Really interesting. Middle class families have advantages getting their kids into Uni for largely academic courses that may now prove a career dead-end. If the best route shifts towards vocational training, will they still retain an advantage in getting their kids into such courses, displacing kids from more typically blue-collar families?
Nando that is a great question. Would imply a completely different way though of child-rearing. I think though this happened in the US in the 1960s/1970s though with the expansion of higher education
I think you underestimate the effect of genetics and overestimate the effect of child-rearing.
Allocation to skiilled work be it blue collar or white collar is highly correlated with a genetic ancestory in that work. There is a reason people with the surname Smith have levels of physical strength above the mean of the population to this day.
Higher education did expand but most white collar work before the expansion was not connected to degrees. The vast majority of people before the 60s became accounts or solicitors as articled clarks, a kind of white collar apprenticeship. This is what my great grandfather did in the 1920s and grandfather did in the 1950s to work as an accountant. The only graduate jobs were doctors, barristers and clergymen.
Before the industrial revolution the vast majority of the population was doing relatively menial and unskilled agriculural work that didn't demand especially difficult levels of dexterity or IQ. Most who moved into imdustrial cities in the 19th century started in unskilled physical labour (of the sort that has been heavily automated now). Over time those with the aptitudes to escape either went into the upper-working class skilled labour or into the growing ranks of lower middle class clerks in the late 19th century. Accountancy was by the early 20th century a boom industry which opened up middle class status to those with the abilities.
So if automation does take off the way you describe, just as it was unrealistic to imagine a few years ago that anyone had the mental abilities to become a programmer, a vast number of the population almost certainlty lack the physical and coordinational aptitude to do highly physical jobs too. That was why even when middle class jobs were less common most of the population found themselves in unskilled jobs.
So a great wedge of the population will have absolutely no means of contrubuting which is where parties promoting UBI will become prominent, opposed by parties voted for by the now culturally powerful bluw collar workers who will resent the taxation this implies.
Interesting position piece but have you got this the wrong way round? High vulnerability to AI could also be read as high productivity potential. Late 19th Century factory workers operating of machinery were poorly paid, but they were markedly better off than the hand-to-mouth penury of a farm labourer or cottage-side piece worker.
So it is here. Where previously a lawyer or accountant might be require a long process of research, review, and repetition to excel in an activity, now their role becomes one of defining the objective, drafting the heads of terms, and standing back. Five people at £100,000 a year become one at £150,000.
Clearly, just as with the Industrial Revolution, the majority of the surplus of production will reside with the possessors of capital. But for those who remain lawyering or accounting, their job becomes better - and better paid.
Super interesting James. Hard to escape feeling we are heady for an almighty smash as multiple waves start cresting in Britain across demographics, society, and the economy. An interesting additional wrinkle on your piece is that - amongst a sub set of white collar professions - there is a lot of hostility to AI, correlated with political identification (anti-Trump, anti-Silicon Valley tech). I think we might see quite active lobbying against some of the AI rollout in some fields (particularly the public sector but you can also see some of this in the creative/media industry). This might create winners and losers and feed into other narratives about resentment.
Hi James, great blog. Can you share the research from this bit: "The Government’s own analysis suggests that the top 5 safest jobs" I can't find it. I'm working in construction so it's very interesting to me.
Thank you. I work in construction and we're always worried about labour supply. Looks like the least exposed jobs are almost all construction. It could be the net beneficiary of this disruption. I'm screwed mind as my job is number 6 on the most exposed list.
I've been thinking about this exact topic a lot recently, so thanks for the detailed analysis. Really interesting read.
As someone has said in the comments, I could see a populist party standing on an anti-AI platform, to pick up this new constituency. Could be the Greens/a new environmental leftist party, as you have the dual narratives of the environmental impact of LLMs and the job losses they are responsible for.
You could then have a party like Reform standing in opposition to this, pushing a narrative along the lines of "none of you gave a shit when this happened to manufacturing in the 80s"
The Tories will be in a bind, if there are mass job losses among the credentialed middle classes in the home counties. Do they tack interventionist/redistributive? Or do they go all in on the Merchant Right?
