The paper states that they created a model per subject. I wonder if it would have been possible to train just one that works for all subjects.
Also isn’t the visual cortex active when dreaming? Would love to see a recording of my dreams some day.
Agreed, I really enjoy his interviews with people from the industry (and occasionally other people as well). Donald Knuth, John Carmack, Demis Hassabis, Brendan Eich, Chris Lattner to name a few.
I only recently discovered Lex Fridman's podcast (2 monthes ago), but I've been obsessed! The Fiona Hill episode (#335) was great. The climate change discussion between Bjorn Lomborg and Andrew Revkin (#339) was very educational. Ray Kurzweil (#321) was awesome as well. I wonder how that guy (Lex) has time to work his real job since he seems to put out 2 episodes a week and they're all 2-4 hours in length and he clearly does his homework!
Yeah, but man does he get some great guests from time to time. Really enjoyed Travis Oliphant and Todd Howard in addition to the ones the parent listed.
I think it's a bit unfair. Overall, I'd say he's a decent interviewer and has interesting guests.
That being said, I find him pretentious, and get a bit tired of some recurring cliches and themes. I'd prefer if he would keep things a bit more "neutral". Also he invites occasionally some crackpots, which would be fine if he'd challenge them more, or invite another guest who can.
yeah, his questions and comments are quite often naive and cringe. But he has incredible guests, from all sorts of domains, and lets them talk for hours.
I suspect these code generating AIs will bring the singularity at some point in the future. Even if we don’t manage to create an artificial general intelligence, they will. I imagine they will learn to code on super human levels through self play just like AlphaGo and AlphaZero did. This will be awesome.
