The problem with your argument is not the goal, but the approach to the goal. There's a whole bunch of researchers who have a paper out "Neural-Symbolic Learning and Reasoning: A Survey and Interpretation" https://arxiv.org/abs/1711.03902 that discuss all kinds of fanciful approaches.
Isn’t that what i have been arguing for, my entire career?https://twitter.com/mpshanahan/status/952892083576156161 …
-
-
-
The correct approach is one where the is no native machinery for symbolic manipulation. Symbolic manipulation is simulated by a connectionist architecture. It is of course inefficient compared to how computers do it, but that's the reality.
End of conversation
New conversation -
-
-
I guess so :-)
-
I can make no such claim. I spent 10+ years working in the symbolic paradigm, then 10+ years working on neurodynamics, and only came around to the view that NNs and GOFAI should be reconciled a couple of years ago
- 1 more reply
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.