Moreover, the conceptual underpinning of traditional leftist redistributive thought, from Marx onwards, is kind of done for. We're moving from a point where "workers are being exploited by capitalists, and we want a fairer share for workers, or else we'll withold our labour and disrupt the means of production" towards "we don't have any work and can't disrupt production, so we need the government to do something about this".
Considering how much of a hard sell traditional leftist policies have been, this new paradigm could be even harder to sell ellectorially, if there's not a large majority of the electorate impact by AI job losses.
The worst scenario could be a large minority of workers displaced by AI, as it'll have a massive impact on people's lives and society, but not enough where a political majority could be formed to address it.
I can't "see a populist party standing on an anti-AI platform" because I can't imagine that an anti-AI platform could be coherent. AI might replace taxi drivers, and taxi drivers won't like that, but OTOH, few people see taxi-driving as a vocation, and it may mean cheaper, more reliable taxis for everyone else. I can see people hating AI taking THEIR job, but not being especially bothered (and even pleased) when it takes somebody else's.
What would an anti-AI party do? How can you ban it? There would be dark web sites where you could download AI, or special pleading by large businesses (if they didn't simply off-shore).
(I considered writing a novel in the 1990s, set around now that had a "Luddite Party" but I didn't work out where I was going with it.)
One thing that messes up all predictions is that degrees just don't seem to signify much education or ability any more. I saw someone today complaining on Twitter than recent graduates couldn't do basic tasks. Whether this is just curmudgeonliness or not, I can't say, but it's a sentiment I keep coming across. I certainly have the impression that large employers no longer rate non-specialist degrees, and that universities haven't been doing whatever it is they did from around Darwin's time for most of this century. In short, I think some of this "flip" would happen even without AI. Student debt is too much given the uncertainty of secure employment for all but the brightest.
Finally, I can't see UBI taking off until we can decide who is a citizen (and should receive it) and who isn't. One thing I think is sure: come here illegally, and you won't get it.
It remains to be seen. Maga is basically a push back against globalization, with loads of other greviences mixed in, most of which are related to the culture of the "winners" of globalization; the liberal elite.
I could something similar in how the push back against AI manifests itself politically, but we just can't see what form it'll take yet.
And yeah, I had similar thoughts to you in terms of graduates. The article was written to imply all/most uni graduates are middle class swots going straight into accounting, medicine, law, etc. That paradigm of uni graduates hasn't existed for a few decades now. The minority of Russell Group grads yes, the bulk now made up of ex-poly students, no.
I'd wager a majority of grads are now those coming out of ex-polies with very average abilities, literally, if half of under 25s are now going to uni, then we have a lot of average people coming out the other end. There was a belief during the Blair years that this will create the knowledge workers of the future, but that never really happened. These grads didn't have a great future already, before AI came along. It's just that now the brighter/more middle class Russell Group cohorts are feeling the same pinch.
The future of redistribution is key here, because the last 40 years or so have leant heavily on a comfortably-off and progressive/liberal middle-class to redistribute _from_.
When instead the burden of tax falls on the merchant right - which it will, there are always so many more people on £80k than £300k that even if you took the latter for every penny, there just aren't the numbers - that's a whole different ball-game.
As someone firmly in the liberal category - comfortably off but not wealthy, card-carrying progressive - I don't mind paying relatively high and redistributive tax rates as long as the system feels reasonably balanced: it resonates with the culture of charitable works and noblesse oblige that has been a strand of middle-class culture for pretty much as long as the middle class has existed.
But also, being an Anywhere, I don't care an enormous amount whether those taxes are spent on helping needy indigenous Brits, Somali refugees in the UK, or as foreign aid to Somalia itself. I want the visible stuff in my present-tense Somewhere addressed, and addressed well - street homelessness, child poverty - and good, universal public services: functioning hospitals and schools, clean trains and buses, and enough visible police to maintain a general sense of order - but after that I don't particularly see one individual person/group as more deserving of aid or support than another. Let the state figure it out based on evidence.
