Even after reading through the discussion I don't understand why he was "wrong". He merely made a statement based on his expertise and offered a direction to fix bias in implementing AI systems.

I hope we can still follow him somewhere where he will not be attacked groundlessly while sharing the newest AI research progress. His research is inspirational.

There was a Twitter kerfuffle about bias in AI: https://www.reddit.com/r/MachineLearning/comments/hdi9hq/dis...

People are exceedingly obnoxious on Twitter, and LeCun got sick of it.

Isn't he uniquely qualified to help rectify bias issues. It seems odd to just leave.

I don't think AI is fixable when applied to humanity.

For example, laws have been passed at the city level (Santa Cruz, Boston) against facial recognition, but we need federal laws.

There is more context here: https://venturebeat.com/2020/06/26/ai-weekly-a-deep-learning...

I appreciated this response thread by Nicolas Le Roux on twitter: https://twitter.com/le_roux_nicolas/status/12754857362597928... especially this: "Why did you reply by repeating what you thought was true, even after many explicitly told you that what you believed to be true was incorrect?" ... "did you watch the tutorials or read the papers that were pointed to you. What did you think of them?"

>"did you watch the tutorials or read the papers that were pointed to you. What did you think of them?"

As if he doesn't have better things to do then give opinions on people's papers and random tutorials about something he doesn't even disagree with. What would it accomplish?

Also: "Fourth, you asked to assume good intent (https://twitter.com/ylecun/status/1275174556152586240), which is known to maintain the balance of power"

Asking to assume good intent is now a power move to the left? This is "known"? Awesome.

Well, mission accomplished, Yann LeCun won't be participating in this discourse any more, and we're all better off, right?