This article touches on an important point of system changes needed in the age of AI, but it is built on a few false arguments/ideas. I think that:
1. AI will not replace jobs in all sectors at once - it is not doing that, and it won't.
2. We cannot make people happy by introducing a China-like regime of political person scoring.
3. Forced token and subscription house renting systems will be met with enormous pressure, because most people own their homes (at least partially), and we cannot build homes that fast.
So I think that the idea behind the post was right, but it disregards what people want, and it takes a very radical AI future viewpoint.
1. You’re right - AI won’t replace all jobs at once. That’s precisely what makes it dangerous. It’s not a clean break, it’s a rolling cascade - sector by sector, function by function, quietly unravelling the labour assumptions that underpin mortgage viability. The damage won’t come from total replacement. It’ll come from gradual collapse that no one is structurally prepared for.
2. I’m not endorsing it civic scoring - I’m warning that in a world of unaffordable housing and post-labour triage, this is the kind of system governments might reach for. Not because it’s right. But because it’s administratively tempting.
3. As for token-based housing models - yes, there will be pushback. Especially from current owners. But the article isn’t predicting next year’s legislation. It’s sketching plausible trajectories when 30–50% of the workforce becomes economically redundant. At that point, we won’t be discussing “what people want.” We’ll be navigating what failing systems can still deliver.
The point of the article isn’t to make an ironclad prediction. It’s a pressure test. And if it sounds radical, that probably says more about how fragile the current system already is than anything else.
Only about a quarter of the houses in the UK have a mortgage on them by the occupants. Also the majority of jobs will survive AI because they don’t involve computers.
The fact that only 28% of homes are mortgaged doesn’t mean the rest of the market is immune.
Mortgaged households set the tempo. They’re the ones exposed to interest rates, wage shocks, and employment disruption. When they crack, prices move. And when prices move, the whole ladder shifts - whether you’ve got a mortgage or not.
Also worth noting - AI isn’t just replacing screen-based jobs. It’s increasingly merging with robotics, computer vision, and sensor tech to automate physical roles too. Although we’re probably a little further out with this type of disruption.
However, your warehouse workers, delivery drivers, factory workers - all redundant the moment autonomous robots get regulatory clearance.
Shelf-stacker, cleaner, coffee shop worker, farmhand - if the task is repetitive and the environment can be mapped, then I fear it’s a matter of when, not if.
The people you mention in your last paragraph probably couldn’t get a mortgage today.
Also, if there are mass defaults, would lenders really repossess? They would only do so if they could find someone who could buy (and who might need a mortgage). Lenders won’t want to own a load of empty houses that would simply degrade.
You’re right that many of those roles might not qualify for a mortgage today. But most were considered perfectly creditworthy not that long ago.
And that’s kind of the point.
AI isn’t just threatening future mortgages. - it’s potentially pulling the rug from under people who already have one. The problem isn’t who gets a mortgage next. It’s who suddenly can’t afford the one they’ve already got.
And yes, lenders don’t want to (or may not be physically able to) repossess en masse. But that doesn’t mean they won’t. If defaults spike and asset values fall, the logic would shift from growth to damage control. Repossession becomes the least-bad option - not because it’s desirable, but because letting the loan book rot is worse.
What may happen next is corporates / private equity step in, buy up distressed stock, and lease it back to the state.
Homes may become yield-bearing welfare infrastructure - with private capital owning the asset. Then the remaining high-end taxpayer covers the rent via UBI subsidies.
along with “Emulated Minds, based on real people, serving/working for whoever owns them”, the idea of “what happens in any sort of mass reshuffling of upper-middle living circumstances in the face of Synthetic Labor” has always been the most interesting question about any AI/Singulitarian-tinged conversation.
and also, similarly to Emulated Synthetic Labor/Minds, the idea of “Digitized Social Credit Scoring tied to meaningful rights and privs” has long felt inevitably *possible* and thus worth considering whether or not you consider it solvents/plausible.
encouraged to see somebody else willing to mention “blockchain tied to virtue” as a way to off-set pure Financial Capital as the basis of rights/privs, there are a lot of Communitarian potential frameworks who havent fully grappled with how those technologies could apply to freedom of labor/to personhood but would do well to.
lastly—and forgive the longass response (this was an excellent article and i say that with 20+ years of considering these issues)—but lastly, given all the above and it’s implications, the absolute unwillingness to take seriously all this on the part of otherwise generally sharp Leftists like Freddie DeBoer and Hamilton Nolan leaves me angry and disappointed.
