The A.I. Civil War

Andrew Heaton (0:07)

Hello, and welcome to The Political Orphanage, a home for thinky people, and increasingly, bots posing as thinky people. I'm your host bot, Andrew Heaton. And I try not to read headlines, in general. But when I do, I find that a lot of them involve AI. And a lot of those headlines are fights over AI. Begging the question, what are they fighting about exactly? When I hear things like, Anthropic pushes back on OpenAI's Q security protocol. What the hell does that mean? Who is fighting who? What are the battle lines here that I'm meant to care about?

When I was a kid, the big tech fight was Microsoft versus Macintosh. And it pretty much boiled down to, do you like Windows 98? Or do you derive fundamental elements of your personality from Apple products because Steve Jobs looks cool in a turtleneck? And then you put Apple logos on the back of your computer because you want to advertise their products for free under the mistaken belief that computer hardware plus corporate branding constitutes some kind of community.

The computer fights, with which I am familiar, basically boiled down to two factors. Whose product actually works better? Blackberries or iPhones? Or whose product is cooler? Laptops or Chromebooks? Whereas the present AI fight that we find ourselves in feels far more apocalyptic, doesn't it? It doesn't feel like team dork versus team hipster. It feels like choose us or everybody literally dies. As though we are not picking software, but high stakes, eschological commitments. Not an AI competition, but an AI civil war.

In that war, who are the players? What are the boundary lines? What are they fighting about?

And what are the stakes? That's what we're going to discuss right here today on The Political Orphanage.

My guest today is the great Justin Robert Young. He is the host of The Attention Mechanism alongside Andrew Main, who is OpenAI's original prompt engineer and former science communicator. And they cover the technology, economics, politics, and increasingly strange, factional battles surrounding the world of artificial intelligence. Justin is also a very good friend of mine and a guest on this program, as well as the host of Politics, Politics, Politics, and along with Jen Briney, co-host of We're Not Wrong. Hello, Justin.

Justin Robert Young (3:16)

I do too many shows. That's what I have realized. I need to stop doing shows. Any new show I have an idea, the answer is no.

Andrew Heaton (3:24)

I think part of the problem is that we figured out a method that works for us, which is doing podcasts. And on top of that, I don't know, probably 50 to 70% of my social life is just podcasting with my friends. And so it kind of, it has a way of bubbling out.

Justin Robert Young (3:40)

The unique male millennial desire to make your entire living talking to your friends.

Andrew Heaton (3:45)

Yes. Although in this instance, it is of benefit to me and listeners because you are friends with people that are at OpenAI. You're working with OpenAI.

And so you've got more access to what's going on in the guts of the AI revolution than just regular garden-variety nerds like myself. So I thought we would talk about that today. And let's kick it off by establishing what the fault lines are. What are the different visions that AI companies have for how you're going to go?

Justin Robert Young (4:15)

Before we go any further, just so I do not have any stolen valor here, I am a podcast producer for the OpenAI podcast. So I am not into the guts of anything. I am in the building probably more than the average Joe that does not work at OpenAI. But most of my knowledge that we're going to talk about today is freely obtained in the way that anybody else who pays attention to this world can find it. And I would encourage people to, because I do think that the fault lines we're going to talk about today are going to be with us for the next probably at least two decades.

Andrew Heaton (4:50)

Okay. Well, noting that, that you're not an executive at Anthropic, I do think you're very knowledgeable on this topic. And it's something that I hear people bring up in conversation, which is they'll hear some fight going on between Anthropic and OpenAI. And other than they are competing companies, we don't really know what any of that means. So what are the, what is the lay of the land in this battle that you note will take two decades to resolve?

Justin Robert Young (5:14)

The two companies that are at what they call the frontier of labs that are developing this technology are OpenAI and Anthropic. Anthropic was founded from people that left OpenAI. So this is essentially a group of people that was assembled out of Google's lab, DeepMind, and various other people throughout Silicon Valley in the original incarnation of OpenAI that was founded by Elon Musk and Sam Altman, along with Greg Brockman and a few others.