As research around AI models heats up, many are following the words of Sam Altman, who has become, in some ways, the face of this community.
Altman, now famous for his work on OpenAI models, sat down with Garry Tan at Y Combinator to talk about his journey, and how it coincided with the rapid advances we’ve made in IT over the past few years.
The two started out with a sort of Voltaire ‘best of all possible worlds’ pronouncement driven by Gary’s question: is this the best time ever to start an AI company?
“It’s the best time yet,” Altman said, expressing hope that humanity will have even better times in the future.
They also discussed the importance of going out on a limb and doing things your own way.
“One of my takeaways, at least for myself, is that no one is immune to peer pressure,” Altman mused, “and so all you can do is … pick good peers.”
Who Will Rescue the Grownups?
Conviction, he suggested, is paramount, as is ignoring some of the pushback you get for pursuing innovation in your own ways. From that, there was this sort of Gen X coming of age realization that the two talked about, that we’re living in an age à la William Golding, in which no one’s really driving the bus:
“Before … I really had this deep belief that somewhere in the world, there were adults in charge, adults in the room, and they knew what was going on, and someone had all the answers,” he said of his early life. “And you know, if someone was pushing back on you, they probably knew what was going on. And (there’s a high) degree to which I now understand that, you can just do stuff. You can just try stuff. No one has all the answers. There are no, like, adults in the room that are going to magically tell you exactly what to do, and you just kind of have to, like, iterate quickly and find your way.”
Throughout, Han referenced Altman’s recent essay on AGI that I covered a couple of weeks ago. (Check that out here.)
In his broadside, he talks about a multigenerational march toward progress, and again references the value of conviction.
“We need to act wisely, but with conviction,” Altman writes.
And his commentary on the inevitability of the modern struck me, illustrated by this quote: “Nobody is looking back at the past, wishing they were a lamplighter.”
More From the Interview
Anyway, back to the interview: in starting out as founders, Altman suggested, it’s helpful to have money. Here’s part of his narration of the beginning of OpenAI:
“It was definitely helpful that I could just like, write the early checks for open AI, and I think it would have been hard to get somebody else to do that at the very beginning,” he said. “And then Elon (Musk) did it a lot, at a much higher scale, which I’m very grateful for. And then other people did after that. And there are other things that I’ve invested in that I’m really happy to have been able to support, and I think it would have been hard to get other people to do it.”
Going back to the early chronology of the tech age, and seeing connections, Altman, a self-professed ‘AI nerd’, talked about the shared legacy that he perceives in the advances of computing:
“There’s this long history of people building more technology to help improve other people’s lives,” he said. “And I actually think about this a lot: I think about the people that made (the early Macintosh). And I don’t know them. You know, many of them are probably long retired, but I am so grateful to them, and some people worked super hard to make (the computer he grew up with) at the limits of technology. I got a copy of that on my eighth birthday, and it totally changed my life, yeah, and the lives of a lot of other people, too. They worked super hard. They never, like, got a thank you from me, but I feel it to them, yeah, deeply. And it’s really nice to get to, like, add our brick to that long road of progress.”
Now, here’s one of the high points for those who really want to analyze where AI is going:
Hidden in this talk is a critical look at the structural framework that some apply to AGI now.
The Race to Artificial General Intelligence
First, Altman talked about how the term AGI itself is “overweighted” and carries too much abstraction and broadness.
With that in mind some of the top people have instituted what they call a five-stage system enumerating levels of progress in AI.
Stage one, as Altman explained, is chatbots – stage two is reasoning. We’re already there, especially with chains of thought in OpenAI’s o1. Stage three is an agent, where agentic AI is just now taking off.
Here’s where it gets really interesting. Stage four is an innovator, where the computer can go out and explore some unknown phenomenon and creates solutions. Stage five is a more, in Altman’s words, “amorphous” situation where the agents can do this at the scale of an entire organization. Think of a company or institute that was founded, staffed and maintained only by AI entities, with no humans involved!
The last words:
“Bet on this trend,” Altman told students, entrepreneurs and investors.
It really is an interesting time. Will we see all-AI businesses start to spring up soon?
