[agarren]

How is it winner-take-all? Training models is expensive, yeah, but big companies are willing to do it, and if deepseek is any lesson, it’s not as expensive as we used to think. You might be building a castle in someone else’s kingdom, but you’ve got a few kingdom’s to choose from. None have a moat and none seem to be going away any time soon.

It’s cheaper to move from one model to another than it is to train a general purpose model yourself (to say nothing of domain-specific smaller models or anything open source.)

I’m not sure about problems turning into black boxes, LLMs are pretty explicit in my experience when producing a solution (good or bad.) _How_ they came about that solution _is_ a black box, but that’s not a new problem.