I do not advocate unsupervised (or self-supervised) learning as a solution to reasoning. I advocate it as a solution to acquiring enough background knowledge for common sense to emerge. Reasoning is a somewhat separate problem. 1/2
Yes, @ylecun you (see quote) try to stifle interest in hybrid deep learning-symbolic models. You advocate unsupervised learning a solution to reasoning, without any real results to show that it works, and yet you dismiss symbols for lack of results. #doublestandardhttps://twitter.com/ylecun/status/1065648782736572422 …
-
-
-
I do not dismiss symbols. I dismiss approaches to reasoning that are incompatible with gradient-based learning. Make symbol-based differentiable, and you win. But you seem to disapprove of gradient-based learning.
- 1 more reply
New conversation -
-
-
OK. Mr. Marcus your are the smartest on the earth.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
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.