Benchmark Breakfast #1 | Jack and Ev

Benchmark Breakfast is a recurring conversation between Benchmark partners Jack Altman and Ev Randle. It’s a window into the firm’s Monday no-agenda partner meetings, where five partners sit around and talk about whatever came up that week. In this inaugural episode, Jack and Ev dig into the Anthropic S-1: what the revenue-to-compute chart actually shows, why growth volatility and consumption-based revenue complicate the public market story, and whether code gen is the most impactful PMF moment in the history of technology. They steelman open source vs. closed source, run the numbers on what the inference cloud market actually looks like, and get into AI’s emerging cybersecurity threat — including “abliterated models,” open-weight models with safeguards removed that adversaries are already using. They close on why bio and robotics may ultimately matter more than all the software gains combined, and what Manhattan Project-scale investment in the physical world could actually unlock. TIMESTAMPS: (0:00) Intro (0:45) Introducing Benchmark Breakfast (2:09) Anthropic S-1 drops (3:59) The revenue vs. compute chart (6:19) Growth volatility and consumption-based revenue (7:11) The race for the best model (10:21) Code gen saturation and software factories (13:39) The AI-bottleneck trade (17:39) Code gen as the most impactful PMF in history (21:27) Durability is hard to call (23:36) Investing in the trend vs. calling the next one (26:09) Circular revenue vs. outside customers (28:15) Open source vs. closed source steelman (34:25) Own your intelligence (36:29) Open source as a % of frontier spend (40:52) AI and cybersecurity (44:35) Abliterated models (48:08) Bio, robotics, and physical abundance (51:52) Why physical feedback loops are slower (54:12) Ev’s dad and the promise of bio (55:13) Halloween plans and the iPhone Duo LINKS: https://x.com/jaltma https://x.com/everettrandle

Speaker names marked ? are inferred from speaking time and the show’s known hosts — they are not voice-verified.

• 0:00Jack Altman?

  And I think I could joke about being just like a little bit bored of.

• 0:04Jack Altman?

  It's like I interviewed.

• 0:05Speaker 01

  Just shit on, just shit on all the VCs you've interviewed.

• 0:07Jack Altman?

  I mean, we could joke about that, though.

• 0:09Speaker 01

  Yeah.

• 0:09Jack Altman?

  They're great. But I asked them all the questions and, you know.

• 0:12Speaker 01

  Learned all the lessons.

• 0:13Jack Altman?

  Maybe we come back, but it was time for, time for an update.

• 0:20Jack Altman?

  All right, well, this is the beginning of a new iteration of the pod where I guess, um, you know, at some point, I mean, I've learned a ton, but at some point you've interviewed a lot of VCs and you say, it's time to just talk to a VC that you talk, that you talk to you 10 times a day instead. But, uh, just changing it up a little bit. So now we're gonna do this thing, Benchmark Breakfast, we're calling it, where we're kinda gonna, I guess, I don't know, mock, you know, a little bit of the time that we spend on Mondays. And I guess, you know, what we, what we do on Mondays at Benchmark has basically been no agenda. We get together as, you know, the five of us and we sit around talking about whatever came up that week or whatever we're thinking about. And Evan and I were just saying, well, that'd be a, that'd be a fun new version.

• 1:03Jack Altman?

  And if I'm losing, if I'm losing steam on the current, you know, the current podcast, like, let's, let's, let's take a new rep. So I'm excited to be doing this.

• 1:10Speaker 01

  A hundred percent. And I feel like whenever we tell anyone, um, about our Monday morning or just our, our Monday partner meetings and the fact that they're typically five to six hours. The two, I mean, the two responses we get is like, that is absolutely insane.

• 1:24Jack Altman?

  A little long.

• 1:24Speaker 01

  Uh, it's like, you know, I, I, you know, people will be like, I can barely stand our hour-long partner meeting. Um, and then the second question's like, well, what's the agenda? And I think that's the really important thing about our Mondays is.

• 1:32Jack Altman?

  You review the pipeline.

• 1:33Speaker 01

  Yeah, yeah, yeah. Their agenda is like, you review a very strict list and, um, you know, it feels like you're, you're like a corporate vice president at Microsoft or something. Um, but our answer is always, there is no agenda. We, we just think about, you know, it's a continuous conversation, as I think our partner Eric says, from one week to the next. Um, and I think, um, you, you know, be, we thought it'd be fun to like, you know, open source or make public, um, uh, like the, the, the form factor of that conversation.

• 2:00Jack Altman?

  So one thing we talked about, obviously, is big news is Anthropic dropped an S1, which is, you know, pretty critical to the industry. So, um, as somebody who can actually read an S1 of our, of our mostly early-stage group, what was, uh, what was your reaction?

• 2:15Speaker 01

  Yeah, it's, um, it's, it's, it was amazing timing because I feel like leading up to the S1 day, you know, there was a lot of market turmoil, there was a lot of carnage in the public AI trade, a lot of thought about Anthropic, like, are they gonna IPO, are they gonna delay it? And then, you know, they come out with this, this S1 and on its face, this is everything that you want from a technology business. And it's funny because most of the numbers in the S1 had been leaked at one point or another. Like, there wasn't like that, that much new information.

• 2:46Jack Altman?

  You mostly could have stitched it together from, you know, if you believe the right things on Twitter. As you always say, Twitter's making us all dumber, but it was, it was almost there.

• 2:53Speaker 01

  But it was like, you know, if you really were good at parsing through, it was all basically there. But I think it was another thing entirely to have all the numbers put together in one place. And, and again, it was, it was like on its face, what you want out of a technology business typically is you want immense growth and you want operating leverage. Operating leverage is the sense that like, you have this big upfront investment in R and D, or sometimes with software businesses, you pay a ton for salespeople upfront and they take some time to pay themselves back. But you invest all this money upfront, you have fixed costs, and then your revenue grows so much that, and your incremental, your incremental margins are so good that you become very profitable very quickly as you scale.

• 3:33Jack Altman?

  Yeah.

• 3:34Speaker 01

  And we don't have to go through every number in stat. Obviously, it's all over X, it's all over the internet. But I think the most important visual from the S1, in my opinion, there's a chart that I'm looking at, um, that you guys can't see.

• 3:44Jack Altman?

  No, no, we'll put it up.

• 3:45Speaker 01

  Oh, well, well.

• 3:46Jack Altman?

  We have the power of technology.

• 3:47Speaker 01

  That's true. We have, we have the power of, you know, maybe even Claude.

• 3:49Jack Altman?

  Maybe we'll use AI.

• 3:51Speaker 01

  But the, the chart is, um, a, is showing the relationship between their revenue and their total compute expenses per quarter.

• 3:59Jack Altman?

  Right.

• 4:00Speaker 01

  And so it's both the compute expenses for serving customers, but then also training. And in Q1 of 2025, um, Anthropic had 400 million-ish, you know, they only, they do it in billions. And so it's only 0.4. So it's like 400 million-ish of revenue against 900 million of compute and infra expenses. And so that's like a point four X ratio, um, where, you know, compute is larger than revenue. Fast forward six quarters, in Q2 of 26, they had 11.6 billion of revenue against 5.6 billion of compute and infra expenses. So a 2.1X ratio of revenue to compute. It's like that, like.

• 4:36Jack Altman?

  Right.

• 4:37Speaker 01

  The, the entire concern that everyone had is like, oh, are the labs or all these AI companies, you know, selling a dollar for 90 cents?

• 4:44Jack Altman?

  Yeah.

• 4:45Speaker 01

  You know, or, or 50 cents, like, are they ever gonna be able to turn a profit? Anthropic in six quarters did what, you know, it takes a lot of software and technology companies, like a decade to prove out.

• 4:56Jack Altman?

  Yeah.

• 4:56Speaker 01

  Which is like, no, these are money printing machines.

• 4:58Jack Altman?

  They're money printing machines.

• 4:59Speaker 01

  In, in a huge way.

• 5:00Jack Altman?

  Yeah. And it's like, the revenue just turned out that it's big enough to make up for all these fixed costs of training. And another surprising thing, and you know, this has played out throughout inference and all sorts of areas of, of the AI stack. But the, the price per gigawatt has just surprised everybody. Like, I don't think anybody.

• 5:18Speaker 01

  Like the monetization?

• 5:20Jack Altman?

  Yeah.

• 5:20Speaker 01

  Yeah.

• 5:20Jack Altman?

  Yeah. And just like what the value of the frontier tokens are.

• 5:23Speaker 01

  That's right.

• 5:24Jack Altman?

  Just, has just worked, has just made it all work.

• 5:26Speaker 01

  It's made it all absolutely work. And I, you know, I, I don't think it, like, I, I think there was a lot of contention around the quality of the frontier lab business model that was somewhat, that was partially answered by the numbers in the S1. Like, you cannot say that, like, Anthropic's business model isn't working on like a per-unit basis or an aggregate basis. Like, they had 8% operate, like, gap operating margins in Q2 2026. So it's clearly working. I don't necessarily think that means that it's going to be a slam dunk out of the gates in terms of how public investors in the public markets value it. And if there's a ton of share price appreciation right out of the IPO. There's two reasons for that. The first is that there's still some battleground topics

• 6:08Speaker 01

  that people have uncertainty over that I think are somewhat fair. So the first is, you know, Anthropic in Q2 grew massively. They grew 145% quarter-over-quarter. So it's like insane sequential growth, but then they didn't grow as much in Q3. Um, and there's a lot of, you know, pretty good reasons for this, um, but they grew 45% quarter-over-quarter, which is still, which is still insane. But, you know, it's like the deceleration from 145% growth quarter-over-quarter to 45% growth. People are like, wait a minute, like, this doesn't, you know, with SaaS, with like, if you go look at like the net new per quarter of like CrowdStrike or Datadog.

