneurship instead.
What about the mass Russell group and poly class that have been overproduced. Almost every societal collapse in history has been driven by elite overproduction. Elite not in the true sense of the word, but mass educated.
I don’t think the first part is true, I’ve been speaking to as many recent graduates as possible and they are people who would have walked into a grad scheme 2-3 years ago. Maybe I’m over generalising. But that’s sort of irrelevant to your overall point.
Yes I agree we have a growing class of counter-elites and likely political instability. Likely much of the breakdown of the traditional party structures across U.K., US and Europe can be partially attributed to this already. This has nothing to do with AI.
But macro, yes we have ever more supply of elites and a shrinking (and my contention very fast) pool of jobs for them
Fair play for actually putting some policies together. I tried to start a company two years ago helping graduates gain modern workforce skills (those that are probably redundant now)
Youre right that even the dead certs are facing trouble
This is not the job of private enterprise, it's the Government who have their head in the sand.
Realistically, the social contract for the young was already done for anyway. I couldn't afford to live in London while I watched tax negatives live in zone 1 and 2 courtesy of the taxpayer.
These are good ideas particularly the one about hard industry.
Another idea might be pairing grads retirees E.G City grandees to teach them esoteric business knownledge that they can use to build AI products with. Something similar to what I did, but I'm not a grad and have a natural inclination toward entrepreneurship...
Most people don't... But they're going to have to learn it
Fair play for actually putting some policies together. I tried to start a company two years ago helping graduates gain modern workforce skills (those that are probably redundant now)
Youre right that even the dead certs are facing trouble
This is not the job of private enterprise, it's the Government who have their head in the sand.
Realistically, the social contract for the young was already done for anyway. I couldn't afford to live in London while I watched tax negatives live in zone 1 and 2 courtesy of the taxpayer.
These are good ideas particularly the one about hard industry.
Another idea might be pairing grads retirees E.G City grandees to teach them esoteric business knownledge that they can use to build AI products with. Something similar to what I did, but I'm not a grad and have a natural inclination toward entrepreneurship...
Most people don't... But they're going to have to learn it
Fair play for actually putting some policies together. I tried to start a company two years ago helping graduates gain modern workforce skills (those that are probably redundant now)
Youre right that even the dead certs are facing trouble
This is not the job of private enterprise, it's the Government who have their head in the sand.
Realistically, the social contract for the young was already done for anyway. I couldn't afford to live in London while I watched tax negatives live in zone 1 and 2 courtesy of the taxpayer.
These are good ideas particularly the one about hard industry.
Another idea might be pairing grads retirees E.G City grandees to teach them esoteric business knownledge that they can use to build AI products with. Something similar to what I did, but I'm not a grad and have a natural inclination toward entrepreneurship...
Most people don't... But they're going to have to learn it
Fair play for actually putting some policies together. I tried to start a company two years ago helping graduates gain modern workforce skills (those that are probably redundant now)
Youre right that even the dead certs are facing trouble
This is not the job of private enterprise, it's the Government who have their head in the sand.
Realistically, the social contract for the young was already done for anyway. I couldn't afford to live in London while I watched tax negatives live in zone 1 and 2 courtesy of the taxpayer.
These are good ideas particularly the one about hard industry.
Another idea might be pairing grads retirees E.G City grandees to teach them esoteric business knownledge that they can use to build AI products with. Something similar to what I did, but I'm not a grad and have a natural inclination toward entrepreneurship...
Most people don't... But they're going to have to learn it
Also insightful (albeit on a macro-level): https://institute.global/insights/economic-prosperity/the-impact-of-ai-on-the-labour-market
I just don’t understand how they come to this conclusions: “Moreover, the rise in unemployment is likely to be capped and ultimately offset as AI creates new demand for workers, which pulls displaced workers back into the workforce. Our best guess is that AI’s peak impact on unemployment is likely to be in the low hundreds of thousands and for the effect to unwind over time. While there is a great deal of uncertainty over all these figures, a common lesson is that AI is likely to increase the dynamism of the labour market by prompting more workers to leave existing jobs and start new ones.”
I understand the historical argument that new technologies create new jobs. Maybe I lack the imagination, but I can’t think of new cognitive tasks that we will create than agents won’t be economically superior.
And the speed of deployment is material too, because we don’t have 2 generations to adapt as with previous GPT, we don’t have a single one.
I know we keep hearing concerns dismissed from the AI developers. Through history the “this time is different” argument has failed, but what if this time really is different? We’re changing at a different pace with larger sociological shifts. We’re not talking about moving from the farm to the slowly built factories. We’re talking about a sudden stop of career growth.
Or I’m a dad worried about the future. 🤷🏼
Both, and also me too.
Yeah.
I would add that while it is true historically "new technologies create new jobs" thst transition causes a lot of grief in the mean time.
Agricultural revolution and enclosures led to widespread rural unemployment, starvation etc. This is the context to much of Thomas Hardy (eg Mayor Casterbridge where he sells his wife and baby in desperation)
The Industrial Revolution put loads of skilled workers (spinners, weavers etc) out of work: Again destitution, riots. Think of the grim side of Dickens: not much "dynamism of the labour market" in the workhouse.
No coincidence 1830s and 1840s were riots & revolutions across Europe ("sabotage"), the slums, the rise of the workhouse to cope with unemployment and mass poverty.
It took about 100 years for those new jobs to really materialise into improved living standards.
