> I'm sure that if he had to decide on a direction to pivot to today, it'd be towards generative AI I'm pretty sure that I agree with you, but what would Face

> I'm sure that if he had to decide on a direction to pivot to today, it'd be towards generative AI I'm pretty sure that I agree with you, but what would Facebook/Meta do with LLMs? You wouldn't want to generate fake people to interact with (well, someone might, but Facebook wouldn't). Because that's what data they had would have been useful for. I mean, if you can predict what a person will say, or how they will react to an ad, that has value... but I'm not sure I know where that value is for a Facebook, outside of better targeting of ads. Maybe predicting consumer/voter sentiment about a policy or product before it's released? The same applies to Google in aa certain respect actually. I assume they have enough data from gmail to effectively recreate a conversation with me. This could be lucrative for some targets (not me!), but probably not in a way that Google would be able to legitimately capture. I could completely see how other actors would be able to use LLMs to predict or impersonate people on the Facebook/Google platforms, just not how the platforms themselves would be able to generate value. Google at least could merge generic LLMs into their search mission to catalog the world's information. But that's a harder pitch to figure out how Facebook would generate value from it. Am I missing something, or am I thinking too much about this like West World?