Cerebras IPO, WarshTime, General Catalyst Ad Reactions | Andrew Feldman, Amy Reinhard, Ben Hylak, Doug O'Laughlin, Eric Vishria, Steve Vassallo

You're watching TVPN.

Today is Thursday, May 14th, 2026.

That's right, May 14th.

We are live from the TVPN UltraDome, the temple of technology.

The fortress of finance. The capital of capital. The temple of capital.

Oh my goodness.

Hey, Ben.

There's Ben, indispensable.

We got multiple Bens over here.

We have a great show. It's Cerebris Day on the show. Cerebris IPO. We'll talk about that. Semi Analysis has a fantastic deep dive on the company. We'll go through that. Doug O'Loughlin from Semi Analysis is joining the show. And then Andy Feldman, the founder and CEO of Cerebras, is joining the show. But we have lots more folks joining. Amy Reinhardt from Netflix, the president of ads. Can you imagine? I didn't think, you know, you think about presidents, the president, you think president of the United States. I think president of ads. That's right. Ben Hylak. Eric Vischer, Steve Vasalo. We got a bunch of folks coming on the show today. It's going to be a fun one. So there's a ton of news. Let's start with Cerebras. The IPO has gone spectacularly well. Cerebras doubled their valuation basically overnight. Brandon Gurel had the good fortune of writing up some of the details of the Cerebras news in the newsletter today, tbpn.com. You can go sign up.

Yeah. And right now, it's sitting at a 64 $54 billion market cap. And a lot of the prediction markets, they didn't even have a category above $50, right? A lot of people were just kind of trading or betting.

And when I wrote the newsletter Friday, Monday, I said a $50 billion IPO and was sort of being optimistic. And it beat those expectations, which is great news. They deserve it. It's a true overnight success. We'll show you some charts of the valuation. Lots of troughs of disillusionment, but Andrew and the team powered through and wound up finding the perfect application for their technology at the perfect time during a mega cycle, which we'll go through. So chip design company, Cerebras, if you're not familiar, they make a big, big chip, big chip company. Instead of—

The biggest chip.

Instead of taking the wafer, putting a bunch of chips on it, cutting it up into smaller chips, they use the whole wafer. It's a genius idea. It's one of those simple ideas taken deadly seriously in some ways. But it's trading at $350 a share on its first day of public trading, which values the company much higher.

$300 now.

$300.

Yeah.

$300. OK. It was.

OK. It was at $350.

It was up at $350. It has since sold off slightly. And they raised around $10 billion. I think they were targeting $6 billion at one point. They've upsized that. I think it was a $3 billion raise initially. Have a good amount of money in the bank now. The price in this IPO has been literally up only. On Monday, the price range was $150 to $160. Then they raised it. That was up from $115 to $125. And today, we're seeing much higher prices.

Go back to that picture.

Who's in the picture? At the NASDAQ?

The picture at the NASDAQ.

Of the Cerebras team standing on stage?

Someone should make a set in LA. You know, they have those fake private jet sets? Imagine if entrepreneurs could have a set where they put their logo in the background, like they're hitting it with a hammer and there's confetti going everywhere.

Yeah, yeah, but it's for your course. Yeah, exactly. And you walk right from there to the Lambo.

And you have 1,000 students in your mastermind.

Yeah, yeah. No, I had this idea back in the day when— do you remember the ice cream museum, this whole thing?

Oh, yeah.

So there was this trend— I mean, really bad news for the museum industry, but they're getting eaten alive. And so some entrepreneurs— I think they did very well— started something called the Ice Cream Museum. Which was not really a museum in the sense of like a presidential library or, you know, the Norton Simon or the Getty or the, you know, Natural History Museum. It was more of like an experiential place to go and hang out, good for first dates, good for, you know, taking kids maybe. And they would maybe give you some ice cream, but most of the museum was just like, very Instagrammable things. So there would be like a ball pit or a bunch of raining confetti and stuff and a huge fabricated statue of ice cream that was not a piece of art that would be sold.

Sprinkle ball pit.

There you go. That sounds real. I don't know if that is real, but it sounds very believable. Yeah, they had that.

They had that?

Okay, yeah. So, and there were a number of other kind of copycats that were trying to jump on and do like, oh, we'll do like the waffle museum or something or the pancake museum, you know, 'cause they just wanted to cash in, and my idea was just the museum of Instagrammable objects, and so it would have all of those. So there would be a private jet set, and then there would be a Lamborghini set, and this one would fit right in. So it's just, they have a big pink wall, so you can go take the pink wall photo, and then there's a beach, and then there was a gym with fake weights, so you could go and look like you're maxing out and benching 500 pounds, and so it just says, "Bring these clothes, or we'll have them for you," and then you move from room to room taking the ideal dating profile photo.

Yeah, gender reveal.