Now consider the merchant right. Basically "Deano" twenty years on, he's worked bloody hard and done well for himself. Guys like that, and I know quite a few, have a fairly wide (and sometimes volatile) range of opinions on appropriate levels of tax and spending - as indeed do Reform. From "I'll keep all I can, I earned it" to "I've done OK, and it's fair to support people who've had a hard life". "Tax the Rich", but that categorically doesn't mean blokes with £1M houses wanting to pass it on to their kids. More importantly though, they have much stronger views than the liberal Anywheres on who is, and isn't, deserving of their help. It's not an approach that's necessarily supported by public-health evidence, but then "common sense" rarely is.
(Side note - Deano is usually identified as white-British or maybe Irish heritage, but IME that's by no means a given; as a white-liberal I'm still at times confounded by the fact that second/third generation immigrants, lacking any inherited sense of Empire Guilt, have often stronger anti-immigrant and anti-foreign-aid views than would be normal or acceptable for indigenous Brits of the same social class / education level. Just look at the Tories' last couple of Home Secretaries.)
This is a great read. I think that concept of the 'merchant right' will emerge as a key dynamic for reform.
There are a large base of people outside of London who don't fit into the left-behind category, nor are they an 'everywhere' liberal-type. Instead I see them as:
- Financially successful or at least secure, but through blue-collar or non-intellectual means - not defined at all by level of education. Often self-employed or work for family businesses, often first-time homebuyers and doing ok
- Right-leaning instincts (patriotic, rooted in family and pride in community, belief in decency and fairness but also self-governance)
- pay high taxes, suffer from increasing bills, increasing difficulty to run businesses
- Not overly political but despise wokery, liberal moaning, government overreach, mass immigration
Farage's messaging of fairness, straight-talking politics etc does very well with this group and they will become increasingly important as the dynamics you talk about come to pass
All very interesting but it is way too early to know what the effect of LLM's on white collar jobs will be. Of course lots of journalists write articles saying that, but that's because it gets clicks. Note that the article you cite doesn't give any actual evidence that the downturn in accounting hiring is due to AI, and that's because there's a more prosaic but less sexy explanation: there's just been a downturn in demand for advisory services, as there is from time to time.
«“Management consultants and business analysts” are listed as the job occupation most vulnerable to automation, and “chartered and certified accountants” the third highest»
Since this conclusion is improbable, the whole "methodology" is rather questionable: the jobs of management consultants and chartered accountants is to design schemes to enrich the purchasing executive or make up and endorse with the "authority" of a third party whatever numbers and conclusions the purchasing executive wants and no LLM etc. can do that. Perhaps the LLM can be used to churn out the boilerplate in which the purchased schemes and numbers and conclusions are wrapped, so interns and graduate employees may be replaced with LLMs, but not further than that.
What LLMs may do is to make more valuable the jobs of people whose main value to their employer is complicity and less valuable the jobs of the rest.
Great analysis, agreed with much. Though I think one aspect that will "protect" a majority of the white collar professional class is the family wealth and inheritance built up by Boomer parents in the previous cycle ... it's probably a less precarious threat than that faced by Blue collar workers in the 70s and 80s
Strong parallels here with the mid-19th century upheavals documented by Christopher Clark in his book Revolutionary Spring. This reminds me of the silk weavers in Lyon. Skilled, respected, and economically secure - until mechanisation crushed them. The AI wave risks doing the same to today’s cognitive middle class. And history tells us: elite overproduction is a key ingredient of revolution.
Nice piece, I'm writing on the topic soon. Due to mass amount of CAPEX and potential >20% profit increases, I expect a new AI-powered middle class. Their behaviour, as they seek to build their capital and have better lives, will put political pressure on health services and housing as demand grows. Logical conclusion, as it stands: gated communities, further sprawl of London. Can MCR build-out in the UK?
Really interesting. Middle class families have advantages getting their kids into Uni for largely academic courses that may now prove a career dead-end. If the best route shifts towards vocational training, will they still retain an advantage in getting their kids into such courses, displacing kids from more typically blue-collar families?
