Sam Altman’s notion of AGI
9 snippets from a wide-ranging conversation between podcaster Lex Fridman and the CEO of OpenAI
A firing and re-hiring from OpenAI, video generation model Sora, and a lawsuit from Elon Musk.
A lot has happened for Sam Altman, CEO of the AI frontrunner company, since the first time he appeared on the Lex Fridman Podcast almost exactly a year ago (episode #367) and until his revisit this week.
In a wide-ranging conversation spanning almost two hours, the pair touches upon the aforementioned events as well as others, including upcoming models, Altman’s perspectives of Artificial General Intelligence (AGI), and how such powerful technology should be responsibly governed.
I’ve watched the episode and picked nine highlights/key quotes from the conversation that you can read below. Link to the particular section in the interview after each quote.
1) Compute as the important commodity in the world:
“I think compute is going to be the currency of the future. I think it will be maybe the most precious commodity in the world, and I think we should be investing heavily to make a lot more compute.” (1:10:01)
2) People who create the underlying data for OpenAI’s models should be compensated:
“Do people who create valuable data deserve to have some way that they get compensated for use of it? I think the answer is yes. I don’t know yet what the answer is.” (40:14)
3) A new Large Language Model will soon eclipse GPT-4:
“I think at this point we are on an exponential curve, and we’ll look back relatively soon at GPT-4 like we look back at GPT-3 now.” (47:46)
4) ChatGPT’s hallucination problem won’t be fixed this year:
“It is obviously an area of intense interest for us. I think it’s going to get a lot better with upcoming versions, but we’ll have to continue to work on it and we’re not going to have it all solved this year.” (53:55)
5) On state actors trying to infiltrate OpenAI and get access to their models:
"They’re trying," but their origin and other details he will not share at this point. (1:28:10)
6) Acknowledging that the name OpenAI isn’t aligned with the company’s shift away from the founding idea of full openness:
“Going back with an oracle, I’d pick a different name.” (28:29)
7) His notion of Artificial General Intelligence (AGI).
“Does the global economy feel any different to you now or materially different to you now than it did before we launched GPT-4? I think you would say no. (..) People define AGI all sorts of different ways. (..) For me, I think that should be part of it.” (1:34:39)
8) No single person should be in control of an AGI.
“I think I have made plenty of bad decisions for OpenAI along the way, and a lot of good ones, and I’m proud of the track record overall. But I don’t think any one person should [be in control of an AGI], and I don’t think any one person will.” (1:40:41)
9) On the personal risk of being so strongly associated with a technology that some people have a strong fear of:
“I think some things are going to go theatrically wrong with AI. I don’t know what the percent chance is that I eventually get shot, but it’s not zero.” (1:12:26)
Any other quotes that you found especially notable? Let me know in the comments.
Episode page here. Thanks to Lex and Sam for doing the interview, and remember to support the show’s sponsors to keep it running.
excitech will share insights about technological development 2-4 times per month. To be notified of other posts, sign up below.