Yeah, this is a justification, but still -- they save single digit billions doing this, while AI capex is $150B (same timeframe) and RL spend is $16B. Feels like you could make the same cut from AI capex and barely notice a difference.
Yeah, this is a justification, but still -- they save single digit billions doing this, while AI capex is $150B (same timeframe) and RL spend is $16B. Feels like you could make the same cut from AI capex and barely notice a difference.
My gut feel was that you can't be right, but it looks like you are: cutting 8000 employees * $500k/year total cost to company (rough but useful ballpark figure) is "only" $4B.
Cross-checking against actual expenditure, Meta spent $118B total last year, with the second largest component of total spending being stock comp at $42B, of which vast slabs went to the top leadership that's presumably also not getting fired.
Though not all of that capex is cash; there's a whole phantom wampum AI economy where the big players are trading promissory notes for compute that doesn't exist yet (and may never exist) some time in the future and booking it as revenue.
Maybe you're thinking of AWS/Azure/Oracle? Meta isn't selling their compute.
Meta plays that game too; they're on the hook to buy compute that CoreWeave has yet to build (and may never be able to build) which counts as "revenue" for CoreWeave and an "asset" for Meta even though no actual money or compute has changed hands.
Meta promised to buy dc capacity for ai workloads. If I remember it correctly, it created a common company with an investment fund as well that took on debt to build capacity.
You calculate the cutoffs as savings for this years while imagining that the future payments are payments only for this year. At the same time the commitments are for 5-20 years ahead and the laid off people would be off the payroll for the same multiple years ahead.
These games are far more dangerous to the industry than "AI now = badcode future".
They're a suicidal bet, because they assume cloud LLMs are efficient and inevitable.
Neither of those is true, or even likely, and we're going to see the consequences by the end of the decade.
The figure I've seen floated is $6B, but yeah, same idea.
It’s slightly more nuanced than layoffs = capex. You’re right, they don’t. That said, they do create free cash flow, which the market uses as one important input into the value of a given stock. Moving FCF positively when capex spending is moving it the other way is the real financial accounting move that is happening here.
salaries are opex, data centers are capex, you can't compare them in the same timeframe.
4B over 5 years is 20B, which is significant.
4B over 5 years is still 4B. It's 0.8B/yr
If I read it right, the 4B ballpark figure is based on total annual per employee cost of 500k * 8000 employees, so the figure is actually 4B/year. 20B over five years.
I think any company that is seen to reduce capex right now is going to be the Bear Sterns of this cycle
Is this because you think the market will short sell them, or because capex is so worthwhile right now that a company which doesn't invest will fall fatally behind?
The former; I think there's simply not the capacity to build the infrastructure that would be required to significantly improve the current tech as much as is expected (and there may never be the capacity to build that much infrastructure).
There's a whole lot of circular funding being passed among the same dozen or so companies right now with very little actual construction or assets to show for it and at some point someone will be holding the bag when actual money is called for, and nobody wants it to be them. The parallels with both 2008 and the '90s S&L crisis are troubling.
"they're not even trying to catch up"
Doing slightly less than 150b looks bad to investors. Or at least it looks small.
Imagine the productivity gains if they just spent $150B on booze and cake for employees!
That is one of my huge complaints about the current levels of AI investments. You can do pretty much anything you want if you got $150B to spend, and then you go burn it on being uncompetitive in the AI space. Then you have the $80B Metas Reality Labs spend on a failed virtual reality and a pair of Ray-Bans.
It's not like Meta has nothing to show for the money it spend, but it seems like they could have spend that money on improving Facebook or Instagram, not that I think Zuckerberg really cares about those product anymore.
he doesn't but increasingly people don't care about facebook, and that will come for instagram as well. These social apps seem to be .. generational. Boomers use facebook and they are dropping like flies. Twitter has it's own cohort, instagram it's own. I don't expect these products to last forever, tiktok took a huge bite out of instagram. These are fad products.
Mark is just looking for the next fad.
And coke. Not the cold drink.
you get another taste of coke every time you clear 100 tickets. not the cold drink.