Gotcha. Well we’ll see how the next year or so goes.
(I do agree we’re in a bubble, and also that there is something shallow about how the current AIs accomplish most of their problem-solving-tasks. But, seems to me like all the pieces are there for RL training on diverse problem solving to take it from here. And, like, the dotcom bubble crashed, but that doesn’t mean the internet didn’t end up dominating the market later anyway)
But, anyways, thanks for clarifying that stuff about Accelerando and welcome to LessWrong! (It sounds like you’d mostly find AI discourse on LW aggravating, but FYI you can click the gear icon at the top of the posts page, and set various tag-topics to “hidden” or “reduced” and anything else that’s interesting to you)
Gotcha. Well we’ll see how the next year or so goes.
(I do agree we’re in a bubble, and also that there is something shallow about how the current AIs accomplish most of their problem-solving-tasks. But, seems to me like all the pieces are there for RL training on diverse problem solving to take it from here. And, like, the dotcom bubble crashed, but that doesn’t mean the internet didn’t end up dominating the market later anyway)
But, anyways, thanks for clarifying that stuff about Accelerando and welcome to LessWrong! (It sounds like you’d mostly find AI discourse on LW aggravating, but FYI you can click the gear icon at the top of the posts page, and set various tag-topics to “hidden” or “reduced” and anything else that’s interesting to you)