• 6:41Jack Altman?

  Yeah.

• 6:42Speaker 01

  It's just like consistent upward trajectory. There's very little volatility to the growth rate. And then 80% of the revenue is consumption-based. And so there's, there's, there's not like these, you know, if you had like, you know, Snowflake or, you know, Databricks or one of these. It's like, a lot of that revenue is like multi-year, long-term contracts.

• 7:00Jack Altman?

  I mean.

• 7:00Speaker 01

  You have more like fixed visibility into the future growth.

• 7:03Jack Altman?

  One of the things that's been surprising to me, which I wouldn't have, I, I would not have guessed it, at least at this point, because like, you know, it's like the last generation of models pretty good now. But it still just seems that so much turns on who has the, the very best model at any given point in time. And the leading edge, even if it's by a little bit, and even if the last model could do most of the tasks.

• 7:23Speaker 01

  Yeah.

• 7:23Jack Altman?

  It just seems like it's worth a crazy amount of momentum.

• 7:26Speaker 01

  Yeah.

• 7:26Jack Altman?

  And we saw this go back and forth between the labs all year. And we saw it, you know, we saw it again sort of in the last month and then you saw it over the summer. And that, that part is surprising to me, that like, people will always seem to go for the highest intelligence, even with some price and sensitivity.

• 7:43Speaker 01

  You, you know, I think it, it, it probably is, is there's just so much surplus. And what I mean by that is like, whether, you know, whether you're paying $0 or $200 for like a Claude Max account, or you're, you're paying like next to nothing with, you know, some free service, the value that you can get as an individual.

• 8:03Jack Altman?

  Right.

• 8:04Speaker 01

  Is probably in like, at least in the thousands of dollars, if not tens of thousands of dollars.

• 8:07Jack Altman?

  This is like a Keith raboism that there's like, you know, two task types. There's the ones where it like, just needs to be completed and it needs to be right.

• 8:14Speaker 01

  Yeah.

• 8:15Jack Altman?

  And you don't get any more bonus points for getting any better past a certain point. And like, you know, that's often in a company context talked about, you know, with regard to like, what type of executive you should hire. Are you protecting downside? Are you looking for unkept upside? And like filing taxes, for example, you kind of either did it right or you didn't. At some point, I just have to believe those types of tasks will get saturated.

• 8:35Speaker 01

  Yeah.

• 8:36Jack Altman?

  But then if you think about writing code or doing sales or discovering, you know, drugs and biology or robotics, it's probably the majority of tasks, people understand it as unbounded upside.

• 8:51Speaker 01

  Yeah.

• 8:52Jack Altman?

  And in an unbounded upside world, if that's the mindset, you're probably always gonna buy the better thing. I mean, I, I do still think. There's, it's obvious that the sellers of tokens are doing really well.

• 9:03Speaker 01

  Hmm.

• 9:03Jack Altman?

  I still think that I believe it'll get there. But I don't think that in all cases, the buyers of tokens are proving to be that much more productive in all cases.

• 9:14Speaker 01

  That's, yeah.

• 9:14Jack Altman?

  Like, there's still a lot of hope going on.

• 9:16Speaker 01

  Yeah.

• 9:17Jack Altman?

  And, you know, obviously code, code got from, you know, it was actually a year ago, I think it was not being productive and people were buying a ton of it. Now I think it looks like it is totally productive.

• 9:27Speaker 01

  Yeah.

• 9:27Jack Altman?

  But I still think we have a way to go until like, you know, Purina Farms can buy AI and we're sure that it's a better business now. And I don't know like how that's gonna.

• 9:35Speaker 01

  You don't know about Purina Farms?

• 9:36Jack Altman?

  I don't know about Purina Farms. I'm just, I'm just guessing.

• 9:38Speaker 01

  Chick-fil-A has an amazing engineering team.

• 9:40Jack Altman?

  Yeah. And you think they're supercharged by AI?

• 9:41Speaker 01

  I bet they are. Everyone, everyone loves the Chick-fil-A.

• 9:44Jack Altman?

  I don't know. A little robotics might help them to make, but.

• 9:46Speaker 01

  Yeah.

• 9:47Jack Altman?

  But so I think there's a little diffusion that needs to happen on the consumption of token side.

• 9:51Speaker 01

  It's a hundred percent true. And I think so then, so you have like weirdness and growth is I think one reason why people will be cautious about the stock potentially. Um, there's debates. I think this is a core debate, which is how much code gen and related TAM. So like for code gen specifically, how long does the fusion take? And then how much just code gen TAM is left?

• 10:12Jack Altman?

  Yeah.

• 10:12Speaker 01

  And then what is the next thing?

• 10:13Jack Altman?

  I mean, how do you think about the code gen TAM? Like, you know, uh, the analogy could be something like in, in Zerp, you know, time.

• 10:23Speaker 01

  Yeah.

• 10:24Jack Altman?

  When everybody went remote.

• 10:25Speaker 01

  Hmm.

• 10:25Jack Altman?

  It was like Zoom got this absurd pull forward.

• 10:28Speaker 01

  Yep.

• 10:28Jack Altman?

  And it was basically like, everybody who wanted a Zoom account, got it. And so Zoom grew ridiculously, and then it had maybe not full saturation, but like roughly it got saturated. And obviously, at least in our corner of the world, everybody who might want to use AI programming has been using AI programming for a while.

• 10:46Speaker 01

  Yeah.

• 10:47Jack Altman?

  And then you have market share trades and you have those kinds of other things, but you don't get 10x year over year from that forever.

• 10:52Speaker 01

  Yeah.

• 10:53Jack Altman?

  But then you probably have a lot of other places in the economy where it's not even close.

• 10:57Speaker 01

  Yeah.

• 10:57Jack Altman?

  So I don't know how you think about like the, the code TAM.

• 11:00Speaker 01

  It's, it's funny because like I think of back to your saturation point of like things that are saturated with like a Zoom call. There's like a pretty easy saturation point.

• 11:09Jack Altman?

  That's like, can I get on, do my headphones correct correctly? But you could always have another coding.

• 11:15Speaker 01

  You could always have.

• 11:16Jack Altman?

  It's like you, you have like 30 agents going at once probably, right?

• 11:18Speaker 01

  All the time.

• 11:19Jack Altman?

  Yeah.

• 11:19Speaker 01

  All the time.

• 11:20Jack Altman?

  Building stuff.

• 11:20Speaker 01

  Just building stuff.

• 11:21Jack Altman?

  Yeah.

• 11:21Speaker 01

  Just, just figuring out how to be mean to you in our, in our partner meeting.

• 11:25Jack Altman?

  In the chat.

• 11:26Speaker 01

  In, but with the, um, but with Zoom, yeah, it's like with Zoom, it's like there, there's a very clear ceiling and asymptote. And I think the really interesting thing that we've seen with code is like, I think there probably is some asymptote, but the asymptote isn't like a crazy place. And what I mean by that is like, some of our newer companies, like the companies that have been started kind of post Opus four, five, I'd say. The way that they build software.

• 11:48Jack Altman?

  Yes.

• 11:48Speaker 01

  Looks almost nothing.

• 11:51Jack Altman?

  It's like nothing. It's.

• 11:52Speaker 01

  Nothing like how you built software in 2019.

• 11:54Jack Altman?

  I met up with one of our founders that we just seeded and he, you know, there's no other, or there's like one other engineer on the team right now. But we went for a walk and he's got like, you know, he's doing some stuff on his phone. What are you doing? He's like, oh, I just, I have like a few agents running and it's going to be done by the time we're done and all this stuff. And I'm just like, it's just a different, there's just a different way to do it now.

• 12:16Speaker 01

  And just like, like the idea of a software factory, um, and it's like a, it's like a new organizational structure for building software products and like a lot of it's not even human driven anymore.

• 12:27Jack Altman?

  So basically within Silicon Valley, there could still be many X to go.

• 12:31Speaker 01

  Yeah. And then I think it's like, if you think that like, okay, you either have to adopt that method of building software products or get left in the dust. You assume that on some timeline, the rest of the economy starts to at least look more like that than not. And that's where I think that like, I'm relatively, I don't know if you can go have like the 10x growth that you had again or something. But in terms of the durability of code gen, I think that you could continue to see like.

• 12:57Jack Altman?

  Double every year for a while. It's insane.

• 12:59Speaker 01

  Hundred percent. And I think, I think that, I think there's like enough to do that.

• 13:01Jack Altman?

  I mean, the other thing around that is, um, all of these startups and markets around code. The right answer has basically just been bet on all of it. And it's been funny because like in a lot of these sort of, let's say code related spaces. You know, you're like, ah, this can't possibly be that big or there's three of them or whatever. They just seem to keep going and it seems like, um, the market pull on some of the code adjacent areas is ridiculous. Which is one of the things we talked about earlier this week.

