Currently Online 
sc94597 said:
I very much support more regulation on AI, and high taxes on institutions that automate away jobs to fund expansive unemployment insurance programs which hopefully would evolve into a UBI. Having said that, I don't think AI can be simultaneously an imminent bubble about to bust soon and and lead to mass unemployment (soon). If it leads to mass unemployment (soon) it isn't about to bust, because right there are the real gains to the investment. The AI companies can start increasing prices for their "agents" to replace workers, and as long as the prices are cheaper than the workers without affecting productivity drastically, they'll have consumers. This is also why AI companies have transitioned from scaling base models and fundamental research (Google and Nvidia being the exceptions) to improving the ability of AI-supported tools to perform tasks. I do think in the medium-term (two - five years from now) we're going to experience a pretty enormous AI-induced crash. But it will be an indirect result rather than an AI bubble. This crash won't be because AI companies weren't able to profit, but will be the result of depressed aggregate demand due to low incomes/high unemployment. What will happen is that profits will stay pretty high for a few years until incomes drop so low that nobody can buy many of the products companies are producing. The FED (and other central banks), worried about a deflationary spiral and an economic slump will reduce interest rates to foster lending and hiring while the government goes into stimulus mode, but that isn't enough. Hopefully, if we're lucky, the U.S will elect a center-left leader, taxes will be raised, and relief will come, probably in the form of something like a new public works program since Americans are too Calvinistic for UBI and the infrastructure needs rebuilt anyway. If we're unlucky, the system collapses into a depression and we can't really predict what happens other than whatever does happen will be radical. I think the latter scenario is more likely, to be honest, and if that happens I hope socialists win rather than fascists/neo-feudalists. |
While the 'OR' scenario is just as bad, failure and bubble burts -> recession OR works and millions of jobs lost -> recession (massive reduction in buying power), it can also be both when for example cheap Chinese AI completely obliterates the hopes of charging a lot for American AI agents.
The latter is the real fear for the bubble bursting, AI is in a race who gets there first. The entire US economy seems to be betting on betting trillions of dollars into being the first to get 'job replacing' AI.
And just like blue collar jobs were outsourced to cheap labor countries, AI can just as easily be outsourced to cheaper systems... Quality nor 'loyalty' was never a priority for capitalism, profits are, so AI jobs will also go to the cheapest bidder, and that shift is a lot easier than building factories in China and offices in India.
So if AI succeeds in taking over jobs, cheaper AI can just as well succeed in taking over the US AI made at enormous investments. It's like building the first ultra expensive car manufacturing factory in Detroit, just to be replaced a year later or even before opening by overseas production. No time to recoup the investments.
https://www.chathamhouse.org/2025/11/low-cost-chinese-ai-models-forge-ahead-even-us-raising-risks-us-ai-bubble
Anyway the US economy is fucked either way :/ Bubble burst, vast unemployment surge or both.
But temporarily profits can go up if AI works to replace those jobs before cheap AI... Until common people run out of money. Then yep, either deflation, UBI or massive unrest. And millions of people sitting at home all day fuming on the internet isn't going to go well either :/
We live in interesting times :(
