It's very in vogue right now to clown on the frontier labs for "not being profitable ," but they are
It's very in vogue right now to clown on the frontier labs for "not being profitable," but they are obviously spending money competitively. Revenue is growing quickly. The market is tight. Everything is close.
- They are clearly spending in anticipation of the market growing bigger.
Do folks seriously think it's the time for OpenAI and Anthropic, etc, take their feet off the gas?
For example: Let's say cheap Chinese models begin to undercut them and spending massively on the next model only pays with a few large customers.
Fine. No problem. They release a cheap product based on old models if that really becomes a problem. But the AI companies still want that most valuable customer, right? The ones who will spend the most to be competitive? Especially because, with diffusion, the ranks of those most valuable customers are likely to grow.
Maybe they won't! But if you're the founder of a frontier lab, I would hope that you would be inclined to believe they will.
When people started using Netflix rather than Blockbuster, and Blockbuster couldn't cover rent on its locations any more, that was an unprofitable company. Fair. But a company spending ahead of projected demand might be technically unprofitable, but it's not really the same thing.
Am I missing something?
Versions · current v2 · 2026-10-11
- Version 1 · frozen pin — 2026-10-09 Fragment from openers.html