• 13:31Speaker 01

  Hundred percent. Yeah. It's, it's funny. So like a lot of people I think know about the sort of like AI bottleneck trade that happened in the public markets. A lot of people probably participated in, in some version of that. Um, but basically like after ChatGPT came out in November of 2022, um, there was like, you know, almost immediately AI demand first for ChatGPT and then other products. Started skyrocketing and people were like, okay, well, LLM training and inference runs on GPU. So it's my Nvidia. So I can be a goes up a ton. But then people were like, shit, what else goes into like an AI data center? But like, what else like goes on? Like what else will be impacted? What derivatives of GPUs are impacted? And so, um, I have like a list of these things that went through. So like 2023 servers and racks, Super Micro went up 10x, Dell went up a ton.

• 14:16Speaker 01

  Um, cooling equipment. So a company called Vertiv went up a ton. 23 and 24, Arista basically doubled in both 23 and 24.

• 14:23Jack Altman?

  It's crazy.

• 14:23Speaker 01

  Um, in 25, um, uh, power generation.

• 14:26Jack Altman?

  Yeah.

• 14:26Speaker 01

  Vistra went up three and a half X. Um, Oclo in nuclear, people were, you know, pulling forward the future of, of nuclear. Bloom Energy went up four X in 2025. Data center equipment, G Vernova went up two X in 25. And then 25 and 26, obviously we've seen memory and storage. So SanDisk, Western Digital, Micron, Seagate all went up at least three X. Uh, I think in 2025, and then now in 26, now the agents are now using more CPUs again. You have Intel up 120% this year. So it was like.

• 14:55Jack Altman?

  It's like falling an egg through a snake. It's just like.

• 14:57Speaker 01

  And it's like a new derivative, like every, like every, every, every year. Or like, there's like three new derivatives every year that you could have gotten like a multi bagger return on in the public markets.

• 15:07Jack Altman?

  Leopold spoke of this, you know, it's just like.

• 15:09Speaker 01

  The profit.

• 15:10Jack Altman?

  Yeah, the profit.

• 15:11Speaker 01

  But then there was like, so then in venture, the version of this that happened is sort of like what, what I think you could call like the code gen PMF derivative.

• 15:19Jack Altman?

  Yeah.

• 15:20Speaker 01

  So like, obviously like in, so, so ChatGPT comes out in November of 2022, but really this really kicks off in November of 25 when Opus four, five comes out and supercharges Claude code.

• 15:31Jack Altman?

  That's right.

• 15:31Speaker 01

  And that creates this mega product market fit moment for code gen. And then like, right after that, you had like basically, um, and again, this is sort of a parallel to all these public categories where you had vibe coding infra.

• 15:42Jack Altman?

  Yeah.

• 15:42Speaker 01

  So you had like backend as a service and databases, Superbase took off. Clickhouse in our portfolios continue to accelerate like crazy, um, uh, deployment hosts. So like Bursell, um, Railway, Cloudflare. Um, the inference clouds, obviously we're large investors in fireworks, which is growing like crazy, Base10, Modal. Um, data companies like Mercore, the benchmark portfolio.

• 16:02Jack Altman?

  Yeah.

• 16:02Speaker 01

  Code sandbox companies like Modal and E2B are growing like crazy. Web search, Exa, also benchmark portfolio company, code review.

• 16:10Jack Altman?

  I'm, I'm glad you're getting all the plugs in as we go.

• 16:12Speaker 01

  I know. I'm like, I'm just hitting all, but it's crazy. It's like, if, if you have, you know, 10,000 X more code in the world, all of a sudden like all of these derivatives.

• 16:20Jack Altman?

  Yeah.

• 16:21Speaker 01

  Of code generation also need to grow that much. And then, um, obviously you have a very excited, I know it's not public yet, but you have a very exciting investment in the CICD GitHub action space.

• 16:30Jack Altman?

  Yep.

• 16:30Speaker 01

  Um, because you, you know, again, durable execution with temporal, browser use APIs, and now supply chain security with, with, um, you know, folks like Chain Guard and Socket. Like all of these categories, every single one I just mentioned have companies in them that are doing like the 10 to 100 to 300. Like, you know, you start, you know, one year, 10, you go to 100 and then you go to like 300 or more.

• 16:52Jack Altman?

  Which never used to happen.

• 16:53Speaker 01

  Never used to happen.

• 16:54Jack Altman?

  And multiple are doing it in the same category at the same time.

• 16:57Speaker 01

  Yes.

• 16:57Jack Altman?

  This is what was wild. I mean, you know, it's like lovable and replit, for example, which I've obviously done great. But it's like, everyone was like, ah, isn't this, you know, isn't this club code? Isn't this codex?

• 17:08Speaker 01

  Hmm.

• 17:09Jack Altman?

  Yeah, it is. And they're going to do a version of it that is form fitted for a different purpose. And the whole thing is so big that it's just all going to work.

• 17:18Speaker 01

  That's right.

• 17:18Jack Altman?

  Which kind of goes to the Eric point of it's all gonna, where, you know, and I think, you know, probably the modification on that is in some of these categories where the thing, where the technology hits that hard.

• 17:28Speaker 01

  Hmm.

• 17:29Jack Altman?

  It's just, it's so easy to underestimate it.

• 17:32Speaker 01

  Yeah, it really is.

• 17:33Jack Altman?

  Yeah.

• 17:33Speaker 01

  It's like, I, I think like code generate, like AI code generation, you can sort of abstract most of the value generation in AI today to code gen. It's like, it's probably the most impactful product market fit moment in the history of technology.

• 17:49Jack Altman?

  Probably right. Yeah. I'm trying to think what else would even, I mean, it's in, it's on, it's on the order of iPhone moment or something like that.

• 17:54Speaker 01

  That's right.

• 17:55Jack Altman?

  Yeah.

• 17:55Speaker 01

  Except for, you know, because so much of it is in the digital, like because the infrastructure was already there.

• 17:59Jack Altman?

  Yeah. It can just go so much wider, so much faster.

• 18:01Speaker 01

  It can go so much faster.

• 18:02Jack Altman?

  Yeah.

• 18:02Speaker 01

  Whereas like Uber and DoorDash and, you know, these, um, companies that were built on the mobile internet that were in iPhones.

• 18:08Jack Altman?

  Yes. They were limited by the growth because they were.

• 18:11Speaker 01

  The growth of mobile, yeah, of mobile phones that had internet.

• 18:13Jack Altman?

  Yeah.

• 18:13Speaker 01

  Um, yeah, it's fascinating.

• 18:15Jack Altman?

  Yeah. And you know, it's funny like for like, um, knock on effects of this. I think part of what's weird for founders and VCs is when you've got a thread like this that's pulling so hard. Not just on code itself, but like you're saying, all of the like accoutrements around it that support it and come before it and come after it. And are just kind of all getting sort of yanked upwards. It's like other categories that are very good.

• 18:40Speaker 01

  Hmm.

• 18:41Jack Altman?

  It's like when the opportunity cost is people are like, well, I can get this ridiculous ride over here.

• 18:46Speaker 01

  Right.

• 18:47Jack Altman?

  That has a big effect.

• 18:48Speaker 01

  Huge impact.

• 18:49Jack Altman?

  And it's weird because at once it's making it so that venture demand in these spaces is so obscene.

• 18:56Speaker 01

  Hmm.

• 18:57Jack Altman?

  And then LPs are obviously doing great and they're funding it. And it's all part of like one rational circus.

• 19:06Jack Altman?

  But then you have these other categories and a founder's like, well, what the hell? I have a good business.

• 19:10Speaker 01

  Yeah.

• 19:10Jack Altman?

  It's hard to raise and all that thing. So it's, it creates a weirdness.

• 19:13Speaker 01

  I like, I had, um, you know, historically, I've, I've spent a fair amount of time in, in growth investing and, you know, as a growth investor, you, you keep like all these historical precedents of like the growth trajectories of all these amazing previous companies. And I'll go back and look and I think like data, I mean, this might be wrong, but I'm pretty sure data dog went like five to 18.

• 19:33Jack Altman?

  Yeah.

• 19:34Speaker 01

  Five to 18 and like five to 18 to like maybe 40 to 50 or something.

• 19:37Jack Altman?

  It's like now five to 18. It's like, I don't know if you can raise. I don't know if there's capital.

• 19:40Speaker 01

  Would you? I don't know. Like, would you look at a five to 18? Or what would it take for you to look at a five to 18?

• 19:44Jack Altman?

  You would need, in this, in this environment, you would need to believe in 18 to 100 story for the next year. Or you would just need to think that they were doing something that was so difficult and unique that it just had insane growth durability forever after that. But on the surface, that used to be enough where everybody would be super interested if all you heard was it's a software company going five to 18.

• 20:05Speaker 01

  Yeah.

• 20:06Jack Altman?

  Whereas now, if that's all you heard, you'd be like, well, why wasn't it five to 40?

• 20:10Speaker 01

  That's right.

• 20:10Jack Altman?

  You know, which is insane.

• 20:12Speaker 01

  It's so insane.

• 20:12Jack Altman?

  But it's all because of this code pull.

