We would be more than happy to host this debate in Australia. If you are both serious and keen about having this debate please let me know and I will organise it with our university partners in Sydney!
Would love to have a debate with @geoffreyhinton about this -- whether AI needs symbol-manipulation and hybrid models.
Anyone want to host and encourage him to join?https://twitter.com/fahd09/status/1143560030861647872 …
-
-
-
I am in if you can get
@geoffreyhinton to join. Cc@maxcoltheart - 5 more replies
New conversation -
-
-
I think the non-symbolic paradigm is a very productive path to pursue. We will learn a lot just as we learned that pure symbolic approaches don't work either. I don't see how to constrain hybrid methods to drive science (vs. doing good engineering)
- 6 more replies
New conversation -
-
-
Canberra could also be an option.
@feraldata - what do you think?Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
Do you think intelligent animals such as primates require symbolic manipulation to do what they can do?
-
i.e., mammal, or MAML (Finn and the Berkeley RL profs).
End of conversation
New conversation -
-
-
I've argued many times (to deaf ears, apparently) that DL suffers from a fatal flaw. It is a representational model and thus blind to patterns/objects it has not learned. Perception without representations is the future, not hybrid models. Marcus is deaf and so is Hinton.
-
This is all a false dichotomy. Symbolic and non-symbolic describe what happen in brains at different levels of description. There is no one privileged level.
End of conversation
New conversation -
-
-
OK email sent
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.