Drop the definitive mic
Geoffery Hinton explains the difference between symbolic AI and deep learning to great applause from the @g7 audience. A must listen:pic.twitter.com/HbeQZX9QkH
Show this thread
-
works fine in lots of cases as an engineering approach to data representation i.e. graph databases, knowledge graphs, web markup etc; but then your processing model ends up being something in the sql, datalog tradition, not so 'intelligent' and not a model of natural intelligence
just curious: are you @danbri indirectly claiming deep learning is a good model of natural intelligence? beyond low-level perception? (@AmirRosenfeld might not even agree there). i think cognition remains a mystery, not well captured by any current approach
-
-
modern ML-style neural nets are a good model for natural intelligence, in same sense that bricks & mortar are good models for many cities. Their representational style (continuous, dynamic, yadda yadda) is at least as interesting as the learning algos; plenty scope for nativists
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.