Getting to AGI using supercomputers is essentially the computing equivalent of using a tank to kill a fly. You might be able to but what's the cost? And is it the best way to get there?https://www.google.com/amp/s/amp.ft.com/content/c96e43be-b4df-11e9-8cb2-799a3a8cf37b …
-
Coming from Dreyfus, this also carries a very different meaning. I think one of the big issues with
@GaryMarcus' proposals is the willingness to re-introduce aspects of GOFAI (e.g., symbols) that Dreyfus so vehemently (and, I believe, correctly!) contested to GOFAI for decades.
And Dreyfus thought AI was effectively impossible, I just think we need some new approaches. (And to be clear, I advocate hybrid, not straight GOFAI)
-
-
Nope. I'm a big fan of Dreyfus. He believed that AGI was impossible with symbolic AI, the type of AI that you want to reintroduce into the field. Not that it has ever left, mind you. DNNs are effectively expert systems. Dreyfus was a non-representationalist, a fan of Heidegger.
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.