NYT DealBook Summit – 2023: Elon Musk

twitter.com

1 point·TheAceOfHearts··3 comments

Some notes on his comments relating to AI:

At 42 mins Elon predicts AGI, an AI capable of matching any top tier human at anything, discovering new physics / maths, etc. in 3 years.

At 46 mins Elon is asked about the copyright issue as it relates to AI training, he says everyone is ignoring copyright. Says it won't matter because by the time the courts start to litigate the issue we'll already have the AI god.

Forgot to note the time, but he also says that he's curious about the Sam Altman situation and that he doesn't know why he was fired. He suspects that OpenAI must have something big happening internally to justify the firing attempt.

Another point that gets made is that many events develop in real-time on the Twitter platform, which means that their data is really valuable for training AI and figuring out what's happening around the world.

Is there any reason to believe he knows more about AGI than actual experts in the field?

I don't know. One argument in his favor would be that he's well connected and employs a lot of AI experts across his companies. This probably means he's had a chance to speak with a lot of experts to help him develop an informed opinion.

One argument against him is that he previously overestimated the arrival of self-driving cars. We'll see if he turns out to be correct or if he's being overly ambitious once more.

From what I've read, even AI experts vary wildly in how soon they predict the development of AGI.

Reply on news.ycombinator.com