31 comments

[ 2.8 ms ] story [ 44.7 ms ] thread
(comment deleted)
My hot take: We’re already in a recession.

We just use AI to justify doing all the same recession behaviors while making it sound like innovation

- not hiring this year > AI is making us productive!

- no wage growth > AI means we don’t need to raise salaries

- layoffs > with AI we can do more with less people

- spending less on offsites, work perks etc > we really need that budget for AI

- not spending money on that new business tool > AI can do it instead

Also RTO mandates as a way to cover up for layoffs.
Q3 GDP was +4.3% We will know Q4 in a bit, but the projection is +2.7%

Hardly a recession

We got rid of a lot of contractors at the end of year. We also aren't hiring anyone new and only replacing some people that have left.
It makes sense that you wouldn't hire in such an uncertain environment. We have a President using emergency powers to affect sweeping, unpredictable, consequential changes to the economy that can dramatically alter unit economics overnight and completely tank a previously viable business. Within this calendar year, the President's ability to do this may be upended by pending court cases, an election, or both. Following those potential changes, the breach of trust created by the previous chaos may mean that trade never returns to normal. I don't envy anyone trying to make long-term business decisions, like hiring, in such an environment.
I was just telling someone the other day how all the talk of civil wars and unrest were overblown, Americans will never resort to that with the economy the way it is. Even in the '08 recession (and even worse scenarios), recovery was a possibility. There would be no jobs at the time, but you can reasonably expect them to be created in a few years. You could go to school and switch careers.

But I think right now there is a dreadful perfect storm of sorts.

- It's not just LLMs, the social/political environment just doesn't favor risk taking, which means less opportunities.

- The US will be alienated from all its good trade partners for a decade plus at best. They'll still do trade, but the era of relying on US companies, or relying on a stable US consumer base is over.

- The dollar will decline, by how much I don't know but it will. Less buying power for American companies.

- Education is in shambles, and skilled & educated immigrants who can leave the US are doing so in droves. Brain drain will be real, the pipeline to replace them will take a generation, and that is if it was fixed today.

- Historically, there is a natural re-balancing of powers that happens as a result of people getting upset and organizing change or some sort. But the ability of the population to organize meaningfully is curtailed because tech, moderation and surveillance capitalism.

- I won't say too much about the current admin, but things are really scary. Not as in "i'm upset about this" or "so much for democracy" but more like "i'm scared for my life" levels of scary.

- Erosion of trust is huge, you don't take risks if you don't have some trust. consumer spending, loans, business spending,etc... and the erosion of trust is fundamental and hard to repair. Trust is also heavily asymmetric, it costs a lot to obtain, but it takes little to lose it. Once lost, gaining it again is usually orders of magnitude more costly.

let me stop there for brevity, but my point was how the US has never truly been in a situation where the economy is doing bad, and the politics is untenable. You have people who are in power and well incentivized to make things even worse, you can cancel elections and deploy troops better if things are really and truly "bad". When people stop having their basic needs met (and I don't mean health care and affordable houses - but food and shelter), I'm concerned there will not be many ways left where the rule of law and peace can be sustained.

It's one of those things, like "you can't unspill the milk". If things get as bad as I fear, right now, today is as good as it will ever get in the US. What scares me is that Americans aren't terrified enough, those that know better are in catatonic state of "what can i possibly do about it?" - and I mean all Americans regardless of politics.

where will brain drain go to? what's the better country for highly educated and desirable workers? especially for americans?
https://fred.stlouisfed.org/series/JTSJOL

Job openings are still very high, ~30% higher than the average for the last few decades.

Arent a lot of job openings known to be resume collection bins and HRs keep them open so that company can say "Oh we created this many jobs but no was willing to work for us"
This is lowest in the past year or so, not the lowest since say ‘08.
As someone who's about to quit to go work on their startup, glad to know things are gonna be terrible if everything goes to shit and I need to return to the job market.
Generally economists say response is necessary at over 4% unemployment. It doesn't take a big swing to unsettle the entire economy.

I guess my point is that even if you think AI's displacement of workforce is vastly overblown, be scared of 5% labor impacts.

Now this is what really is “AGI”.
AI is only going to reduce jobs further.

My own personal example working at a small govt contractor in which we had a team of 12 working on an app for 10 years. We lost the re-bid to a smaller team (5 to 6)/budget using AI. We have been informed going forward the company will be following the competition (small teams using AI and bidding like the competition via smaller budget needs).

Corporate America is such a weird place. They rather hire immigrants than retrain or educate existing workforce. Everything about the US economy is plug and play then discard. And if they can’t plug and play, they will just scrap the whole thing. That’s exactly what’s happening right now.
Where is the evidence that corporate America prefers immigrants over existing workforce? I've never heard such a claim before.

> if they can’t plug and play, they will just scrap the whole thing

This isn't just corporate America, it's all of America and all of humanity. Anything that's not plug and play has higher costs, higher risk, higher delays, and less chance of success.

Sudden and unexpected changes that increase uncertainty make it harder to invest in retraining, rather than easier. Increased uncertainty always decreases investment, because there's higher risk.

These are very very basic management considerations to anybody that runs a group, a lab, a small firm, or a large firm.

> Sudden and unexpected changes that increase uncertainty make it harder to invest in retraining, rather than easier.

This largely ignores the uncertainty and costs of a revolving workforce, and the value of having a workforce that all knows each other, the way the company works, and cares whether it succeeds.

When I was younger, I did a bunch of short-ish stints at various companies (I think average tenure was a hair under 2 years) because someone else would offer more money, more exciting work, etc. It was incredibly inefficient for the employer. "Waiting for my laptop/credentials to be issued" was like a full percentage point of my time there. I barely cared whether the company did well because I wasn't going to be there long enough for my RSUs to swing wildly either way. I never got to the point where I knew offhand who to talk to about niche parts of the product, and never became "that guy" for anyone else.

Frankly a lot of our stuff was higher risk and took longer because of the revolving doors. People important to a project would leave in the middle, or the person who wrote an important system would quit so we were left with whatever tribal knowledge we had.

Things worked well when the lady that wrote our invoicing system 20 years ago was still around but in security now. Things went poorly when she quit 6 months ago and now I have to reverse engineer it to figure out why I get stack overflows on invoices that contain an image.

Many job openings are just a way for recruiting to harvest resumes, or to give the impression there's some growth going on, or to conceal the actual open headcounts for teams. Ghost jobs.

This impacts vulnerable people looking for jobs and makes job search highly inefficient.

Makes sense. I know of multiple companies that reduced their headcount by a factor of 10 after Opus 4.5. Their new business model is basically: CEO + CTO + CFO + hundreds of Opus 4.5 agents.
There is going to be plenty of jobs for soldiers, nurses and gravediggers when WW3 starts.
I think this is somewhat normal? companies slow down hiring at the end of the year