Yeah, exactly, exactly. Oh, you had kids? Here, you in the hospital. You can live an entire life through this fictional museum of Instagrammable objects. More of a meme than a real business idea. But—

No, John. The Museum of Ice Cream now has 7 locations.

OK, so they're cooking.

They're global.

They're global. They're doing well. Well, let us know in the comments.

Anyway, nice little tangent there.

Miami is the capital of gimmick museum. Gotta do a whirlwind tour. Anyway, let's go back to the serious stuff. Cerebras, it's a complicated company because they are so deep in the AI supply chain, but we'll break it all down for you. So Semi Analysis has a fantastic deep dive. It's a longer read, so we're not gonna go through it all, but there are some very interesting tidbits in here that we can sort of summarize and contextualize for you. And then of course we'll be talking to Doug O'Loughlin from Semi Analysis at 12:30. So, There's a bunch of interesting takeaways, some really solid positives. Cerebras chips work, which was something people were not expecting for a while. There was a lot of FUD around this company. Just the idea of like, oh, that'll never work. Like what if the architecture changes? What if we go away from transformers or something? What if we need something quote completely different? Or maybe like the yields will never work because there was this idea that if you're using the entire wafer, typically as you're, as you're etching the chips onto the wafer, sometimes there's little defects and it's not a problem if you're gonna break up a wafer into like 64 chips 'cause you just throw away one. But if there's one defect on basically every wafer, well then your yield is gonna be super low. We talked to Andrew about how he solved that by creating redundant cores and they don't actually activate all the cores. And so they sort of built in that redundancy and got through that. But that was an early critique of the strategy. So they work. Yeah, and you can use— You can use it.

You can use Cerebras chips today as—

In Codex 5.3 Spark. And so they are very fast. And I think the most important thing that Semi Analysis points out is that token consumers, customers, businesses have shown this revealed preference for and a willingness to pay for speed. And they sort of contextualize it and they quantify it based on their own usage and their experience with Anthropic's Opus models. So Opus 4.6 fast mode famously, I like that they use famously because it's like famous, to like 100,000 people, but famously charges 6 times the price for 2.5 times the interactivity, although it's now under 2x faster. So effectively you're paying 6 times the price for 2 times the speed. That's disproportionately more money for what you're getting. You would think you'd pay 6 times the price for 6 times the speed potentially. But there were a lot of questions about would people really pay for more, that much more for faster models, faster inference? And Andrej Karpathy, Sam Altman was saying like, do you want faster models or smarter models? And he was like, I think, and Sam's point was sort of like these models are very intelligent, but using them faster is sort of more of a magical superpower. And Sam was, I felt like Sam was sort of leading it towards like speed is really important as the next leg up on productivity. And Audrey Karpathy was like, no, I just want smarter. I'll just let it run overnight. I don't mind that. But that's not what everyone is feeling. Some people, especially the Semi Analysis team, leaned more towards interactivity or speed over raw intelligence power.

Well, yeah. And then there's the other aspect, which is just capability, right?

Mm-hmm.

Capability, speed, and intelligence.

Yeah, I think—

Like that's a, the question people have had is like, okay, what is, is there a 250 IQ model or is there just a much more capable model?

Yeah, the unhobbled one.

That can use tools more efficiently and is really quick.

Yeah, and that's actually important to Cerebras because as we'll get into, the chips do face a hurdle with scaling in terms of longer context windows, all that stuff. But Semi Analysis shared their breakdown. They were run rating $10 million on AI spend in April. And they said that April was the peak, which is interesting because that was sort of, I would've expected a straight line, continued growth, but they might've, you know, been really pilled, tried it all, and then eventually said, oh, okay, well for this we can probably use a cheaper model. We can probably optimize. We don't want, you know, 95% of our revenue going towards tokens. We want a little bit of margin to pay, our team and obviously it's a business, they need to make profits. So Semianalysis was spending 80% of their AI spend on Opus 4.6 Fast. And so they were willing to pay that 6x, like 80% of their spend, disproportionately more, even though when their sort of expectation, as they put it, was that they would always want the smartest model, they would be very cost conscious, they were, in reality saying, I'm going to hammer fast mode, I want to spend on fast mode. And then I think the price was significant. And so there's probably sort of a renegotiation about when is the right time to use fast versus when do you want to leave something running overnight. But OpenAI is clearly very pilled on Cerebras. Cerebras has a big 750-megawatt deal with OpenAI and the chips are already serving GPT-5.3 in Codex under the name Spark, as we mentioned. And I've used it. You should use it. It's a very interesting experience because I think a lot of people have interacted with LLMs and chatbots and they're sort of used to this token streaming in and it's sort of cute because the phone vibrates and it feels like you're talking to someone who's typing, but it's way better when you just land on a Wikipedia page, the full thing loads and you can just scroll however much you want. And that's the experience that I think people want and will demand across everything, especially if they're firing off a coding task, they just want the code immediately. And so, You can also just go talk to the model like it's ChatGPT. You don't need to use Codex 5.3 Spark in a coding context. You can ask it whatever you want and it will just act like a normal LLM. And I personally think there will be huge demand for faster inference across all parts of the AI economy. There's this old—

Yeah, another way to think about it is like if you have two employees with the same, skill set, the same capability, but one is just 5 times faster, right? That person can create way more value in the organization, right?

