[abuani]

I'm surprised the article or the source for it didn't dive deeper into the impact the changes to H1B have had on the tech job numbers. Mind you I'm not trying to argue for or against them, but I find it hard to believe the changes aren't contributing at least a small amount to the drop.

The other thing not mentioned is the impact on the end for ZIRP. Every tech company with a pulse over hired such a staggering amount during the pandemic. It's not surprising these companies are returning to reality and not hiring back to the same levels.

[the_real_cher]

Its impossible to ignore when a company laysoff thousands but then has thousands if verifiable H1Bs in the H1B logs.

I agree with you. I'm not arguing for or against it.

I do think leaving it out of the conversation is a willful choice and blaming job loss solely on AI is becoming propaganda at this point.

[alephnerd]

Employees on work visas are the first and easiest employees to lay off because we don't see our insurance premiums rise when laying them off (it's what most of us did early during COVID). And now with the $100K fees we have no incentive to hire most employees on a work visa anyhow.

Hear that sound? It's the sound of GCCs opening in Poland, Romania, Israel, and India with those employees who were on work visas in the US now in charge.

If you remain in a major tech hub like the Bay or NYC where you are close to early stage capital, you are secure as it gives you a density of established and early stage employers which makes job hunting easier.

If you are in an inshoring hub like Atlanta, RTP, Denver, Minneapolis, or Pittsburg you're in big trouble.

If you are remote first and lack a network of friends and colleagues in the Bay and NYC who can personally vouch and refer you for roles at their companies you're in big trouble.

If you are a bootcamp grad, you are also in trouble so get the cheapest online degree you can that forces you to take a real algorithms and systems programming class.

Edit: can't reply

> The matter of fact nature of your reply sounds highly, personally anecdotal. Given the confluence of moving parts from AI and policy/politics alone, let alone potential regional instability and the effects on US financial markets, it’s really impossible to speak to the next quarter, let alone year/years. Not saying your prediction is not correct, but few are

Well, duh. This is an anonymous forum, not a research paper. That said, I am speaking from my personal experience, and a couple of other decisionmakers on HN have also voiced similar sentiments on here.

It's free advice - take or don't. It's not my problem. To quote Buck Strickland, "I ain't yo daddy".

[testbjjl]

The matter of fact nature of your reply sounds highly, personally anecdotal. Given the confluence of moving parts from AI and policy/politics alone, let alone potential regional instability and the effects on US financial markets, it’s really impossible to speak to the next quarter, let alone year/years. Not saying your prediction is not correct, but few are.

[porridgeraisin]

Meta MLE hiring seems to still be going strong. They recently picked up a whole slew of people - all H1B - from one of the UCSD labs to work at HQ.

[alephnerd]

> Meta MLE hiring

> from one of the UCSD labs

That's why.

For bleeding edge R&D, we still hire domestically because the capacity, network, and capital remains in the US, but these roles will overwhelmingly be in the Bay or NYC as I mentioned above, and will overwhelmingly recruit from a handful of top programs with an established track record.

Most HNers are median or below median (that's how medians work because life is a Gaussian distribution) so a large portion will be negatively impacted.

Sam who did his BSCS at Michigan State and Sandeep who did his MSCS without a thesis at SUNY New Platz will be negatively impacted, but Sarah who did her BSCS at UT Austin or Sarita who did her MSCS with thesis at UIUC will be hired.

Similarly for L1/2, it ain't happening unless you have a proven track record at a US comparable office like MSRL Bangalore, but that candidate would have ended up at a top CS program doing an MS/PhD anyhow and have a relatively easy path to immigrate to the US.

Basically, the same funda that exists in India about university tiers for hiring has returned to the US, but a lot of HNers aren't ready for this level of competition.

The difference between Kamal who is slogging at a mid-level SWE at Infosys in Indore and Karl who's working as a SWE at Cisco's NC office is marginal - both are going to be made structurally unemployed, with Kamal's work going to be automated with a CodeGen LLM in the hands of a junior SWE, and Karl's job being moved to a Cisco GCC in India, Eastern Europe, or LatAm.

The only way either can protect themselves is to find a way to end up at a top employer in a major hub.

