Of course, computational power isn't the real limiting factor for general #AI, anyway. We will need fundamental advances in understanding common sense knowledge and reasoning to achieve extrapolative AI rather than our current interpolative AI.https://twitter.com/Grady_Booch/status/1081740548627550209 …
-
These are good points. I guess I was thinking about whether it COULD be done with a pure learning approach. Of course the most practical approach might be related to the Belbute paper I mentioned: wrap ML around a physics engine.
Belbute-Peres et al certainly is fascinating, @tdietterich, but if truth looked like that, @ShlomoArgamon & I would celebrate. It has a huge amount of innate structure, eg position & velocity vectors for objects, constraints on friction, even priors re color schemes in Breakout
-
-
Indeed. The thing is, we don’t know how to create that sort of innate structure for much common sense physics, not to mention social interaction
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
By "truth" I assume you mean human intelligence. I totally agree. But I think it could be an excellent engineering solution.
-
I would feel vindicated there as well, inasmuch as adding innateness to AI is what I advocated here: https://arxiv.org/pdf/1801.05667.pdf …
- 2 more replies
New conversation -
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.