I'm temperamentally not a predicty person, but it looks to me like the coming wave of LLM-centric solutions to new-school platform governance is going to get a whole lot of resources and attention because it's the saving throw/deus ex machina for redeeming the combination of global scale + hyper-acceleration + a generally neoliberal-to-libertarian sense of what's good. I hope some boring-to-capital, sensible, and human-centric things break through.
I think that's all I will say about that.
Oops sorry not done, I think the other, quieter reason LLMs-as-AI are going to get shoehorned into everything governance is that they allow technologists to keep focusing on things that sound neutral or blandly prosocial rather than “political” and not being overtly political is a comically huge part of being perceived by powerful people as a legitimate and reasonable force. Just another structural reality.