Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Leaked OpenAI financials show $38.5B loss and compute burn

runtimewire.com

Post reply on HN
Print view View original

271–278 of 278 posts

Re: Leaked OpenAI financials show $38.5B loss and compute burn

#271

Post by danielmarkbruce »

dofm wrote: post #262
danielmarkbruce wrote: post #258

Earlier quoted context omitted.

Look at the history of farming. Tractors also don't go to disneyland.

Right, but at best (putting aside the tractor's magnifying effect in the Great Depression) you would be looking at employment displacement from one sector to another. Those people went on to other jobs. Here you are talking about a shift taking employment out of the economy full stop — taking the money that would have gone to "a good chunk of labor" , to use your words, and giving it to AI firms. You are necessarily…

No, I'm not. I was using it to size the market for AI. I don't imagine those people just sit idle. They do something else. Tractors took a good chunk of labor out of farming, those people didn't sit idle.

Re: Leaked OpenAI financials show $38.5B loss and compute burn

#272

Post by danielmarkbruce »

dofm wrote: post #263
danielmarkbruce wrote: post #259

Earlier quoted context omitted.

I meant by dollar amount, not headcount.

OK, but look, again, if you take away a good chunk of the money that goes to employing people — that is what you mean, right? This money is lost to the labour market one way or another — then don't they buy less stuff because they have less money? It's the same as if you had taken half their jobs and forced everyone into underemployment through job sharing. Your formulation forswears them finding replacement jobs bec…

No, it doesn't. The economy can grow. Money isn't some fixed amount thing.

Re: Leaked OpenAI financials show $38.5B loss and compute burn

#273

Post by willsmith72 »

himata4113 wrote: post #226
willsmith72 wrote: post #201

Earlier quoted context omitted.

> R&D cost is static since when? do you think their R&D in 2026 is the same as in 2022?

22->26 was scaling up, right now they're at their peak R&D spending from today onwards for the next forseeable future until nvidia starts rolling out their new room-sized systems this will be the upper limit of what can be trained without diminishing returns to where you're just burning money for less than single digit gains. R&D grows because you have more money to spend, not because the R&D costs are increasing, th…

I doubt it lets see who's right. 2028 r&d is it higher or lower than 2026.

Re: Leaked OpenAI financials show $38.5B loss and compute burn

#274

Post by surgical_fire »

nl wrote: post #269
surgical_fire wrote: post #231

Earlier quoted context omitted.

> Revenue is higher than cost of revenue and revenue is growing faster than cost of revenue. That's not necessarily true. It depends on two things that are impossible to derive from the leaked numbers: 1 - How much of their compute costs are subsidized. 2 - How much of that R&D chunk can actually be reduced for the company to continue working. What goes into "R&D"? Is that only things such as training for new models?…

All model training is in R&D > How much of that R&D chunk can actually be reduced for the company to continue working. Worth noting that almost all engineering (including software engineering) is always included in R&D. Development is important! Most software (and indeed engineering!) companies couldn't continue if they stopped R&D. Ford couldn't continue as a company without R&D.

> All model training is in R&D

Yes. But is all R&D model training? That was the actual question.

Everybody that is presuming "inference is profitable" here is just naively deducting "cost of revenue" from "revenue" and running with that.

Re: Leaked OpenAI financials show $38.5B loss and compute burn

#275

Post by awillowingmind »

danielmarkbruce wrote: post #271
dofm wrote: post #262

Earlier quoted context omitted.

Right, but at best (putting aside the tractor's magnifying effect in the Great Depression) you would be looking at employment displacement from one sector to another. Those people went on to other jobs. Here you are talking about a shift taking employment out of the economy full stop — taking the money that would have gone to "a good chunk of labor" , to use your words, and giving it to AI firms. You are necessarily…

No, I'm not. I was using it to size the market for AI. I don't imagine those people just sit idle. They do something else. Tractors took a good chunk of labor out of farming, those people didn't sit idle.

Can you give an example of what people may transition to?

Re: Leaked OpenAI financials show $38.5B loss and compute burn

#276

Post by danielmarkbruce »

awillowingmind wrote: post #275
danielmarkbruce wrote: post #271

Earlier quoted context omitted.

No, I'm not. I was using it to size the market for AI. I don't imagine those people just sit idle. They do something else. Tractors took a good chunk of labor out of farming, those people didn't sit idle.

Can you give an example of what people may transition to?

lol if i knew the next big industry i wouldn't be sitting here chatting on hacker news.

New industry after new industry has been created over the past 100 years. It's difficult to believe that all of a sudden that stops.

Re: Leaked OpenAI financials show $38.5B loss and compute burn

#277

Post by ted_dunning »

pu_pe wrote: post #232
ted_dunning wrote: post #91

Earlier quoted context omitted.

Add in the fact that they claim 900 million weekly uniques. Pretending the growth and cost rates compound as described in the article, they will need to generate about 100x current revenue to growth out of their current hole. That sort of implies that they will have 10x the entire world's population as weekly uniques at that time. That seems unlikely.

It's not like their 10x revenue growth year-to-year was based on expanding their customer base 10x though. The vast majority of weekly users are not paying, so they are actually a cost right now, but could be converted into paying users or audience for ads in the future.

That's a fair point.

My own experience with converting unpaid customers is that it is really hard.

Re: Leaked OpenAI financials show $38.5B loss and compute burn

#278

Post by gamblor956 »

HWR_14 wrote: post #58
gamblor956 wrote: post #38

Earlier quoted context omitted.

That's only true once they become public. As a private entity they can calculate COR any reasonable way that their investors are okay with. (For tax purposes for most businesses it doesn't matter whether an expense is CORS or not.)

COR is a GAAP concept.

COR is an accounting concept, and GAAP does not dictate what has to go into COR because that is industry specific. (Though in that sense, the industry-specific practices determine GAAP for that industry.)
Post reply on HN