https://digitaleconomy.stanford.edu/wp-content/uploads/2025/08/Canaries_BrynjolfssonChandarChen.pdf
We find that since the widespread adoption of generative AI, early-career workers (ages 22-25) in the most AI-exposed occupations have experienced a 13 percent relative decline in employment even after controlling for firm-level shocks
I imagine this is largely an effect of hiring freezes while they try to pressure their current staff to make up the difference with LLMs. This would achieve both:
- Reduced labor costs long-term (if it works)
- Generate culture of employees being pressured to leverage LLMs instead of requesting an increase in staffing.
But, if I’m right, that doesn’t mean that the current hiring freeze is actually panning out for these companies. I feel like all managers have essentially been tasked with this for the last 1.5 years.
You must have 5 years experience to apply, all jobs that give you experience are now done poorly by chatbots
Assuming this country survives the next 10 years, what would happen when people retire and they need senior workers?
It’s tragedy of the commons type shit. We’re seeing this big in software engineering firms. Nobody wants to hire juniors anymore because you have to train them, and if you’re the loser company that hires juniors and trains them you’re basically just training juniors to be eaten up by Big Tech as soon as they become seniors, so it’s just a “waste of money.” Of course if everybody thinks this way (as they are) then nobody trains juniors and you have no senior engineers in 15 years. But executives can only think of the next quarter so uh who cares.
to some of us it’s comedy of the commons
Executives want to think of their employees as replaceable cogs that can be abused and replaced to turn a profit. Training is a cost to them and of course your whole logic holds. It is really a contradiction of capitalism. Tragedy of the commons is a capitalist apologetic invention, believe it or not, to justify privatization. Capital needs to maximize profit, it will do so best with a skilled workforce, but it is against the interests of individual capitalists to train workers that they believe will just leave after being trained.
Extending the contradiction, these employers are the same ones that avoid promoting from within or giving raises to retain workers. They think they’re maximizing profits by avoiding raises for 2-5 years but then eat the larger costs of constantly interviewing candidates and paying them even more than those who left. And don’t even think about the dead weight of the bloated management structure, another contradiction intended to discipline labor and validate the capitalists.
That’s the next CEO’s problem
These companies want to never have senior workers. They only keep them on by economic necessity. Senior workers are more expensive, so they really want to “de-skill” their workforce in order to decrease how much they will pay and make it so that they don’t lose as much when firinh anyone for any reason.
When companies fail at this, which is what most of these will do, they reach a profitability crisis and either go bankrupt or get bought out by more liquid monopolies. If the big monopolies run into trouble they get the state to bail them out. Those that survive and try to be productive will need to pay a premium for senior labor and consider increasing levels of training they will fund. But this really depends on how much money they can throw around, which will be not much when this bubble bursts unless the fed goes ham.
Was the study funded by AI firms?
Stanford Digital Economy Lab and Institute for Human-Centered Artificial Intelligence. So effectively yes.
Replacing/substitution implies maintaining productivity, something the authors did not investigate at all. They did not even consider the alternative hypothesis that these are really layoffs - cynical or based on hype that hasn’t realized much substitution for actual labor. It also does not account for businesses failing.
This paper is mostly just, “cum hoc ergo propter hoc”. It points to some other correlations to suggest that workers are losing employment in industries where LLMs are, “more likely” to succeed (a position they never justify).
Lol. Cum hoc.
The AI bubble burst will be K shaped.
K you say???