My optimistic take: there's a technological time lag between when the models are proposed to model cogsci experiments and when they can be implemented at scale. Maybe when we have bigger computers and fancier inference algorithms everyone will use hierarchical Bayesian models.
-
-
Replying to @tallinzen @gchrupala
Pretty sure Hinton wasn't "believed" and was very alone before Deep Networks took off.
2 replies 0 retweets 3 likes -
By "pretty sure" I'm being facetious — I mean "I am 100% sure".
1 reply 0 retweets 4 likes -
Also you're saying he's not influential (Josh) but he's a keynote at the biggest ML conference.
1 reply 0 retweets 3 likes -
-
Replying to @gchrupala @tallinzen
I mean sure, maybe he isn't... He's also younger, so there's some confounds here in the comparison. Is your point that CogSci is being left behind?
2 replies 0 retweets 1 like -
I don't wanna be rude but you came here and said "all the cool work in your field died out", so I think it's fighting words!
2 replies 1 retweet 1 like -
I'd love to see some evidence supporting the above assertion "all this cool work in cogsci pretty much died out." otherwise one may consider it a pretty blunt comment.
1 reply 0 retweets 1 like -
let's not take
@gchrupala's statement out of context - he was referring specifically to foundational modeling work that makes an impact on ML, not to CogSci more generally.1 reply 0 retweets 2 likes -
I don't think I am. I still like to see some evidence.
3 replies 0 retweets 2 likes
It was kind of a strange thing to say IMHO but I'm glad it's become/is softer than face value!
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.