• 20:14Speaker 01

  That's right. A hundred percent. Which, you know, and like a related topic that we've been talking about is that there's sort of two archetypes, or there's almost like a continuum of like how levered you are to the trend. And then do you have like the old hallmarks of like durability, enterprise stickiness?

• 20:36Jack Altman?

  Yeah.

• 20:36Speaker 01

  Um, like some of the things that we loved about, you know, the, the, the scions of the last era, like data dog and CrowdStrike and, you know, even data bricks, even though they're not public, um, there, there seems to be some form of trade off between like how levered you are to the AI, like basically the code gen pull or just like the AI trade in general.

• 20:57Jack Altman?

  Yeah.

• 20:57Speaker 01

  Um, and some of the like more traditional, um, uh, like signs of durability and defensibility from like an enterprise.

• 21:06Jack Altman?

  Yeah.

• 21:06Speaker 01

  Context where it's like, oh, this is like three year long-term contracts.

• 21:09Jack Altman?

  Yes.

• 21:10Speaker 01

  And all these things. What have you, like, what, what do you think about sort of that trade off? Um, and, um, and if you're attracted to one or the other, or anything that you think about that?

• 21:20Jack Altman?

  Well, for, uh, and this might be wrong, but for me, um, I'm in a mindset right now where durability is really hard to call. Like I think in the last era by let's say the late 20 teens. I think you had a good enough feel for cloud and enterprise at that point that you could, not perfectly, but you could kind of name it. When you're like, that's durable or that's going to get competed. That's sticky. That's got a high switching cost. Like we got to a place where I think you could make those calls somewhat correctly because you had too many examples that looked like it. And you just knew how software worked. You also had situations where enterprises would look at software for

• 22:00Jack Altman?

  years before they were buying and like, I know I'm going to buy Workday, but I'm not going to buy it yet. But when I do, I'm going to use it forever. You just had dynamics like that where right now, I don't know that durability is nearly as easy to identify. You know, it's like even take like, take application software. I don't know. Six months ago, everybody was kind of like, it's dead.

• 22:22Speaker 01

  Yeah.

• 22:22Jack Altman?

  At the moment, I think people are like, it's alive and it's good. And why should the labs do it? And there's all these edge cases to do and they can actually like, you know, be value add to the intelligence tokens. And there's all these reasons to put your data in there and blah, blah, blah. And you need a UI and you know, that's great. And six months from now, people might flip again. It'd be like, never mind. Code got so good that that's all not true anymore and all these things. So calling the durability seems difficult to me.

• 22:52Speaker 01

  Yeah.

• 22:53Jack Altman?

  I think, um, you know, I was thinking about like, what is it about the five to 18? So I think if you are, if you are trying to identify the next code, the next thing that's going to pull like crazy. I think that's probably the reason that people get really attracted. You know, like as an example, I think we all believe robotics will work.

• 23:13Speaker 01

  Yeah.

• 23:14Jack Altman?

  We would all probably say it's not exactly working yet. We can't have a robot in here doing, you know, a bunch of things that, you know, a human could do. You can't really have one at home just yet. But I think we all believe we'll have one. Like you and I both love our Teslas. You like see the beginnings of it and you know, you go to these, you go to cool companies, like do you visit like Pi or something like that? You're like, this, you know, you could see the future.

• 23:37Speaker 01

  Yeah.

• 23:37Jack Altman?

  So it's going to come. And so you can bet ahead on things like that. Bio is another one where it seems like, you know, the smartest people at the lab seem to be saying it's going to work. There seems to be early indications that, you know, target identification and other things are possible that didn't used to be. And so I think that to me is more the spirit of investing at the moment is

• 23:58Speaker 01

  Yeah.

• 23:59Jack Altman?

  either on the trend or you're calling the next one.

• 24:01Speaker 01

  Yeah.

• 24:01Jack Altman?

  And obviously the returns are highest if you call the next one.

• 24:03Speaker 01

  A hundred percent. You're calling, yeah. I, I feel like that is, um, it's funny, like as soon as you can sort of put into chat, um, you know, like, you know, for, for obviously the segment of like, what are all the derivative things? I went to chat. I'm like, well, we've seen these work. What else, what other trends have worked? As soon as like chat knows.

• 24:21Jack Altman?

  That means it's

• 24:22Speaker 01

  The derivative trade.

• 24:23Jack Altman?

  Everyone's talking about it.

• 24:23Speaker 01

  Everyone's talking about it and that the companies that you could invest in are probably already at like a billion plus valuation or something like that.

• 24:29Jack Altman?

  Yes. And you can still make that call because you can believe that the market will still somehow just not appreciate how far it will go.

• 24:35Speaker 01

  Yeah. A hundred percent. A hundred percent.

• 24:37Jack Altman?

  And that's fine.

• 24:38Speaker 01

  It's super fun. Yeah. I think, I think on the, um, on some of the, the, like, a, a final thing on the, the difference or the comparison between, um, even both categories that are growing and like what is healthy growth versus, you know, riskier growth. And there, there's like two sort of, um, conflicting views. One is that like, well, if you want to see like a really healthy, like a, a sign of health for the AI ecosystem would be money that is not in the system, coming into the system.

• 25:08Jack Altman?

  And so this is You don't like when it just circles around?

• 25:10Speaker 01

  Yeah. So, well, I, I think there's two arguments, but I think the argument here is like, you know, okay, circular revenue and just like if you're, if the only people buying AI stuff are AI companies, that is like, you know, it's just, it's more and more leverage. Like that's more and more leverage on the system because if it doesn't work out, then it all comes crashing down. Whereas like, if large banks or large healthcare systems in like Ohio and New York and Florida are getting immense value and buying a ton of AI, that is really healthy because that's, that is like an independent, non-AI oriented end market.

• 25:41Jack Altman?

  And it goes back to, those are buyers of tokens.

• 25:43Speaker 01

  And those are buyers of tokens.

• 25:44Jack Altman?

  Which is ultimately what needs to happen here.

• 25:45Speaker 01

  Yes.

• 25:46Jack Altman?

  Spending a lot of money to make these tokens.

• 25:47Speaker 01

  Yes.

• 25:48Jack Altman?

  They better be useful.

• 25:49Speaker 01

  They better be useful. And, and like, I think AI companies are much more willing to be risking their token spending or token buying relative to like a big healthcare system.

• 25:58Jack Altman?

  Yes.

• 25:58Speaker 01

  With, you know, eight percent profit margins. They're not going to blow up their P and L.

• 26:02Jack Altman?

  Well, again, if you're, if you are an AI application company buying tokens.

• 26:06Speaker 01

  Yeah.

• 26:07Jack Altman?

  And then doing something to them and then giving them to somebody else. Somebody still needs to make people's lives better. Like at the end of the day, somebody needs to turn the token into a better life in a way that you couldn't have done without the token, or at least not as well.

• 26:19Speaker 01

  That's right. Yeah. But then the other, so the other side of the argument that I think also has a lot of merit is like, but look like the, the AI companies that are, you know, also buying a lot of these tokens, creating software factories. They're going to become the massive companies of the future because they're going to outcompete all the companies not doing this.

• 26:35Jack Altman?

  Yeah.

• 26:36Speaker 01

  And so like, this is like, when you look at Stripe, like Stripe didn't get, now they work with a lot of old kind of economy companies. But they didn't get big because they worked with old economy companies. They got big because they worked with DoorDash and Shopify and Instacart.

• 26:49Jack Altman?

  Lyft.

• 26:50Speaker 01

  And Lyft and like these, and Uber, you know, a lot of the sort of like Y C era.

• 26:55Jack Altman?

  Yeah.

• 26:55Speaker 01

  Internet native companies.

• 26:56Jack Altman?

  Just got big.

• 26:57Speaker 01

  That got massive. And like, that's what, and then once they were massive, then there, it was much easier to attract.

• 27:02Jack Altman?

  Yeah.

• 27:02Speaker 01

  Old economy companies because they were so like, they're like, hey, we just helped DoorDash.

• 27:06Jack Altman?

  Yeah.

• 27:06Speaker 01

  You know, kick the hell out of you. So like, you should probably use this as well.

• 27:09Jack Altman?

  Yep.

• 27:09Speaker 01

  And so I think that like, I, I have no problem. I've, I, I don't really have any problem with AI companies buying AI company, like, you know, AI tokens.

• 27:16Jack Altman?

  Yeah.

• 27:16Speaker 01

  Like the circularity, because I'm like, look, like these are company, they're just at the frontier. And just like internet native companies were, you know, at the frontier of the internet and older economy companies didn't really buy into what they were doing until it was too late and those companies were very large. I think the same thing's going to happen in AI.

• 27:32Jack Altman?

  Yeah, that's probably right.

• 27:33Speaker 01

  Yeah.

• 27:33Jack Altman?

  How do you feel about the, um, the closed versus open source stuff? Like we've talked about this a bunch.

• 27:38Speaker 01

  Yeah.

• 27:38Jack Altman?

  And obviously, you know, we are, we are exposed to a lot of the open source world through fireworks and, and other things.

• 27:50Jack Altman?

  What we've seen in general, I think, has been a rise in open source this year.

• 27:54Speaker 01

  Yep.

• 27:54Jack Altman?

  At the same time, obviously, the, you know, the frontier labs have gotten massive. And, you know, the price per token is very different. The use cases are similar but different. So anyway, like, what's your latest on sort of open and closed stuff?

