Debunking Statistical AI – Noam Chomsky, Gary Marcus, Jeremy Kahn [video]

At Web Summit, November 2022, The father of modern linguistics, Noam Chomsky, joins scientist, author and entrepreneur Gary Marcus for a wide-ranging discussion that touches on why the myths surrounding AI are so dangerous, the inadvisability of relying on artificial intelligence tech as a world-saver, and where it all went wrong.

Chomsky is saying what he has basically been saying for a while - deep learning neural networks are "strong" (as he called it) in engineering, but "weak" in being scientific.

Chomsky’s gotta hate all forms of reinforcement learning successes. After all, what else are those if not the proof that his arch enemy BF Skinner was right.

Chomsky came to fame by criticizing Skinner’s book “Verbal behavior“, arguing, there was more to language and intelligence than statistics. Now here we are, 30 years after Skinner’s death, and the closest thing we have to general artificial intelligence is based exactly on what Skinner described.

"Just a waste of energy in California." (in regards to large language models)

Chomsky is insufferable.

Anybody believe people think like linear algebra, does not understand how linear algebra works.

Url changed from https://yewtu.be/watch?v=PBdZi_JtV4c, which points to this.

I also changed the title to take out "Great AI Lie", which is obvious linkabit. From https://news.ycombinator.com/newsguidelines.html: "Please use the original title, unless it is misleading or linkbait."

If anyone can suggest a better (i.e. more accurate and neutral) title, we can change it again.