My takeaway from this is that it's incredibly validating as a business model. Inference is _highly_ profitable. Of course, like any company that has ever tried to grow at breakneck pace, you run at a loss until you "win."

Isn't this what all of the big companies that spend a lot on R&D and engineers promise?

And then the reality turns out not to be the case - you have to continuously spend on R&D to avoid getting your lunch eaten by someone else.

This isn't a social media network with lockin either. People can and will just switch to whatever whenever they feel like it. Maybe it becomes a defacto standard like google but if someone is much better than you, well...

>> Inference is _highly_ profitable.

Totally untrue.

> My takeaway from this is that it's incredibly validating as a business model. Inference is _highly_ profitable.

The problem is you can't just separate training costs from inference costs. If OpenAI just didn't train a new model for the next five years, sure, they'd do OK. Assuming all those dirt cheap Chinese models nipping at their heels don't make up the gap while OpenAI is resting on their laurels.

Without being a frontier model (read: continuous, incredibly expensive training), they effectively don't have much to sell. So inference and training costs are intertwined to some extent.

Yes, it is like a new era - the startups have huge direct revenue on real products instead of "users" which yet to be monetized.

And the network effect which ruled for the last 20 years seems to have relaxed its death grip just a bit (of course it is still there as having more customers using your tools and models provides more training data, etc., yet the current network effect doesn't seem to have that high exponential value like before)