• 28:09Speaker 01

  Yeah. I mean, if you'll, if you'll entertain me, I, um, I, I feel like we probably have a similar synthesis of what we feel now. But I think what would be interesting to do is like, maybe we steel man each side of the argument.

• 28:23Jack Altman?

  Great.

• 28:23Speaker 01

  And then like that, I think, once you have the steel man for each argument, because I think there's a sort of like a lot of like, there's a lot of talk online about open source versus closed source, which I think is sort of off track.

• 28:33Jack Altman?

  Right.

• 28:33Speaker 01

  From like the reality on the ground. So I think if you steel man both sides, you can actually understand like, okay, well, what's the synthesis that probably blends the two?

• 28:41Jack Altman?

  Great.

• 28:41Speaker 01

  Steel man. So like, I'll, I'll do, I'll do closed source, you do open source.

• 28:44Jack Altman?

  Yeah, you go first.

• 28:44Speaker 01

  Um, so I, I think, so the steel man of, of the closed source argument, and this is, and what, what we mean by this is like, you know, why would a company, um, you know, um, get inference from a closed source frontier API rather than from an open source API inference provider like a fireworks using Kimi K three or any of these other models? Um, US or Chinese open models. Um, and people say, well, look, intelligence is fungible, or it's like sort of a commodity. And if we take that seriously, that idea that like, you know, a unit of intelligence is this fungible thing, there's a lot of history and a lot of good literature about who wins those types of markets. And, um, when you're, when you're optimizing for the cost of a fungible output.

• 29:27Speaker 01

  And it's, um, oftentimes the winner is something we call a low cost producer. And so, uh, the defining feature of one of these low cost producers is that you are structurally, as a company, you have a lower cost to deliver a comparable outcome. And so you can either charge lower margins, or you can charge lower prices at the same margin as competitors, or, uh, you can earn higher margins than your competitors at the same price. And so two famous examples in tech, SpaceX is probably the easiest one to understand. SpaceX has the lowest cost to them in order to launch a kilogram into space.

• 30:00Jack Altman?

  Yep.

• 30:01Speaker 01

  In the world. No one can launch a kilogram into space as cheaply as SpaceX can. And that gives them a pseudo monopoly on launch, uh, at least at, at the same level, unit level of, of capability. Um, a, a previous one in computing is AWS. They had massive aggregated demand that created economies of scale. They're able to amortize a lot of investments over a massive customer base, which gave them lower cost relative to competitors in the CPU cloud business. So for AI, the equivalent, um, here would be, how do you become the low cost producer of a unit of intelligence? So we could say this is like a given level of capability, latency, context length and reliability for some task.

• 30:44Speaker 01

  Like who can produce that task at a given level of capability, the lowest reliably? Uh, and I think proponents of the closed source argument would say the leading frontier labs have the best models in the world, many of which are not released to the public, are not released via API. And they're using those models and pointing them at their infrastructure to improve their serving efficiency. Um, and, and, you know, co-design sort of their serving infrastructure with the actual models that they're training. Um, they have the most aggregated demand, they're the largest, and so they have economies of scale. Like the data center space that they're buying is cheaper for them on a per unit basis than anyone else, because they're the largest purchasers of compute.

• 31:22Speaker 01

  Um, and if all of these are true, if you have economies of scale, if you have like, if we are in some weak form of RSI right now, where the models are actually designing, like, you know, actual inference efficiencies, um, at, at, at any sort of rate that, that is compounding. Uh, that means that you can have a low cost producer of intelligence in form of, in the form of an AI lab. And that means that this argument that like open source equals cheaper is a misnomer. And is not necessarily true, because at the same level of margin, if this was true, a frontier lab at any point of the intelligence curve, like the capability curve, would be cheaper than a inference provider serving on an open source system. So that's sort of the economic argument for closed source, and then I'll let you take open source.

• 32:08Jack Altman?

  Okay. So, I mean, there's, I'll, I'll start with the economic piece. And obviously, as you know, I mostly, I mostly buy it, but economically, one difference is that because the, you know, to, there's the model cost, and then you got to do the inference on it.

• 32:24Jack Altman?

  The model in theory has some training that's been amortized into the price. And so to the extent that the labs want to be profitable or have, you know, enough margin that they make up for it, they do need to do something to make up for the training costs. Versus, you know, if you're using Kimi, you don't, you don't, the model, the model is free. So there's no amortization there. And then obviously you still need to run the inference and I probably, you know, it's probably right that, you know, on a per sort of inference unit basis, the labs are still going to have the advantage over the inference companies, but you maybe in theory that's, that's made up. The other thing is, you know, do the models care to take their margins down to that level? And is that actually worth it? Or is that really expensive for them to do? And so, you know, the.

• 33:09Speaker 01

  Then they don't, then they don't have as much money to train.

• 33:10Jack Altman?

  They don't have as much money to train. And also, it's like, you know, if they're really focused on the, the real frontier, and they're like, you know what, use your open source for the taxes. Use it for recommending which, you know, iPhone you should buy, and should you get the Duo or not? And we're going to go solve biology, we're going to go do robotics, and it's just not worth it to go optimize two models ago to go do the basic stuff. And so there's some, there's basically some argument along the lines of

• 33:42Jack Altman?

  we're going to get saturation on many tasks. And as open source of the future is better than, you know, Opus five, five and Astra are today, that's going to be good enough for most tasks. And the model, the, the frontier labs are not going to care to go reduce their margins, optimize two models ago. They just need to keep their eyes on the frontier.

• 34:04Speaker 01

  Hmm.

• 34:04Jack Altman?

  And so maybe there will be some, some space for that. But what I would say probably is more than the money pieces. And I, I wish I actually had, because to your point, the, the, the specific numbers on the margin are where it all turns on.

• 34:18Speaker 01

  Right.

• 34:19Jack Altman?

  You know, so like, I, and I don't have those on hand, but I think more important is, do you want to own your own intelligence? And this is like the Alex Karp stuff.

• 34:26Speaker 01

  Yeah.

• 34:27Jack Altman?

  Of just like, you cannot trust, as an enterprise, you cannot trust any company to own all of your data.

• 34:33Speaker 01

  Yeah.

• 34:34Jack Altman?

  And to be fully reliant on any company, and you need to own your own models, you need to own your intelligence. And I think there is an argument to be made there. And I think for certain industries, that trade is going to be worth it.

• 34:44Speaker 01

  Yeah.

• 34:45Jack Altman?

  And particularly as the open source models get good enough, even if you're not saving any money, you're in control. And that's what's something.

• 34:52Speaker 01

  And then even like, think about the branding of like, if you, let's say you're all, all your, all your tokens are anthropic tokens. And all of a sudden, anthropic enters your space.

• 35:01Jack Altman?

  Yeah.

• 35:02Speaker 01

  And then your customers are like, well.

• 35:04Jack Altman?

  Yes.

• 35:05Speaker 01

  You know, like, what, what model are you using?

• 35:07Jack Altman?

  Yeah.

• 35:07Speaker 01

  And you're like, a hundred percent of our models that we're using is anthropical, like, then why don't we just go to anthropic?

• 35:10Jack Altman?

  Yeah. If you're Eli Lilly.

• 35:11Speaker 01

  No special sauce.

• 35:12Jack Altman?

  If you're Eli Lilly and all of your data lives in anthropic.

• 35:15Speaker 01

  Yeah.

• 35:16Jack Altman?

  And then anthropics like, hey.

• 35:18Speaker 01

  By the way, BTW.

• 35:19Jack Altman?

  We got some drugs coming. You're like, okay, well, I don't feel so good now.

• 35:22Speaker 01

  Right.

• 35:23Jack Altman?

  So I think there are some reasons that people won't want to do that.

• 35:25Speaker 01

  Yeah.

• 35:25Jack Altman?

  And then I think, you know, there's, there's situations where you can post-train on open source, and that can be useful. And if you've got a lot of data, that data can outperform, you know, the extra intelligence that you might get from the next model. And so you'll see some of that too.

• 35:38Speaker 01

  I feel like we hear that pretty consistently from, from founders, um, that are using, you know, they don't necessarily, aren't even selling an AI product, but just have a big product corpus and they're using AI throughout their products. Where they're like, look, you can really like, if you really spend the resources to fine tune, and you have a really good trace data from the task that you're trying to do, like, yes, like maybe vanilla, Kimi K three or whatever, isn't quite as good as, as, as whatever the model, but like, you can actually get a really small model, like a, like a really, really small open source model. And if you fine tune the hell out of it.

• 36:08Jack Altman?

  Yeah.

• 36:09Speaker 01

  Like it, it gets to like, you know, a K three or a Sonnet level performance, and it is dirt cheap.

• 36:14Jack Altman?

  Yeah.

• 36:14Speaker 01

  It is dirt cheap. I think, yeah, we, I, I, yeah, we had a, a, a large public CEO come in and talk to us about that. It's like, there's clearly all these, these, these use cases where that's true.

• 36:23Jack Altman?

  Yeah. But then, you know, it's like. The the the nice thing is, we know the answer. The market's telling us the answer. And the answer is that the inference companies in combination are doing mid-single digits worth of total revenue.

• 36:36Speaker 01

  Yeah.

• 36:37Jack Altman?

