Comment by senordevnyc
15 hours ago
Pointing at the API margin and calling the whole business profitable is the financial equivalent of weighing yourself with one foot off the scale.
I didn’t call the business profitable, I said that inference is profitable. I was responding to your assertion that they’re speculating by selling below cost. Which isn’t true; they’re selling inference, profitably. They’re losing money because they’re investing in the next model. The company isn’t profitable, it might never be profitable, but the product they’re selling is profitable. So calling it speculation based on selling something below cost is just factually incorrect.
Granted on the narrow point: inference itself runs at a positive margin. Where it falls apart is the implicit claim that the training spend is separable.
It isn't. Frontier model training is the cost of having a product to sell inference on next year. Stop training and the inference margin decays on the timescale of the next competitor release, which in 2026 is measured in weeks. So "the product is profitable, the company is just investing" describes a business where the investment is structurally non-optional and structurally larger than the product margin. That's the definition of selling below cost at the level that matters, which is the level you're hedging at when you hold 49%.
McDonald's is profitable because a Big Mac in 2027 costs roughly what a Big Mac in 2026 cost to make. OpenAI's product depreciates to zero on a 12-month cycle unless they spend ~$40B keeping it ahead. That's the disagreement, and "but inference itself has positive margin" doesn't resolve it, it just relocates it.