OpenAI's Code Red, Sacks vs New York Times, New Poverty Line?

All right, everybody. Welcome back to the number one podcast in the world, the All-In Podcast. In the news, in your feed, we're ready to go. We've got the original quartet here. The band's back together. All right, first up on the docket, a code red has been called by Sam Altman. He sent a memo on Monday, told employees to stop working on side quests, you know, like ads, et cetera, and focus on the core, ChatGPT. The core experience, make it faster, make it better. And, uh, I think we all know why, because Gemini and Grok and Claude from Anthropic have been crushing it. ChatGPT-5, let's call it what it is, was a bit of a flop. It didn't por- perform to expectations. We discussed that a couple of weeks ago, or months ago. Anthropic is beating OpenAI in enterprise revenue starting this summer, and I talked about how, you know, previous episodes on the streets, I'm seeing more startups want to use Anthropic's API, uh, and also Google, Gemini's API, and, and they don't essentially trust OpenAI to not steal their business. So it's big changes right now. Here is, uh, a chart based on data from July. Most of Anthropic's revenue here is corporate. Most of OpenAI's, as everybody knows, is consumer. And here's another chart, OpenAI's new infrastructure deals versus revenue, and this is just for 2025 alone. A lot of deals being made, but competition is fierce. Chamath, your thoughts here on the game on the field. Code red for people who don't know in our industry is when everybody reports to the office and gets focused on one thing, and, uh, that's apparently what Sam's doing. How do you interpret it?

Look, I think that there are two things. Let's do the strategic lens and then the tactical lens. The strategic lens is that this is, uh, an incredibly vibrant and dynamic market, and I think it's too difficult and too early to pick winners, other than at the silicon layer, where largely that die has been cast. I think that it's going to grow, so we can talk about how there'll be more competition, but it's roughly NVIDIA plus AMD plus Google plus a bunch of inference silicon. So that's sort of that market, but above it at the model market, it reminds me, frankly, a lot of when we were building Facebook. I remember sitting around our senior executive team, six of us, looking at Myspace, who was an order of magnitude bigger than us, and at some point, we were like, "You know what? Our product is just fundamentally better than theirs." And they had 100 million plus users, and we had sort of 15, but we knew that we were eventually gonna beat them. Nobody else knew. And I feel like this market is similarly evolving, which is that you have these early winners, but there's still so much work to do. There's still so many consumer expectations to define that it's too early to know who's gonna win. And ultimately, what we are learning, especially as all these markets converge, is that distribution still matters a ton, which favors Google. It favors Meta, although Meta's quite behind, and now it will still favor OpenAI because they have 800 million monthly actives. So then the tactical thing is, what do you do knowing that this dynamic is set up to have a lot of competition? And I think what you have to do is streamline the focus and try to make a crisis out of every opportunity, because look, the companies tend to just grow. This positive entropy tends to cause people to hire at every level, and you look around, and there's thousands of people that you didn't know even six months ago. And so I think if Sam can use different points in time to tighten the core focus, they'll be better off, and I think that is what Google did a while back. If you remember the whole black George Washington thing, they were able to use that as a rallying cry to streamline the organization and to focus and to get their best and smartest people to work on the most highly leveraged tasks, and what you see now is an incredible overperformance from where they were. I've said this before, but Gemini is, is incredible. So I think that that's what's happening. Too early to call. It's a three or four horse race, and Sam needs to batten down the hatches, and I think he used this opportunity to stop a bunch of peripheral activities.