  That's a reasonable proxy for open source, because that's that's where it's running. And that's, you know, whatever percent of the labs companies.

• 36:45Speaker 01

  Yeah.

• 36:45Jack Altman?

  And you can argue for that there'll be some mix shift, but what we're basically seeing is that, you know, open source might be might be five, it might be twelve, it'll be some percentage like that unless there's some change.

• 36:55Speaker 01

  That's right.

• 36:55Jack Altman?

  Of the overall spend, and that's what'll be there.

• 36:57Speaker 01

  Yeah, I think, yeah, so that, so I think the synthesis between these two is that, you know, again, shout out Eric Vishria. Um, it all works.

• 37:04Jack Altman?

  It all works.

• 37:05Speaker 01

  Um, but also, like, the way it all works is like, these, these actually, like, they're probably not even rival risk to each other. And I, like, I don't know if this would be the case, but like, also, I'm sure the frontier labs, you know, they're probably even happy about there being an open source ecosystem, because it probably creates all, like, from an antitrust perspective.

• 37:21Jack Altman?

  Hundred percent.

• 37:22Speaker 01

  If it's like, oh, they're getting too powerful, it's like, no, there's a healthy open source ecosystem, where people can also use, and that they're like, I think that's probably also healthy, I have no idea if that's an actual thing that they think about, but like, right now, so, like, this will be excluding the.

• 37:34Jack Altman?

  Yeah, I'm not sure they think about it that way, but yeah.

• 37:36Speaker 01

  Yeah, the, um, but if you look at, so like, right now, if you look at the, we'll exclude the hyperscalers, because we have no idea how much, um, open source inference they're they're doing. Um, but if you look at, you know, just the inference clouds, as we've talked about, and then you look at just OpenAI and Anthropic, so this is very imperfect, but it gives us some, some okay proxy. By the end of this year, I think the main, the top three inference, three or four inference clouds will probably do like six billion of run rate.

• 38:01Jack Altman?

  Yep.

• 38:01Speaker 01

  Ish.

• 38:01Jack Altman?

  Yep.

• 38:02Speaker 01

  Um, and then, you know, let's say, OpenAI and Anthropic will do like one hundred and seventy billion of run rate, um, ish. Um, if they, you know, if their growth kind of extrapolates from public disclosures. So that's like three-ish percent thus far.

• 38:13Jack Altman?

  Right.

• 38:14Speaker 01

  Between the inference clouds and the frontier, and the top two frontier labs. If the growth continues.

• 38:18Jack Altman?

  Yeah, maybe if you add the clouds in, maybe it's, maybe it's double at most, or something like that.

• 38:22Speaker 01

  That's right, so maybe, maybe it's like, so we can say like five to ten percent.

• 38:25Jack Altman?

  Yep.

• 38:25Speaker 01

  Um, in terms of revenue.

• 38:27Jack Altman?

  Yep.

• 38:27Speaker 01

  For open source versus versus closed. But if you have the growth that the labs have had, and that people, I think, expect, again, this isn't a guarantee, so this isn't a projection, but if you do have the growth for three to four years on the frontier side, and the labs combined earned a trillion.

• 38:40Jack Altman?

  Yeah.

• 38:41Speaker 01

  And open source takes ten percent of that, that is like a hundred billion dollars of revenue. And these inference, like, the inference market has consolidated. Like, there's really only like three players.

• 38:49Jack Altman?

  That's right.

• 38:50Speaker 01

  Like, we have fireworks.

• 38:51Jack Altman?

  Yep.

• 38:51Speaker 01

  We have Base Ten, we have Modal, and it's like, I mean, there's, there's other ones that are good and all that stuff, but like, they really are like the three champions.

• 38:57Jack Altman?

  That's right.

• 38:57Speaker 01

  Of the market, and I think we're gonna see more consolidation to those three than less. So you have like probably two or three companies at that time, splitting a hundred billion dollars of revenue, which is incredible.

• 39:08Jack Altman?

  Which is incredible, I mean, it means one of those companies could be worth a hundred billion or more.

• 39:11Speaker 01

  Probably both. You know, you have like a forty billion dollar company, with, like, even if it's fifty percent gross margin, you know, it's probably a two hundred billion dollar company.

• 39:18Jack Altman?

  Yeah.

• 39:18Speaker 01

  You know, you have like hundreds of billions of enterprise value for open source, and then you have the trillions, uh, in frontier labs, and like, everyone is very happy with that, like, that's,

• 39:26Jack Altman?

  Yeah.

• 39:27Speaker 01

  That's huge, like, humongous companies all around, and it's through the synthesis of like, open source will always, on a revenue basis, be some percentage of frontier, and whatever that percentage is, like, the, the, there's so much room for, for the inference platforms to grow.

• 39:41Jack Altman?

  By the way, and then we can, we can move on from this, but, um, it, it's like, a couple of years ago, I feel like it was, um, it was, there was a worry of like, man, it seems like AI is either gonna wake up the Terminator, Sam God, and everything's gonna be really bad. It's like too good.

• 39:57Speaker 01

  Yeah.

• 39:58Jack Altman?

  Or it's just like not gonna matter that much, crypto all over again, head fake. And at the moment, and, you know, this is why I'm, I'm, I'm happy to see the market reacting positively to Anthropic. It's like, at the moment, like, unless, you know, RSI takes us into a weird place, like, we might actually, actually be splitting the middle landing, where like, the labs are extremely important, but they are not sixty trillion dollar companies, where it is not useless technology that doesn't help anybody. Like, that, that, that's a nice thing.

• 40:31Speaker 01

  Seems like the happy path.

• 40:32Jack Altman?

  It seems like that, it, I feel like my, my odd, my, my P happy path is higher than it used to be, and that's awesome.

• 40:38Speaker 01

  Amazing.

• 40:39Jack Altman?

  Which is amazing.

• 40:39Speaker 01

  That's amazing.

• 40:40Jack Altman?

  It's amazing.

• 40:40Speaker 01

  Um, do you wanna talk about, um, bio? Obviously, you've, you've got, I, I feel like you've been really interested in robotics and bio.

• 40:47Jack Altman?

  Yeah, I have been, I mean, the, I was gonna go to the dark place of cyber.

• 40:51Speaker 01

  We can do cyber first.

• 40:52Jack Altman?

  Let's do cyber.

• 40:52Speaker 01

  Let's do cyber.

• 40:53Jack Altman?

  Because like, that would be, that is one of the scary things, and this, this last week, you know, there was like the PHP thing.

• 40:58Speaker 01

  Mm-hmm.

• 40:58Jack Altman?

  And people are very afraid, and understandably so. It does feel like we are in a point in time where one of the risks that is most imminently being expressed is a cyber risk.

• 41:09Speaker 01

  Yep.

• 41:10Jack Altman?

  Where, you know, we, we, we've spoken to some like prominent CEOs in the last few weeks, who have told us that like, they've pointed frontier models inward and said, hack us. And the model's like, I'm gonna hack the shit out of you, and it does.

• 41:23Speaker 01

  And, yeah, they, they said that, like, day after day, they were getting company-ending vulnerabilities surfaced.

• 41:30Jack Altman?

  Yes.

• 41:30Speaker 01

  That like, if it had been actually hacked from an external force.

• 41:33Jack Altman?

  And these are great companies with great founders, and, you know, with great engineering teams.

• 41:38Speaker 01

  Like tech-forward companies, these aren't like, you know.

• 41:40Jack Altman?

  Yes.

• 41:41Speaker 01

  Old companies.

• 41:41Jack Altman?

  And, you know, obviously, when you attack internally, you have crazy advantages that outside attackers don't have. You've got all the data, you've got the code base, it's just a different thing. But still, it seems like, you know, fear could be on the horizon, and then, you know, you see all this stuff. So, um, obviously, it's probably, it's a good time to be a security company.

• 42:01Speaker 01

  Yep.

• 42:02Jack Altman?

  But, um, it is also a scary, it's a scary, you know, first exposure of real risk, I would say.

• 42:07Speaker 01

  Hundred percent.

• 42:07Jack Altman?

  Yeah.

• 42:08Speaker 01

  Yeah, the, um, uh, and I think that the really scary thing is that

• 42:14Speaker 01

  cyber attacks, at least for the public, I feel like cyber attacks historically have been sort of an abstract thing, where like, at worst, they're like, your, your data was, was like, your password for this website was, was hacked. And then you get like some notice in the mail, and you like, you know, you change your password, but like, nothing actually happens.

• 42:32Jack Altman?

  Yeah.

• 42:33Speaker 01

  And recently, um, there's been a lot creepier, and like more things that feel really personally intrusive, that are both now possible and have actually been happening.

• 42:43Jack Altman?

  Hmm.

• 42:43Speaker 01

  And so like, there, like, I pulled up like a few examples where, like, there, um, there was a researcher who used Claude to disable a webcam's recording indicator.

• 42:54Jack Altman?

  Wow.

• 42:55Speaker 01

  Um, so started recording without, like, the light coming on, um, and access the microphone and the firmware.

• 43:03Jack Altman?

  That's scary.

• 43:03Speaker 01

  All those things, like, a French supermarket suffered an attack that manipulated refrigeration controls and actually spoiled food. So I think that the issue is that, like, now, like, the systems that are now, that we thought were secure, because these are open source systems that have existed for fifteen years with humans.

