@fchollet That is basically what the Chinese Room is trying to get at. Symbol grounding.
Language is an interface between the inner worlds of humans. Current NLP AIs are mimicking the interface, with none of the representations.
-
-
-
@Pawelotti I don't think so, the Chinese Room is an idiotic argument based on semantics. It boils down to a tautology. - Show replies
New conversation -
-
-
@fchollet I would replace 'none' by 'few', at least for multimodal embeddings (text + image + speech). -
@ogrisel@fchollet I can find http://www.cs.toronto.edu/~zemel/documents/mnlm2014.pdf … and http://web.stanford.edu/~bgirod/pdfs/ChenHuizhongACM_MM14_multi-modal-language.pdf …. That is few indeed or are there others? - Show replies
New conversation -
-
-
@fchollet I agree perfect mimicry is probably not possible without experience / embodiment but I don't think 'mimicry' it-self is the issue.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.