Biological Anchors: A Trick That Might Or Might Not Work
Astral Codex Ten Podcast
00:00
Openfield's Estimate of a G I Is a Double Edged Sword
Openfield: Maybe the deep learning paradime can't produce a g i, and we'll have to wait decades or centuries for some one to have the right insight. Eliazer takes this objection seriously, but feels like, on balance, probably new paradimes will speed up ai rather than slow it down. So he still thinks openfield's methods are more likely to overestimate the amount of time until a g i,. Rather than underestimate it. What more's law giveth, platt's law taketh away. Two years before the end of the world, there'll be another published biologically inspired estimate showing that a g i is 30 years away. And it will be
Transcript
Play full episode