Nando that is a great question. Would imply a completely different way though of child-rearing. I think though this happened in the US in the 1960s/1970s though with the expansion of higher education
I think you underestimate the effect of genetics and overestimate the effect of child-rearing.
Allocation to skiilled work be it blue collar or white collar is highly correlated with a genetic ancestory in that work. There is a reason people with the surname Smith have levels of physical strength above the mean of the population to this day.
Higher education did expand but most white collar work before the expansion was not connected to degrees. The vast majority of people before the 60s became accounts or solicitors as articled clarks, a kind of white collar apprenticeship. This is what my great grandfather did in the 1920s and grandfather did in the 1950s to work as an accountant. The only graduate jobs were doctors, barristers and clergymen.
Before the industrial revolution the vast majority of the population was doing relatively menial and unskilled agriculural work that didn't demand especially difficult levels of dexterity or IQ. Most who moved into imdustrial cities in the 19th century started in unskilled physical labour (of the sort that has been heavily automated now). Over time those with the aptitudes to escape either went into the upper-working class skilled labour or into the growing ranks of lower middle class clerks in the late 19th century. Accountancy was by the early 20th century a boom industry which opened up middle class status to those with the abilities.
So if automation does take off the way you describe, just as it was unrealistic to imagine a few years ago that anyone had the mental abilities to become a programmer, a vast number of the population almost certainlty lack the physical and coordinational aptitude to do highly physical jobs too. That was why even when middle class jobs were less common most of the population found themselves in unskilled jobs.
So a great wedge of the population will have absolutely no means of contrubuting which is where parties promoting UBI will become prominent, opposed by parties voted for by the now culturally powerful bluw collar workers who will resent the taxation this implies.
Interesting position piece but have you got this the wrong way round? High vulnerability to AI could also be read as high productivity potential. Late 19th Century factory workers operating of machinery were poorly paid, but they were markedly better off than the hand-to-mouth penury of a farm labourer or cottage-side piece worker.
So it is here. Where previously a lawyer or accountant might be require a long process of research, review, and repetition to excel in an activity, now their role becomes one of defining the objective, drafting the heads of terms, and standing back. Five people at £100,000 a year become one at £150,000.
Clearly, just as with the Industrial Revolution, the majority of the surplus of production will reside with the possessors of capital. But for those who remain lawyering or accounting, their job becomes better - and better paid.
Amazing article, feels like our political parties and narrative are miles behind this.
Appreciated
Super interesting James. Hard to escape feeling we are heady for an almighty smash as multiple waves start cresting in Britain across demographics, society, and the economy. An interesting additional wrinkle on your piece is that - amongst a sub set of white collar professions - there is a lot of hostility to AI, correlated with political identification (anti-Trump, anti-Silicon Valley tech). I think we might see quite active lobbying against some of the AI rollout in some fields (particularly the public sector but you can also see some of this in the creative/media industry). This might create winners and losers and feed into other narratives about resentment.
My friend Ed Newton-Rex has suggested something very similar - with explictly anti AI parties pitching...
Hi James, great blog. Can you share the research from this bit: "The Government’s own analysis suggests that the top 5 safest jobs" I can't find it. I'm working in construction so it's very interesting to me.
https://www.gov.uk/government/publications/the-impact-of-ai-on-uk-jobs-and-training
Thank you. I work in construction and we're always worried about labour supply. Looks like the least exposed jobs are almost all construction. It could be the net beneficiary of this disruption. I'm screwed mind as my job is number 6 on the most exposed list.
I've been thinking about this exact topic a lot recently, so thanks for the detailed analysis. Really interesting read.
As someone has said in the comments, I could see a populist party standing on an anti-AI platform, to pick up this new constituency. Could be the Greens/a new environmental leftist party, as you have the dual narratives of the environmental impact of LLMs and the job losses they are responsible for.
You could then have a party like Reform standing in opposition to this, pushing a narrative along the lines of "none of you gave a shit when this happened to manufacturing in the 80s"
The Tories will be in a bind, if there are mass job losses among the credentialed middle classes in the home counties. Do they tack interventionist/redistributive? Or do they go all in on the Merchant Right?
