Garry Tan on Naval
Live in the Future
Mostly a duplicate of the July 2 Naval appearances, with familiar arguments about personalized AI writing, frontier models, and household robots. Skip unless you specifically want the more speculative exchanges about nationalization or AI companionship.
Well, then they'd nationalize it and then, you know, maybe we could get the government to. Do they really nationalize it? They didn't nationalize Google, they didn't nationalize. You know, Apple's way bigger than that. It depends on how bad it gets. I mean, like search, it's a good utility. Like, there are lots of reasons. And if you nationalize it, it might slow down and lose because now the DMV is running it and China wins. It's not like the Chinese are asleep. There's an ASML machine supposedly somewhere in China. Now, did you hear about that? Howard Litnick called ASML and was all mad because an ASML machine somehow leaked into China. That's the lithography machine. There's only 184 of them.
Garry Tan contributes a strongly pro-adoption perspective, emphasizing that people should use AI directly to understand its capabilities and build systems around it. He discusses AI writing, general models competing with specialized software, the risks of concentrated control, and the prospect that robots could expand personal abundance.