We’re in a general economic decline masked, in some key aggregate measures, by the AI bubble. That’s the reason for the job cuts.
Shuffling money around between AI adjacent companies may (but I doubt it) be enough to keep the AI bubble going through the downturn, in which case stockholders in those key companies manage to ride it out without pain, and executives too because they serve their constituency well.
Both "everyone" and "to fund AI" are doing a lot of work there. There probably was a lot of hiring in excess of requirements in tech over the past five years or whatever--and a bunch of people who jumped on board the ship primarily to capitalize on the gold rush are going to end up suffering as a result. Doesn't mean that tech generally isn't still a big opportunity going forward but may very well mean that outsized tech money opportunity on the west coast of the US declines. (And cost of living declines will lag) It's happened before with respect to the other coast (and the Midwest to some degree).
While there are points here that I agree with, I do not agree with the premise that they are firing people to fund AI. Funding of AI (keyword funding, not usage) I would say is an adjacent relation, but not the reason people are being let go.
I think it's a combination of factors including:
- Overhiring in the past, most of these companies were/are still employing way above the pre-pandemic numbers when overhiring was the norm. Many of these companies legitimately probably have more people than they need still.
- No matter what you think, AI is certainly already capable of improving productivity enough to make up for many of the jobs lost. Just look at how much AI can do today compared to even just a couple years ago. Practically every engineer who is worth their salt can put out way more than they previously could. Yes there is the question of quality and whether that's going to keep up, but you simply cannot argue at this point that AI is not able to increase productivity enough for the decreased headcount. Theres a reason junior hiring has fallen, they expect senior engineers to both output more using AI and to have the capability of effectively reeling it in and reviewing.
- As much as I hate to say it because it gets people angry and political: offshoring. Every company that I've worked on or worked at in the last few years has increased headcount of offshore employees and contractors.
- And lastly is simply expectations. Executives can see that other companies that have let go of people are able to still effectively run much slimmer.
The poster is kind of missing the point. If I see a tool that could make a lot of my employees 20% more efficient today, and maybe more than that in the future, I’m going to bet big on it.
Layoffs aren’t just to “fund AI” - layoffs (a) always happen and (b) likely genuinely make sense at some scale due to efficiency gain from AI.
It’s not “cut employees, send they money in circles to prop up stock prices” it’s “cut employees, like we always do (but more than usual since we have actual AI efficiency gains).” Separately, let’s yeet excess money in circles to prop up our stock price and AI is a great hype cycle vehicle for that.
Yeah layoffs always suck, especially big ones, but they literally always happen and IMO are not caused by AI circlejerking at the c level.
Also, for anyone curious about “AI gains”. In my role at bigtech company I need to have a high/medium level understanding of like 30 products and teams. LLM summaries of docs, tickets, slack, etc mean I can get a basic understanding and history of any topic in seconds, rather than spending hours reading all those sources. My role is not unique or special
Doesn't pass the smell test as someone who has been on boards that made these decisions.
This would be like saying cloud spend was circular because F1000 tech companies represent the majority of cloud spend from a revenue perspective, but tend to sell or partner with those same organizations. For example, AWS and Zscaler, or GCP and Broadcom.
Reality is, a lot of product lines were formed or overhired with little-to-no business justification during the late 2010s and early 2020s. Most of us are now cleaning shop and axing those product lines and features that cannot be attached to tangible revenue generation. Inevitably this leads to ICs getting axed.
If you want to blame someone, blame EMs, PMs, Principal SWEs, and VPs who decided to launch product lines and features for their resume instead of for business justification. (Edit: that said, BeFlatXIII is right that we probably won't have hired most people if this didn't happen at all)
Additionally, comp expectations just got too high - most new grads are not worth $150k-200k TC (some absolutely are though)
The "spending money on each other" part is untrue. It is mostly going in one direction, towards Nvidia, and that's the reason they crossed $5 Trillion in market cap recently.
I remember watching the firings of people right before the dot com bubble burst. Instead of pay employee's, companies thought they could double down and started putting MORE money into vaporware companies that eventually folded up and went bankrupt. I'm guessing were almost to that last stage again...
Most companies using AI are just paying for a subscription. They aren't actually building LLMs or other machine learning applications that take over a known task. You could easily replace the term AI with just subscription.
When they say they are laying off workers, they are saying "we signed up for Cursor, and enabled some AI features on Jira and Salesforce"
The layoff is just an excuse until they realize that just because Jira can write a summary of the ticket, doesn't mean the ticket gets done any faster.
I rarely post comments on LinkedIn, but I saw some of the dumbest content today and felt compelled to leave a response:
> Just met a founder who fired his entire team because he was able to individually beat their entire productivity with Claude Code
My only thought when I read this is that this absolutely embarrassing...for founder who wasted dozens of hours recruiting, qualifiying, interviewing, onboarding, and training these resources. Then paying their salaries for who knows how long!
It just came across as some of the most small-brain thinking that anyone would then "Like" this content and not see it for what it is: idiotic management and leadership and just complete lack of basic foresight.
