“I’m not saying I want to forget deep learning... But we need to be able to extend it to do things like reasoning, learning causality, and exploring the world .” - Yoshua Bengio, not unlike what I have been saying since 2012 in The New Yorker.https://www.technologyreview.com/s/612434/one-of-the-fathers-of-ai-is-worried-about-its-future/ …
No, you just haven’t read it. There’s a very detailed analysis of what happens in various language models with and without operations over variables in Chapter 3 of The Algebraic Mind, recently replicated and extended by @LakeBrenden and others.
-
-
Well, I suspect that a majority of AI practitioners don't understand specifically the kind of 'symbol manipulation' that you are advocating. Perhaps a short elevator pitch is in order?
-
I think it's most interesting to consider
#AI systems that can learn to emulate symbol manipulation reasonably well but without it being their native mode of operation. Like humans, these systems would likely be prone to errors in manipulating those symbols. - 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.