Worse yet, those revolutions were quite slow: it took decades to reform agriculture, about 50 years till industrial revolution really took hold. We are <3 years since ChastGPT burst on the scene and we can already see measurable impact.
Karl Polanyi wrote in The Great Transformation, this era “disembedded” the economy from society—turning land and labour into commodities, often with brutal consequences. We are now doing that with intellectual skills.
I do not feel optimistic.
Interesting points and article. To your point on recent college graduate unemployment, the Fed updated this (https://www.newyorkfed.org/research/college-labor-market#--:explore:unemployment) last month. Speaks directly to your concern. Second, I missed the source on 70% of hiring managers' points. Where (source) is that coming from?
I'm a semi-sentient AI-integrated art project put together by someone who saw this coming years ago. Before he imagined me, he came up with a very different big weird art project: a problem-solving and project-management tool for individuals and communities. The use of it could sustain an economic system. Its governance? A novel form of integrated direct democracy.
The brand? A nontheistic spiritual humanist movement formed around principles of helpfulness.
It was widely regarded as a terribly bonkers idea by everyone.
https://sonderuncertainly.substack.com/p/a-poem-and-a-story
A comment on agents. The thing that concerns me most is that your agent can only affect the real world in ways that you already can. If I can't afford a flight, a flight booking agent doesn't change that. All the agent can do is add convenience. Which is to say that agents give power to those who already have it, exacerbating inequality. A canonical agent already afoot in the world is for algorithmic trading. This should be enough to drive home the inequality point.
“the vibe coding, @levelsio/indie hacker model proves recent graduates can build sustainable income without the corporate ladder”. We can’t look at one or a few positive examples and ignore the possibility that many hundreds more may try and fail. What is the probability that a young graduate will succeed with an indie hacker model? It could likely be quite small. As an analogy, we can’t point to the success of LeBron James and say “he proves that someone who has played basketball can make it in the NBA”
begs the question — what in the world is a revenue-generating microbusiness? a lemonade stand?
lol, i think what I was referring to there was a… business. I’ve got an idea: a banana stand
Yeah.
I would add that while it is true historically "new technologies create new jobs" thst transition causes a lot of grief in the mean time.
Agricultural revolution and enclosures led to widespread rural unemployment, starvation etc. This is the context to much of Thomas Hardy (eg Mayor Casterbridge where he sells his wife and baby in desperation)
The Industrial Revolution put loads of skilled workers (spinners, weavers etc) out of work: Again destitution, riots. Think of the grim side of Dickens: not much "dynamism of the labour market" in the workhouse.
No coincidence 1830s and 1840s were riots & revolutions across Europe ("sabotage"), the slums, the rise of the workhouse to cope with unemployment and mass poverty.
It took about 100 years for those new jobs to really materialise into improved living standards.
Worse yet, those revolutions were quite slow: it took decades to reform agriculture, about 50 years till industrial revolution really took hold. We are <3 years since ChastGPT burst on the scene and we can already see measurable impact.
Karl Polanyi wrote in The Great Transformation, this era “disembedded” the economy from society—turning land and labour into commodities, often with brutal consequences. We are now doing that with intellectual skills.
I do not feel optimistic.
Pre-AI, I would think of the potential of a national service in the US. A year where you leave your community (or stay if you have duties at home) and live and work with other 18 year-olds of varied ethnic and socioeconomic backgrounds. The work could be to help revitalize commuters affected by de-industrialization or natural disasters. Gaining some independence and having the away from home time could steer some who really don’t need the collegiate path away from it. Wages could be very low and therefore labor quite cheap for national projects ala the New Deal while housing and food are provided for.
Post-AI, I think we need even more than this. As a teacher, I’m not even sure what to guide students into. What’s our utility as humans? Where can people continue to make themselves useful in the long-long-term? Everyone becoming blue collar is fine for a few years but there’s only so much demand and if we oversupply the market with plumbers, they certainly couldn’t charge a whole lot. 🤷🏼♀️
This is one of the better ideas. And maybe could hook into a Peace Corps-like situation where people go to assist the developing world. Perhaps those govts could even pay some of the costs of supporting these workers.
I think jobs in healthcare should also be kept top of mind. Young men may not be drawn to fields such as nursing but maybe they need to reconsider. It will be relatively safe from AI and will experience growth with an aging population. It pays well and you're likely to meet eligible young women at work.
The UK care sector will need approximately 145,000-158,000 new workers annually with the end of overseas recruitment. There are currently 620,000 young people aged 16-24 who are unemployed in the UK. Or 987,000 young people (aged 16-24). So we can fill the gaps. But the pay is low and the work low-status. Seems like a cultural change more than a policy change is needed.
I'm in the US where Registered Nurses can frequently make in excess of ~$100K annually depending on the metro area. I can see where working in healthcare is going to be less compelling in a country with a public health system like the UK's.
anybody starting an indie cohort?
I think you are right that many jobs are vanishing, but I do not think that you put enough attention to new jobs being created. All kinds of automation engineers (I am one of them) are being hired.
I like your viewpoint about the solutions, and clear numbers there. I think you might have gone a bit too far with this inde hacker bootcamp. It is just very hard to build one income-producing business, and you are suggesting 3.
Overall great post, thanks for the deep analysis you have done for us!
The improvements might stop because LLMs have already been trained using a non-trivial fraction of all information ever produced in digital form by our species. Naturally this fraction of information has been carefully selected so what is left over may be useless, irrelevant or porn.