Why are you hiring absolutely replaceable people in the first place?
This is turning out to be more wasteful than arms race. Companies are locked into a vicious spiral built by themselves. The only hope appears to be a bubble burst that wakes up investors.
There has never been a better time to build. The vast majority of executives likely don't comprehend AI and it's capability. (I'm not talking about silicon valley) I'm talking about everything else.
The highly likely scenario is that:
- These people will spend money on "AI" to solve stuff to keep their job
- These people will be so slow to respond that they are ripe for disruption/exit.
The consumption side has so far been held up by higher income consumers. Now that so many of them are being fired all that’s left to prop up the economy is AI spend. I’ve heard some think it could go on for a few more years but I’ve also increasingly seen bubble / circular finance posts so it feels like we’re getting closer.
It’s quite possible that they’d want to do the layoffs regardless, and the AI narrative just gives them a convenient excuse, with the added bonus that it also happens to promote AI.
22 comments
[ 4.2 ms ] story [ 42.9 ms ] threadI'm not terribly surprised.
Shuffling money around between AI adjacent companies may (but I doubt it) be enough to keep the AI bubble going through the downturn, in which case stockholders in those key companies manage to ride it out without pain, and executives too because they serve their constituency well.
More likely, it just delays the reckoning.
I think it's a combination of factors including:
- Overhiring in the past, most of these companies were/are still employing way above the pre-pandemic numbers when overhiring was the norm. Many of these companies legitimately probably have more people than they need still.
- No matter what you think, AI is certainly already capable of improving productivity enough to make up for many of the jobs lost. Just look at how much AI can do today compared to even just a couple years ago. Practically every engineer who is worth their salt can put out way more than they previously could. Yes there is the question of quality and whether that's going to keep up, but you simply cannot argue at this point that AI is not able to increase productivity enough for the decreased headcount. Theres a reason junior hiring has fallen, they expect senior engineers to both output more using AI and to have the capability of effectively reeling it in and reviewing.
- As much as I hate to say it because it gets people angry and political: offshoring. Every company that I've worked on or worked at in the last few years has increased headcount of offshore employees and contractors.
- And lastly is simply expectations. Executives can see that other companies that have let go of people are able to still effectively run much slimmer.
Layoffs aren’t just to “fund AI” - layoffs (a) always happen and (b) likely genuinely make sense at some scale due to efficiency gain from AI.
It’s not “cut employees, send they money in circles to prop up stock prices” it’s “cut employees, like we always do (but more than usual since we have actual AI efficiency gains).” Separately, let’s yeet excess money in circles to prop up our stock price and AI is a great hype cycle vehicle for that.
Yeah layoffs always suck, especially big ones, but they literally always happen and IMO are not caused by AI circlejerking at the c level.
Also, for anyone curious about “AI gains”. In my role at bigtech company I need to have a high/medium level understanding of like 30 products and teams. LLM summaries of docs, tickets, slack, etc mean I can get a basic understanding and history of any topic in seconds, rather than spending hours reading all those sources. My role is not unique or special
This would be like saying cloud spend was circular because F1000 tech companies represent the majority of cloud spend from a revenue perspective, but tend to sell or partner with those same organizations. For example, AWS and Zscaler, or GCP and Broadcom.
Reality is, a lot of product lines were formed or overhired with little-to-no business justification during the late 2010s and early 2020s. Most of us are now cleaning shop and axing those product lines and features that cannot be attached to tangible revenue generation. Inevitably this leads to ICs getting axed.
If you want to blame someone, blame EMs, PMs, Principal SWEs, and VPs who decided to launch product lines and features for their resume instead of for business justification. (Edit: that said, BeFlatXIII is right that we probably won't have hired most people if this didn't happen at all)
Additionally, comp expectations just got too high - most new grads are not worth $150k-200k TC (some absolutely are though)
When they say they are laying off workers, they are saying "we signed up for Cursor, and enabled some AI features on Jira and Salesforce"
The layoff is just an excuse until they realize that just because Jira can write a summary of the ticket, doesn't mean the ticket gets done any faster.
It just came across as some of the most small-brain thinking that anyone would then "Like" this content and not see it for what it is: idiotic management and leadership and just complete lack of basic foresight.
Why are you hiring absolutely replaceable people in the first place?
The highly likely scenario is that: - These people will spend money on "AI" to solve stuff to keep their job - These people will be so slow to respond that they are ripe for disruption/exit.
Google is not at its all-time high, but still pretty close:
https://www.macrotrends.net/stocks/charts/GOOG/alphabet/numb...
Amazon: similar.
https://www.macrotrends.net/stocks/charts/AMZN/amazon/number...
Microsoft: all time high?
https://www.macrotrends.net/stocks/charts/MSFT/microsoft/num...
It sucks if you need to find work, but these are still major employers, even after large layoffs. They can lay off thousands without much of a dent.
I'm old, I remember when Google was 20k people and I thought that was enormous.