To be honest, I would take a private space agency 7 days out of the week with that kind of capital. We have no fundamental proof that LLMs will scale to the intelligence levels that we imagine in our heads. The industry application for LLMs is even weaker than computer vision, and the public sentiment is almost completely against it. Sam's product is hype; eventually people are going to realize that Q* and Strawberry were marketing moves intended to extend OpenAI's news cycle relevancy and not serious steps towards superintelligence. We were promised tools, and they're shipping toys.

I could tell you in very plain terms how a competitor to Boeing and SpaceX would benefit the American economy. I have not even the faintest fucking clue what "AGI" even is, or how it's profitable if it resembles the LLMs that OpenAI is selling today.