This is good, thanks. In brief reply to your bullets:
Yeah, agree; this seems complicated.
I agree that progress isn’t inevitable. But to some extent it’s fine if you do-the-thing and don’t publish your research. But to some extent ideas leak.
I think LLMs are now sufficiently promising that if DeepMind, OpenAI, and Anthropic disappeared, the field would be set back a year or two but other labs would take their place.
This is good, thanks. In brief reply to your bullets:
Yeah, agree; this seems complicated.
I agree that progress isn’t inevitable. But to some extent it’s fine if you do-the-thing and don’t publish your research. But to some extent ideas leak.
I think LLMs are now sufficiently promising that if DeepMind, OpenAI, and Anthropic disappeared, the field would be set back a year or two but other labs would take their place.