Moreover, the conceptual underpinning of traditional leftist redistributive thought, from Marx onwards, is kind of done for. We're moving from a point where "workers are being exploited by capitalists, and we want a fairer share for workers, or else we'll withold our labour and disrupt the means of production" towards "we don't have any work and can't disrupt production, so we need the government to do something about this".
Considering how much of a hard sell traditional leftist policies have been, this new paradigm could be even harder to sell ellectorially, if there's not a large majority of the electorate impact by AI job losses.
The worst scenario could be a large minority of workers displaced by AI, as it'll have a massive impact on people's lives and society, but not enough where a political majority could be formed to address it.
I can't "see a populist party standing on an anti-AI platform" because I can't imagine that an anti-AI platform could be coherent. AI might replace taxi drivers, and taxi drivers won't like that, but OTOH, few people see taxi-driving as a vocation, and it may mean cheaper, more reliable taxis for everyone else. I can see people hating AI taking THEIR job, but not being especially bothered (and even pleased) when it takes somebody else's.
What would an anti-AI party do? How can you ban it? There would be dark web sites where you could download AI, or special pleading by large businesses (if they didn't simply off-shore).
(I considered writing a novel in the 1990s, set around now that had a "Luddite Party" but I didn't work out where I was going with it.)
One thing that messes up all predictions is that degrees just don't seem to signify much education or ability any more. I saw someone today complaining on Twitter than recent graduates couldn't do basic tasks. Whether this is just curmudgeonliness or not, I can't say, but it's a sentiment I keep coming across. I certainly have the impression that large employers no longer rate non-specialist degrees, and that universities haven't been doing whatever it is they did from around Darwin's time for most of this century. In short, I think some of this "flip" would happen even without AI. Student debt is too much given the uncertainty of secure employment for all but the brightest.
Finally, I can't see UBI taking off until we can decide who is a citizen (and should receive it) and who isn't. One thing I think is sure: come here illegally, and you won't get it.
It remains to be seen. Maga is basically a push back against globalization, with loads of other greviences mixed in, most of which are related to the culture of the "winners" of globalization; the liberal elite.
I could something similar in how the push back against AI manifests itself politically, but we just can't see what form it'll take yet.
And yeah, I had similar thoughts to you in terms of graduates. The article was written to imply all/most uni graduates are middle class swots going straight into accounting, medicine, law, etc. That paradigm of uni graduates hasn't existed for a few decades now. The minority of Russell Group grads yes, the bulk now made up of ex-poly students, no.
I'd wager a majority of grads are now those coming out of ex-polies with very average abilities, literally, if half of under 25s are now going to uni, then we have a lot of average people coming out the other end. There was a belief during the Blair years that this will create the knowledge workers of the future, but that never really happened. These grads didn't have a great future already, before AI came along. It's just that now the brighter/more middle class Russell Group cohorts are feeling the same pinch.
The future of redistribution is key here, because the last 40 years or so have leant heavily on a comfortably-off and progressive/liberal middle-class to redistribute _from_.
When instead the burden of tax falls on the merchant right - which it will, there are always so many more people on £80k than £300k that even if you took the latter for every penny, there just aren't the numbers - that's a whole different ball-game.
As someone firmly in the liberal category - comfortably off but not wealthy, card-carrying progressive - I don't mind paying relatively high and redistributive tax rates as long as the system feels reasonably balanced: it resonates with the culture of charitable works and noblesse oblige that has been a strand of middle-class culture for pretty much as long as the middle class has existed.
But also, being an Anywhere, I don't care an enormous amount whether those taxes are spent on helping needy indigenous Brits, Somali refugees in the UK, or as foreign aid to Somalia itself. I want the visible stuff in my present-tense Somewhere addressed, and addressed well - street homelessness, child poverty - and good, universal public services: functioning hospitals and schools, clean trains and buses, and enough visible police to maintain a general sense of order - but after that I don't particularly see one individual person/group as more deserving of aid or support than another. Let the state figure it out based on evidence.
