Any theories of what Elon's motivation is with the increasingly strident public alarm bells he tweets out re AI danger and his view that regulation is needed sooner rather than later?
Friday night he sent out a new one claiming AI is more dangerous than North Korea - the same night he publicized the fact that an OpenAI bot beat the world's best Dota 2 players. Both the tweet and the OpenAI stunt smell of really cheap, low level fear mongering and publicity seeking. I was impressed with the Dota 2 achievement until I read the only piece I could find on the game - in which the author essentially claimed that the OpenAI bot probably "cheated" - in the sense that the bot likely was tapped into the game's API and thus had much more information available to it than the human opponent. Denny Britz is a Stanford educated AI researcher most recently working at Google Brain. You can read his thoughts on the Dota 2 match here: Denny Britz
The Dota 2 match was not pre-announced - it was sprung on the world for maximum PR effect with no build-up. Turn out the bot had been tested on several high level players prior to the public match - so there was no risk involved of it losing when it went on TV.
Saying anything with "North Korea" in the phrase generates a lot of attention.
So you have Elon doing whatever he can to draw maximum attention to his claims that AI is dangerous - and now he's engineering publicity stunts to try to show it. This match was not in the same league as the Deep Mind vs human go player match or the much earlier Watson vs chess master match-up.
The question is - why does Elon put so much effort into calling for regulation when his private business is in the business of using AI to make money? I have several theories - mostly revolving around my belief that he is trying to position himself as the world's public face of "prudent" and "reasonable" AI use. IE so Joe Sixpack thinks "Well, if Elon Musk, of all people, says this here self-steering car is safe, I guess it is."
I dunno - your thoughts?
Friday night he sent out a new one claiming AI is more dangerous than North Korea - the same night he publicized the fact that an OpenAI bot beat the world's best Dota 2 players. Both the tweet and the OpenAI stunt smell of really cheap, low level fear mongering and publicity seeking. I was impressed with the Dota 2 achievement until I read the only piece I could find on the game - in which the author essentially claimed that the OpenAI bot probably "cheated" - in the sense that the bot likely was tapped into the game's API and thus had much more information available to it than the human opponent. Denny Britz is a Stanford educated AI researcher most recently working at Google Brain. You can read his thoughts on the Dota 2 match here: Denny Britz
The Dota 2 match was not pre-announced - it was sprung on the world for maximum PR effect with no build-up. Turn out the bot had been tested on several high level players prior to the public match - so there was no risk involved of it losing when it went on TV.
Saying anything with "North Korea" in the phrase generates a lot of attention.
So you have Elon doing whatever he can to draw maximum attention to his claims that AI is dangerous - and now he's engineering publicity stunts to try to show it. This match was not in the same league as the Deep Mind vs human go player match or the much earlier Watson vs chess master match-up.
The question is - why does Elon put so much effort into calling for regulation when his private business is in the business of using AI to make money? I have several theories - mostly revolving around my belief that he is trying to position himself as the world's public face of "prudent" and "reasonable" AI use. IE so Joe Sixpack thinks "Well, if Elon Musk, of all people, says this here self-steering car is safe, I guess it is."
I dunno - your thoughts?