Edit: can't reply

> What do you mean

Basically, it's best for all of us (be us SWEs, PMs, VCs, or SEs) to assume that we are average or below average, as that allows us to think strategically and defensively, thus allowing us to reduce our risk and take any advantage that we can.

That requires a level of competitiveness and strategic thinking that some may find distasteful, but it's the only way you can survive in the industry long term.

It doesn't mean hurt or negatively impact anyone always, but it requires one to be realistic and recognize that we are not in a stag-hunt but a Hobbesian trap.

[eddbuildsthings]

> Most HNers are median or below median (that's how medians work because life is a Gaussian distribution)

What do you mean?

[porridgeraisin]

My point was that H1B hiring still seems strong. Despite the fees and stuff.

Otherwise yeah what you're saying makes sense.

[alephnerd]

It depends on the tier of H1B hiring. I think you study/studied at an INI so your peer group has more in common with Sarah and Sarita instead of Sam, Sandeep, Karl, or Kamal.

In a lot of cases, Sandeep used to be Kamal, and these are the Desis on H1Bs or L1s who are getting laid off or forced to consider moving back to India, the UK, the EU, Australia, etc.

[gravisultra]

I don't know what a GCC is but I assure you that no one is opening offices in Israel.

[alephnerd]

I can assure you otherwise - hell two of my Israeli/American startups are about to come out of stealth in the coming days as RSAC is approaching.

Even with the current Iran War and the Gaza War, it's been business as usual.

Edit: can't reply

> You may be putting your money there, but consumers aren't having it

Most companies aren't B2C or B2B2C. Enterprise and B2B is solely outcome driven.

[gravisultra]

Look at the blowback on Vercel after the CEO went to Israel. You may be putting your money there, but consumers aren't having it.

[disgruntledphd2]

Not right now, but there's a big tech scene in Tel Aviv and that won't go away for a while.

[bayarearefugee]

> It's not surprising these companies are returning to reality and not hiring back to the same levels.

The common claim from the "dont worry about AI stealing jobs" crowd is that there is nearly limitless demand for new software to be written.

Even if over hiring is the reason for a lot of current job losses, the fact that over hiring is possible makes it obvious their Jevron's paradox claims are either lies or an attempt at self-soothing.

[el_benhameen]

The process of figuring out what to build has always been harder and more drawn out than the process of building it. I’m not arguing one way or the other for the Jevon’s paradox claims, but the steelman argument that you’ve missed is that job losses can happen very quickly (“last week’s version of Claude code is good enough that we can fire Joe and have Sarah do twice the work”), but the recovery can take a long time as the tech slowly diffuses throughout the economy and slowly spurs new ideas.

[dangus]

I think a more common claim is that the current downturn is way more obviously explained by Covid-era overhiring.

Basically all the charts on that subject are incredibly eye-opening.

To me it shows that the tech industry has actually been extremely resilient. Despite clearly going on an insane hiring spree that was not justifiable in the end, there hasn’t really been the kind of collapse we should expect. Tech companies have still maintained revenues and we haven’t seen any real sign of collapse. Companies that have gone trough major layoffs generally still have more employees than they did before 2020.

What we are seeing are some tech employees who were used to nearly a decade of easy work act like spending 3-6 months looking for a job is industry apocalypse.

I think you actually need to support the idea that there isn’t a nearly endless demand for software to be written.

Software is just business logic.

What businesses don’t involve software?

What businesses are you seeing that are saying “our software is done, there are no more ways to optimize our business, and there’s no need to evolve business processes to compete, we don’t need to update it anymore!”?

I can’t think of any business vertical that isn’t expecting constant improvement with their software.

[lelandbatey]

I think the timelines are too short for trends to be completely apparent yet. You can typically hire people faster than you can scale your income sources, even in the face of tremendous demand. Right this moment there's factors pushing folks to fire, but I also do see some companies delivering more (not a lot, but noticably more) and seeing increasing sales as a result. Those are in conflict, and we'll see which way the trends push through time.

[visarga]

AI work creates surplus, we eat that away by specializing and becoming more dependent. Work doesn't stop, it becomes higher stakes, we depend on each other and AI now.