Zitron just reported that OpenAI lost $38.5 billion dollars in 2025:

https://www.wheresyoured.at/exclusive-openai-financials/

Exclusive: OpenAI Losses Increased Nearly 8X in 2025, With Spending Hitting $34 Billion

Soundtrack: In Flames - Colony To further support my independent journalism, please subscribe to my premium newsletter. It’s $7 a month or $70 a year. If you’re subscribed to the free newsletter and logged in, you should see at the bottom right hand corner of your screen a little

Ed Zitron's Where's Your Ed At

@ludicity I don't think I've ever seen Zitron be that... plain about something.

Honestly kind of terrifying.

@cxberger He didn't even text me to crow this morning, a tradition after every single post, which means the world is ending
@ludicity Maybe if they try hard, they can lose $100
billion! That's how the economy works now, right?

@theorangetheme why are we setting our sights so low? Aim for the stars, go for one trillion in losses!

@ludicity

@ludicity
Okay, but hear me out:
Could we get rid of all the billionaires by assigning each of them full ownership of an AI company?
@ludicity we should describe this in "bird sites lost", eg "OpenAI lost 0.875 bird sites in 2025". This is not ecological advice.
@ludicity that's good, right? 🤓
@bweller The more money you lose, the gooder it is!

@ludicity

OpenAi: _rips receipt_

"Give this bill to the taxpayers at once."

@ludicity that means Sam Altman is running for president, no?

@ludicity looks like, at a surface level, inference is bringing more revenue than cost of revenue. Will depend a lot on the details though - e.g. I expect a lot of the CoR will come from deeply discounted rented capacity. But we’re going to hear a lot of gloating about it being “profitable”.

I also don’t see the net income - operating income less depreciation/amortization and interest payments. I expect those will grow worse.

Hopefully we get to see the full GAAP trio soon

@jacques @ludicity I hear more and more about executives complaining about tokens being too expensive. If LLM companies can keep the cost of inference under control, they shouldn't have allowed these stories to happen
@ludicity Losing money with every customer is fine. We'll make it up in volume.