• 43:20Jack Altman?

  Yeah.

• 43:20Speaker 01

  Looking to patch them and hack them in all sorts of ways. Um, like, these, these critical systems to like firmware and hardware.

• 43:28Jack Altman?

  That's what's tough.

• 43:28Speaker 01

  The overall internet, like, these are now getting exposed. That, that only have been able to be exposed via these AI models.

• 43:34Jack Altman?

  Yeah, and you can patch the software, but can't really patch the hardware.

• 43:37Speaker 01

  Yeah, and so it's like, you know, people are like, oh, like, you know, uh, like, update your smart TV, and it's like, what the fuck?

• 43:43Jack Altman?

  Yeah.

• 43:43Speaker 01

  My TV?

• 43:44Jack Altman?

  What's wrong with the TV?

• 43:44Speaker 01

  Like, what, like, you know, and they're like, oh, disconnect it from the internet, and I'm like, well, that's really scary.

• 43:48Jack Altman?

  If you have a Samsung Frame, smash it immediately.

• 43:50Speaker 01

  Yeah, like, yeah, smash all of your smart devices in your home, or disconnect it from the internet, so I'd be like, it just feels so much more widespread and intrusive, um, than, than before, which is why it's now sort of, like, I, I, you know, I definitely bought Ubiquis.

• 44:04Jack Altman?

  Yeah.

• 44:04Speaker 01

  Um, a couple of days ago, like, I, I wanna go through the whole, um, actually, Andre Karpathy has a really good blog post on, um, digital hygiene.

• 44:11Jack Altman?

  Yeah.

• 44:12Speaker 01

  And just things you should do, and I feel like, for the first time, given the, the parallelism and the scale that people can use, and just like the extent to which these things that were seen as like strongholds, like PHP.

• 44:23Jack Altman?

  Yeah.

• 44:24Speaker 01

  Um, are, are being hacked, it's like, okay, this is, this is for real.

• 44:27Jack Altman?

  Yes.

• 44:27Speaker 01

  Um, I think that the, like, really interesting thing, so there, there's one other, um, piece of the cyber thing, is this idea of obliterated models.

• 44:34Jack Altman?

  Mm-hmm.

• 44:35Speaker 01

  So an obliterated model is an open weight model that has its internal sort of safety guards removed, and, um, to help, help with its work, because, you know, ultimately, like, when the hugging face incident happened, you know, they, they had some complaints where they're like, we had to use an open source model because we couldn't turn on, we couldn't get, you know, the frontier labs to turn off the safety guards in time to help us with the hack. Um, and so obliterated models are literally open weight models that have no safety guards. So it's very scary, and like, I think one of these is actually what, um, some, some groups in China have been using, I think it's called Artex, um, uh, have been using to, to, you know, commit some of these hacks globally. It's also an incredible, um, defense tool for enterprises, because you take an obliterated model,

• 45:19Speaker 01

  you do what that CEO told us, and you just say, hack us, like, hack us to death, but like, obviously, like, you're the one using it, and so nothing actually happens from it. And like, that is, I think, going to be like an incredible tool, is like, you need these models that will, because guess what, the adversaries have obliterated models. The adversaries are not working with safety controls.

• 45:38Jack Altman?

  Right.

• 45:38Speaker 01

  You need to have that, that, you know, you need to see the tools that they're using and use it on your own infrastructure, and I feel like we're gonna hear a lot more about this idea of obliterated models.

• 45:46Jack Altman?

  By the way, I think, you know, like, the, um, the tech world, which I think is mostly a good thing, you know, at first, I would say, reacted negatively to like pacing the frontier type stuff, and there's much more of just like a let it rip mindset.

• 45:59Speaker 01

  Mm-hmm.

• 45:59Jack Altman?

  Which is, you know, it's like very libertarian, tech-forward good, like, I think that's basically a good thing. But this is, the, the, this security dynamic is an instance where it not going too fast would be very helpful.

• 46:10Speaker 01

  Yeah.

• 46:10Jack Altman?

  Because you actually do want the defenses to penetrate before we have, you know, four generations forward of models.

• 46:17Speaker 01

  That's right.

• 46:17Jack Altman?

  Because you do need things to get a little time to harden against this stuff.

• 46:21Speaker 01

  It is, but, but, and this is the crazy thing, is like, how long do we think it's gonna take to harden all this stuff?

• 46:27Jack Altman?

  It's gonna take a long time.

• 46:28Speaker 01

  Like, think about like, I don't know, think about like a hospital system in like West Virginia.

• 46:32Jack Altman?

  Yeah.

• 46:32Speaker 01

  How long is it gonna take for them to, because I think if you talk to some of these cybersecurity professionals, they're like, we basically have to re-architect. This is like the biggest code modernization project in all of history.

• 46:43Jack Altman?

  That's what we need cognition for.

• 46:44Speaker 01

  That's right, that's what we need, that's what we need cognition for, um, you know.

• 46:49Jack Altman?

  But, I mean, I think like, um.

• 46:50Speaker 01

  I was gonna try to do like an ad read there.

• 46:52Jack Altman?

  Yeah. Cognition.

• 46:53Speaker 01

  Cognition, update your code.

• 46:55Jack Altman?

  Yeah.

• 46:55Speaker 01

  In security.

• 46:56Jack Altman?

  That's good, but it's like, um, it, it's gonna take a long time, but I mean, the, the hope I have is that, you know, like we were talking about, you know, in the cloud era, it's like people bought, you know, Workday and Salesforce, some of these are really elongated timelines.

• 47:11Speaker 01

  Yeah.

• 47:12Jack Altman?

  They were slow, and, you know, if you were talking to a government or a hospital or a law firm, it was like thirty-six month cycles on some of the stuff.

• 47:19Speaker 01

  Mm-hmm.

• 47:19Jack Altman?

  That's not really how it's going right now on the purchasing of software side.

• 47:22Speaker 01

  Right.

• 47:22Jack Altman?

  I think people are pretty alert.

• 47:24Speaker 01

  Yeah.

• 47:24Jack Altman?

  So the hope would be that people see the security stuff for real, people are making it a top priority, and everybody reacts a lot faster than you think. I still don't think that that means it happens in three months.

• 47:35Speaker 01

  Yeah.

• 47:36Jack Altman?

  But maybe in twelve months, people could do a lot.

• 47:39Speaker 01

  That'd be amazing.

• 47:40Jack Altman?

  People are, I think people see the urgency.

• 47:42Speaker 01

  Yeah.

• 47:43Jack Altman?

  Hopefully. Yeah, maybe as a, as a last topic, the, um, I think what's cool is, the bio thing is interesting to me, because.

• 47:53Jack Altman?

  I've kind of been of the view, I still am of the view, that all the software improvements that we're making are awesome. But there is a ceiling to how much that can improve the world. And if you like, touch some grass and you get out of San Francisco, and you're like, what will actually make the world a better place? It's not that we're shipping software way faster and way more easily. It's things like people have better housing, or people live longer, or people live healthier, or people have food, or people have transportation. Like, it's physical things. I just think there's a, there's a, there's a limit to what you can get. When it's all sort of in the, in the screen.

• 48:32Speaker 01

  Yeah.

• 48:32Jack Altman?

  And bio to me would be like, one of the very best possible things. If we're discovering drug targets and, you know, finding amazing new, you know, drugs quickly. Or at all, that would be, what are you laughing about?

• 48:47Jack Altman?

  So anyway.

• 48:47Speaker 01

  Oh, you were on a good tirade.

• 48:48Jack Altman?

  I was on a good tirade. So anyway, but I think that it is both really promising, and it seems like it's on the horizon.

• 48:54Speaker 01

  Yeah.

• 48:54Jack Altman?

  Same with robotics, by the way.

• 48:55Speaker 01

  Yeah.

• 48:56Jack Altman?

  Those are the, those are the two to me where you see the, you see the labs both talking about it. Building teams around it, purchasing huge amount of data around it. You know, they're writing about it publicly. And so I think that that to me is likely an indicator that it's there. I mean, I haven't yet seen the labs be wrong.

• 49:15Speaker 01

  Yeah.

• 49:15Jack Altman?

  When they say something's coming.

• 49:16Speaker 01

  That's right.

• 49:17Jack Altman?

  Getting the timeline right is really difficult, you know, with self-driving it took, now it's fully here, but it like took a while.

• 49:22Speaker 01

  Yeah.

• 49:23Jack Altman?

  And, you know, self-driving, you know, Tesla's obviously dominant.

• 49:25Speaker 01

  Yeah.

• 49:26Jack Altman?

  But a lot of that was because of just like infinite data.

• 49:28Speaker 01

  Right.

• 49:29Jack Altman?

  And once you get enough data at these things, like, maybe it just works.

• 49:32Speaker 01

  I think that's, and that, that's the, there's this really weird sort of like inverse relationship between, I completely agree, I feel like. You, you, you, do you want like, the, you want the rest of America? Who doesn't have like an amazing, I think, view on AI, or at least a lot of people don't. Throughout the world and in America, don't have an amazing view on AI. Like, do you want to change perception of AI? Like, solve disease, save people's lives, help people live longer?

• 49:56Jack Altman?

  Yeah.

• 49:57Speaker 01

  Build more housing so housing prices are lower, like create physical abundance?