This article touches on an important point of system changes needed in the age of AI, but it is built on a few false arguments/ideas. I think that:
1. AI will not replace jobs in all sectors at once - it is not doing that, and it won't.
2. We cannot make people happy by introducing a China-like regime of political person scoring.
3. Forced token and subscription house renting systems will be met with enormous pressure, because most people own their homes (at least partially), and we cannot build homes that fast.
So I think that the idea behind the post was right, but it disregards what people want, and it takes a very radical AI future viewpoint.
Thanks for the comment.
A few quick thoughts in response to you points:
1. You’re right - AI won’t replace all jobs at once. That’s precisely what makes it dangerous. It’s not a clean break, it’s a rolling cascade - sector by sector, function by function, quietly unravelling the labour assumptions that underpin mortgage viability. The damage won’t come from total replacement. It’ll come from gradual collapse that no one is structurally prepared for.
2. I’m not endorsing it civic scoring - I’m warning that in a world of unaffordable housing and post-labour triage, this is the kind of system governments might reach for. Not because it’s right. But because it’s administratively tempting.
3. As for token-based housing models - yes, there will be pushback. Especially from current owners. But the article isn’t predicting next year’s legislation. It’s sketching plausible trajectories when 30–50% of the workforce becomes economically redundant. At that point, we won’t be discussing “what people want.” We’ll be navigating what failing systems can still deliver.
The point of the article isn’t to make an ironclad prediction. It’s a pressure test. And if it sounds radical, that probably says more about how fragile the current system already is than anything else.
Only about a quarter of the houses in the UK have a mortgage on them by the occupants. Also the majority of jobs will survive AI because they don’t involve computers.
Appreciate the comment.
The fact that only 28% of homes are mortgaged doesn’t mean the rest of the market is immune.
Mortgaged households set the tempo. They’re the ones exposed to interest rates, wage shocks, and employment disruption. When they crack, prices move. And when prices move, the whole ladder shifts - whether you’ve got a mortgage or not.
Also worth noting - AI isn’t just replacing screen-based jobs. It’s increasingly merging with robotics, computer vision, and sensor tech to automate physical roles too. Although we’re probably a little further out with this type of disruption.
However, your warehouse workers, delivery drivers, factory workers - all redundant the moment autonomous robots get regulatory clearance.
Shelf-stacker, cleaner, coffee shop worker, farmhand - if the task is repetitive and the environment can be mapped, then I fear it’s a matter of when, not if.
The people you mention in your last paragraph probably couldn’t get a mortgage today.
Also, if there are mass defaults, would lenders really repossess? They would only do so if they could find someone who could buy (and who might need a mortgage). Lenders won’t want to own a load of empty houses that would simply degrade.
You’re right that many of those roles might not qualify for a mortgage today. But most were considered perfectly creditworthy not that long ago.
And that’s kind of the point.
AI isn’t just threatening future mortgages. - it’s potentially pulling the rug from under people who already have one. The problem isn’t who gets a mortgage next. It’s who suddenly can’t afford the one they’ve already got.
And yes, lenders don’t want to (or may not be physically able to) repossess en masse. But that doesn’t mean they won’t. If defaults spike and asset values fall, the logic would shift from growth to damage control. Repossession becomes the least-bad option - not because it’s desirable, but because letting the loan book rot is worse.
What may happen next is corporates / private equity step in, buy up distressed stock, and lease it back to the state.
Homes may become yield-bearing welfare infrastructure - with private capital owning the asset. Then the remaining high-end taxpayer covers the rent via UBI subsidies.
I think a lot of people might think that’s a good outcome.
along with “Emulated Minds, based on real people, serving/working for whoever owns them”, the idea of “what happens in any sort of mass reshuffling of upper-middle living circumstances in the face of Synthetic Labor” has always been the most interesting question about any AI/Singulitarian-tinged conversation.
and also, similarly to Emulated Synthetic Labor/Minds, the idea of “Digitized Social Credit Scoring tied to meaningful rights and privs” has long felt inevitably *possible* and thus worth considering whether or not you consider it solvents/plausible.
encouraged to see somebody else willing to mention “blockchain tied to virtue” as a way to off-set pure Financial Capital as the basis of rights/privs, there are a lot of Communitarian potential frameworks who havent fully grappled with how those technologies could apply to freedom of labor/to personhood but would do well to.
lastly—and forgive the longass response (this was an excellent article and i say that with 20+ years of considering these issues)—but lastly, given all the above and it’s implications, the absolute unwillingness to take seriously all this on the part of otherwise generally sharp Leftists like Freddie DeBoer and Hamilton Nolan leaves me angry and disappointed.