Gary Marcus

Hinton’s “favorite Gary Marcus quote” argues that pure #deeplearning fails to understand language. I stand it by it: no current #AI system can understand a stories, articles, etc. If you disagree w the quote, I’d like to hear your thoughts why.https://twitter.com/rsalakhu/status/1187518370276614145 …

The issue is that "understanding" is a sliding scale. Obviously today's AI systems (which are virtually all hybrids) understand some things and not others. There is plenty of room for improvement!

operationally: answer questions about who did what where when why and how. what motivations were, what agents intended to achieve, what obstacles they faced, what sequences of events was etc. stuff we expect of any normal reader. discussion: http://rebooting.ai  chapter 4

Yes, I agree that those are all things that are beyond the state of the art. The part of the glass that's half full includes lots of queries to google, alexa, etc. that are understood correctly, with significant contributions from deep learning.

Re the Google/Alexa queries, I’d not use the phrase “understood correctly”. There’s little to no understanding going on (although there is lots of pattern matching).

Agreed. Try typing "largest lake not in North America" into Google and the results are dominated by the largest lakes _in_ North America. This is consistent with pattern matching to me ...