Now consider the merchant right. Basically "Deano" twenty years on, he's worked bloody hard and done well for himself. Guys like that, and I know quite a few, have a fairly wide (and sometimes volatile) range of opinions on appropriate levels of tax and spending - as indeed do Reform. From "I'll keep all I can, I earned it" to "I've done OK, and it's fair to support people who've had a hard life". "Tax the Rich", but that categorically doesn't mean blokes with £1M houses wanting to pass it on to their kids. More importantly though, they have much stronger views than the liberal Anywheres on who is, and isn't, deserving of their help. It's not an approach that's necessarily supported by public-health evidence, but then "common sense" rarely is.
(Side note - Deano is usually identified as white-British or maybe Irish heritage, but IME that's by no means a given; as a white-liberal I'm still at times confounded by the fact that second/third generation immigrants, lacking any inherited sense of Empire Guilt, have often stronger anti-immigrant and anti-foreign-aid views than would be normal or acceptable for indigenous Brits of the same social class / education level. Just look at the Tories' last couple of Home Secretaries.)
This is a great read. I think that concept of the 'merchant right' will emerge as a key dynamic for reform.
There are a large base of people outside of London who don't fit into the left-behind category, nor are they an 'everywhere' liberal-type. Instead I see them as:
- Financially successful or at least secure, but through blue-collar or non-intellectual means - not defined at all by level of education. Often self-employed or work for family businesses, often first-time homebuyers and doing ok
- Right-leaning instincts (patriotic, rooted in family and pride in community, belief in decency and fairness but also self-governance)
- pay high taxes, suffer from increasing bills, increasing difficulty to run businesses
- Not overly political but despise wokery, liberal moaning, government overreach, mass immigration
Farage's messaging of fairness, straight-talking politics etc does very well with this group and they will become increasingly important as the dynamics you talk about come to pass
All very interesting but it is way too early to know what the effect of LLM's on white collar jobs will be. Of course lots of journalists write articles saying that, but that's because it gets clicks. Note that the article you cite doesn't give any actual evidence that the downturn in accounting hiring is due to AI, and that's because there's a more prosaic but less sexy explanation: there's just been a downturn in demand for advisory services, as there is from time to time.
«“Management consultants and business analysts” are listed as the job occupation most vulnerable to automation, and “chartered and certified accountants” the third highest»
Since this conclusion is improbable, the whole "methodology" is rather questionable: the jobs of management consultants and chartered accountants is to design schemes to enrich the purchasing executive or make up and endorse with the "authority" of a third party whatever numbers and conclusions the purchasing executive wants and no LLM etc. can do that. Perhaps the LLM can be used to churn out the boilerplate in which the purchased schemes and numbers and conclusions are wrapped, so interns and graduate employees may be replaced with LLMs, but not further than that.
What LLMs may do is to make more valuable the jobs of people whose main value to their employer is complicity and less valuable the jobs of the rest.
Great analysis, agreed with much. Though I think one aspect that will "protect" a majority of the white collar professional class is the family wealth and inheritance built up by Boomer parents in the previous cycle ... it's probably a less precarious threat than that faced by Blue collar workers in the 70s and 80s
yes really good points Alexander. I'm going write more about the wealth, income, education, interaction soon which might cover this
Strong parallels here with the mid-19th century upheavals documented by Christopher Clark in his book Revolutionary Spring. This reminds me of the silk weavers in Lyon. Skilled, respected, and economically secure - until mechanisation crushed them. The AI wave risks doing the same to today’s cognitive middle class. And history tells us: elite overproduction is a key ingredient of revolution.
The 'Footing the Bill' graph needs a comparable with the Y-axis instead of population being total income per decile.
Nice piece, I'm writing on the topic soon. Due to mass amount of CAPEX and potential >20% profit increases, I expect a new AI-powered middle class. Their behaviour, as they seek to build their capital and have better lives, will put political pressure on health services and housing as demand grows. Logical conclusion, as it stands: gated communities, further sprawl of London. Can MCR build-out in the UK?
I can definitely see an "augmented" segment winning out. But those gains might be time limited? Hard to know precisely at this point