Last week: Deep learning is on verge of understanding natural language. Problem of tracking world change over time solved 3 years ago. This week: Nobody ever said statistical language models understand anything.https://twitter.com/tdietterich/status/1191756945226522624 …
-
-
Replying to @GaryMarcus
Now now. I didn't say they understand nothing; you know very well I wouldn't claim that. The "extend this prompt" demos show that the new models extend the span of "local coherence" far beyond what n-gram models would capture. But they also reveal major semantic failures...
2 replies 0 retweets 6 likes -
Replying to @tdietterich @GaryMarcus
I would say
@tdietterich the statistical language models are limited to incremental improvements because of science. As Yoshua Bengio says, DL is lacking in understanding & generalization out side the training data. That's not generalization required for scaled NLU to@PatisNLU2 replies 1 retweet 2 likes
and as gary marcus says, and has been saying since 2001.
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.