You're listening to Robert Wright's Non-Zero podcast. You're listening to Robert Wright's Non-Zero Podcast.

Hi, Stephen.

Hey.

How you doing?

I'm good, thanks.

Thanks for having me.

Well, it's a pleasure to have you. I'm very excited about this conversation. Let me introduce us.

I'm Robert Wright, publisher of the Non-Zero Newsletter. This is a non-zero podcast. You're Stephen Adler, publisher of the aptly named Stephen Adler's Substack newsletter at stephenadler.substack.com.

You are also somebody who until a few months ago worked at OpenAI. You'd been there since

2020. You decided to leave. And after you left, you expressed some concerns about the pace of

AI development. You were quoted as saying an AGI race after you left, you expressed some concerns about the pace of AI development.

You were quoted as saying, an AGI race, you know, meaning the race toward artificial general

intelligence, is a very risky gamble with a huge downside.

So I want to talk about that.

Now, you also are among a number of former Open AI employees who were part of an amicus brief, if that's the right term, kind of contesting the aspiration of Open AI to shed its nonprofit status. Now, since you filed a brief,

there's been a development. They say they are shedding it, but it's kind of, I don't know,

I kind of have my, my reasons for skepticism, but we'll talk about that and, and how happy you

may or may not be with what what they're doing um and you know

I I uh I also want to just just get into the whole uh uh the logic behind your concerns and

kind of the range of concerns because there's a whole bunch of different kinds of safety

concerns. But before we do any of that, I would like to ask you to recall a certain part of your

experience about Open AI, because you were there at a very special time, right? The, I think it was in 2022 that the,

...