• 50:01Jack Altman?

  That's what makes a better world.

• 50:01Speaker 01

  That's what makes a better world.

• 50:02Jack Altman?

  Yes.

• 50:03Speaker 01

  Um, and, and the, the weird thing though is that like, there's almost this inverse relationship with those things, the things that, like, when you actually touch. The physical world could actually make the physical world a better place. And the things that have been able to get better very quickly from AI. And so if we take like, you know, like AI just tends to improve fastest. When it can generate large amounts of trustworthy feedback cheaply, I think is like a good way to abstract it. So if you look at like code, um, you know, it's like this incredible area for this because it's extremely verifiable. Like, you know, you can do a unit test, you know if it works or not. You can scale up experiments extremely cheaply, they're like infinitely scalable, it's all in the digital space. And so when you think about like Claude Code, or like how Cursor improved its models very quickly, or even ChatGPT.

• 50:47Speaker 01

  When they started getting feedback from users, all of these were in a digital realm where there was an API. And customers could come on, uh, at the click of a button. And start, you know, basically contributing into the leviathan of training data for these labs.

• 51:02Jack Altman?

  Yeah.

• 51:02Speaker 01

  And in the real world, robotics and now bio, the tough thing is that like you, you have real bottlenecks to that feedback loop. Like in robotics, obviously like, robots are kind of expensive, like they're expensive to deploy, expensive to maintain. There isn't like a big customer base that can press a button. I mean, you could order a robot and you could start doing these things and Pi has an API now and you could do the Sunday beta program or any of these things.

• 51:22Jack Altman?

  Yep.

• 51:23Speaker 01

  But it isn't like ChatGPT where it's like you can get a hundred million people in a few months.

• 51:26Jack Altman?

  Right.

• 51:26Speaker 01

  Via website.

• 51:27Jack Altman?

  Yeah.

• 51:28Speaker 01

  And so like, that's just very different. Bio, you also have natural rate limiters, like. The equivalent of like running a unit test in biology is a physical experiment.

• 51:35Jack Altman?

  Yeah.

• 51:36Speaker 01

  And like that, you know, like it could, you know, like, and even that experiment might only answer like a small part of the question you actually care about.

• 51:43Jack Altman?

  And a lot of things do just take time.

• 51:44Speaker 01

  They just take time. Like, you know, it's like, I don't know, a physical experiment, like even if you speed it up, um, you know, making your own.

• 51:50Jack Altman?

  No, you have to parallelize. That's what it is. And I think so in bio, it's like, you can imagine a virtual. Sort of, or, or like a real lab, but where you're sort of, you know, kind of run somewhat autonomously or at massive scale.

• 52:00Speaker 01

  Yeah.

• 52:01Jack Altman?

  Where you're just doing a ridiculous number of thoughtful, you know, trials at once.

• 52:06Speaker 01

  Hundred percent.

• 52:06Jack Altman?

  I think that could be like, you know, you, the, the, time will hold you back, but the parallelizing will still speed things up quite a lot, I think.

• 52:13Speaker 01

  Yeah. And I, but I think like, this, and this is, I, I know like you, you have an investment that you recently did in, in this general space for robotics. Um, we've, we've been very interested in things happening in bio as well, but like, the infrastructure to. I mean, we've obviously seen it with the AI labs as well, like Merckor as a business has grown to an unbelievable amount of scale.

• 52:31Jack Altman?

  Yes.

• 52:31Speaker 01

  Because there's basically, there's just no limit.

• 52:33Jack Altman?

  Very unintuitively, by the way. Everybody kind of thought that like the data thing would run out and they would buy the data and then they bought the data and now they can move on.

• 52:39Speaker 01

  The day's gone. So like, turns out.

• 52:41Jack Altman?

  It's just turned out to be wrong.

• 52:42Speaker 01

  Yeah, reality has a surprising amount of detail, as they say.

• 52:44Jack Altman?

  That's right.

• 52:44Speaker 01

  And, um, there's just so many different things that these labs want to train on. There's so many different data types that emerge. And I think like, how much we've seen in the digital space, in the digital data space, we're going to see. I think you have to see more, like Brendan, I, I don't know if he's the first one to come up with this, but Brendan Foodie at Merckor says. Um, the internet was a one-time subsidy on the product, on the production of L L Ms.

• 53:06Jack Altman?

  Yep.

• 53:06Speaker 01

  And these physical spaces, they don't have as much of that one-time subsidy, they don't have the internet. Um, they have some data, there's a lot of bio data, there's some robotics data, but like nowhere near.

• 53:14Jack Altman?

  And you also need a ridiculous diversity of data, it turns out, to build these models well.

• 53:18Speaker 01

  Hundred percent.

• 53:18Jack Altman?

  And so it's just, it's just not going to happen overnight.

• 53:20Speaker 01

  But I think it's, it's going to be like a civilization scale thing.

• 53:23Jack Altman?

  Yes.

• 53:24Speaker 01

  Uh, maybe that's maybe economy scale, civilization scales may be a little dramatic, but like to get these things where they need to be.

• 53:29Jack Altman?

  Yeah.

• 53:29Speaker 01

  It's going to take, I think it's going to be like, you know, Manhattan Project level scale.

• 53:33Jack Altman?

  Yes, it'll be even, it'll be a bigger, it'll be a bigger effort to make it actually happen.

• 53:38Speaker 01

  Yeah.

• 53:38Jack Altman?

  But if you imagine what's really possible with, you know, good AI robotics. It's probably bigger than, it's probably bigger than all the digital stuff so far, if it actually works.

• 53:48Speaker 01

  I guess, I mean, even like, if you think about the GDP impact of, I mean, maybe it could get scary because if you extend lifespans, then you have like a even more growing. Population, but we don't have to get too much into macro.

• 53:58Jack Altman?

  Yeah.

• 53:58Speaker 01

  But like, you know, it's like, you know, not to get too personal with it, but like, again, my dad got diagnosed with a very advanced form of cancer. I look at like the survivor cohort rates.

• 54:07Jack Altman?

  Yeah.

• 54:07Speaker 01

  It was like, you know, eight percent by year two or whatever. Yet immunotherapy drugs produced in the last 10 years have given him like a 10 plus year prognosis.

• 54:16Jack Altman?

  Yes.

• 54:17Speaker 01

  Which would like, like, you know, if he got the disease 12 years ago, he's dead.

• 54:21Jack Altman?

  Yeah.

• 54:21Speaker 01

  And now he's in remission. It's like, like, and we are just at the start of probably a hundred to a thousand times that amount of. Things happening where it's like every single disease area, every single cancer. Uh, like, you know, we've solved everything else from AI, and when we solve bio. Through all the scale of investment we're going to need to make, it's going to make probably a much, much larger, um, impact than any of these other things.

• 54:47Jack Altman?

  Well, I think that was enough for this one.

• 54:48Speaker 01

  Yeah.

• 54:49Jack Altman?

  This was lovely.

• 54:49Speaker 01

  This is great. What are you going to do for Halloween with the family?

• 54:51Jack Altman?

  Uh, we're gonna, we're gonna, we're gonna go trick or treating in an extremely dense neighborhood.

• 54:57Speaker 01

  Wow.

• 54:57Jack Altman?

  And I'm gonna just try not to lose the kids.

• 54:59Speaker 01

  That's great.

• 54:59Jack Altman?

  That's always my goal.

• 55:00Speaker 01

  Is it in, in NSF?

• 55:01Jack Altman?

  In the, on the peninsula.

• 55:02Speaker 01

  On the peninsula?

• 55:03Jack Altman?

  Yeah.

• 55:03Speaker 01

  Hmm, yeah.

• 55:04Jack Altman?

  Yeah, not.

• 55:04Speaker 01

  Which you're welcome to come to. Well, Brett and I are going to be in, in San Francisco.

• 55:08Jack Altman?

  Why don't you guys come down?

• 55:09Speaker 01

  We, we might, we might come back.

• 55:10Jack Altman?

  We'll talk about it.

• 55:11Speaker 01

  And then do we, do we get a duo review, or is that next episode?

• 55:13Jack Altman?

  The duo, the duo is unbelievable.

• 55:15Speaker 01

  Do you have the duo?

• 55:16Jack Altman?

  Yeah. The duo is unbelievable, with the, with the open.

• 55:21Speaker 01

  Wow.

• 55:21Jack Altman?

  It's incredible.

• 55:22Speaker 01

  You liking it? So this is day two?

• 55:23Jack Altman?

  This is, this is, yeah, I'm 24 hours in.

• 55:26Speaker 01

  And you love it?

• 55:26Jack Altman?

  I'm obsessed with it.

• 55:28Speaker 01

  Amazing.

• 55:28Jack Altman?

  But if, uh, next episode, let's see if.

• 55:30Speaker 01

  Yeah, yeah, we'll see, we'll see the.

• 55:31Jack Altman?

  We'll see if I've returned it.

• 55:32Speaker 01

  Yeah.

• 55:32Jack Altman?

  Yeah.

• 55:32Speaker 01

  All right.

• 55:33Jack Altman?

  I'll see you in an hour.

• 55:33Speaker 01

  Thanks, Jack.

• 55:34Jack Altman?

  Goodbye.

• 55:34Speaker 01

  Bye.