And for a lot of things, if they're 2 times faster, they do command 6 times the price. Because over the course of the year, a sales rep that sells twice as much, or someone who is twice as effective at their job, might actually command a salary that's 5 times, 6 times. The actual price. And so there's lots of other context across different business lines that you could draw to. There's also this old adage or saying about e-commerce that may or may not be real, but it's probably been transposed so many times in think pieces, I don't know the real quote, but it goes something like, "Every 100 milliseconds of latency costs Amazon 1% in sales." I don't know if that's the right way to think about it, but basically as Amazon was scaling, they realized that there were a bunch of things that they could do on the UI side, a bunch of things they could do on the layout side. Where does the buy button go? Where does certain information go? The price, the discount, all of this stuff, the images. They were tweaking the front end, but as they did that, they added bloat and the pages would slow down. And what they noticed was that the slower the page was, the lower the conversion rate because people were waiting for— Amazon.com to load, click on the page, it takes a second, they get distracted, they go somewhere else. And I think that that's happening in LLM use cases all over the place. People fire off a query and they're like, ah, it's taking too long, I'll go scroll Instagram Reels. There's always an Instagram Reel. And they'll be like, oh, I kind of forgot about what I was asking about. I didn't get my answer. And that's certainly true in business contexts as well. So this is currently playing out in AI inference. Companies are paying disproportionately more for faster inference and this is good for Cerebras. But Semi Analysis does point out a number of potential headwinds and problems that the team at Cerebras will have to solve or contend with over the next few years. Mainly, Cerebras chips are not currently as capable of holding larger models in the limited memory that they have or networking multiple chips together to serve larger models. We've heard about the NVL72, racks that wire a whole bunch of Nvidia chips together can serve these really large models. That has potentially been a challenge. So Semi Analysis says, moreover, the industry is tending, trending towards larger context windows ad infinitum. 128K context will certainly not be acceptable for long, especially with the prevalence of agentic workloads. And it doesn't look like there's a simple solution of just scaling the wafer size larger. Because TSMC is set up with a standard wafer size, and adding, or adding more memory to the existing architecture because Cerebras' whole design depends on a lot of SRAM, static random access memory directly on the wafer. But SRAM is no longer shrinking as much with each new semiconductor node. So the last version of the Cerebras chip, they've done WSE-1, 2, and 3. They're on 3 now, but WSE-2 had 40 gigs of memory. WSE-3, you would expect, oh, we want a doubling, right? We want a 10x or something. It got 44. So a 10% increase over one process node, one iteration. And some of the analysis is asking the question of like, okay, is there an easy way to double this? Is there a question like how will this scale as the models get bigger to add more SRAM? You might have to sacrifice compute area because everything is being done on one wafer. If you want computation or memory, there's a direct trade-off because you only have so much space on the actual wafer. And so they, there might be much harder 3D wafer bonding approaches doing stack stuff. Like there are other potential ways, but there's not like a linear, oh yeah, of course the next version's gonna double again. And so that is a potential problem that they need to work through. But in an agentic workflow, I think it's entirely possible that you want like the biggest, most powerful model, like the vice president delegating things. You want the vice president handling things.

Senior or junior?

Senior vice president, maybe just the president handling the critical work. So future models might not, and that might not be on Cerebras, that might be on NVL-72 or TPUs or something. But I imagine that we will quickly jump from the agentic age where you're firing the best smartest model at the full workload to the orchestration age Illustration age, and there will be hybrid approaches. So the biggest and best models will delegate certain tasks to smaller, faster models, just like they go and do database queries these days, or they go and search the web these days, and that's CPU bound. There will be certain workloads that the larger, smarter agent model, like the boss model, can sort of delegate to the cerebrus speed workers, the faster workers. So if you need to, I mean, just yesterday I was asking Ben about like, He pulled all our guests together. He's like, "I'd love to know the geolocation of every single guest." And it's like, okay, run that same inference query of like, look up this company, figure out where it is across 1,000 or 2,000 individual rows. And something like that's highly parallelizable. If you want that to happen really fast, it might not need a GPT-5.5 class model, right? It might be okay to run faster on a smaller model that works on Cerebras. And so it's hard to predict the exact mix of chips that will power large networks of agents. But these different designs, to me, they seem more complementary than directly competitive. Like a year or two ago when Daniel Gross wrote AGI Bets and was sort of like, is NVIDIA underpriced? I don't know if NVIDIA— he may have said that on Semi on Stratechery, but Like, you know, we entered the AI age and everyone was like, oh, GPUs are the future, NVIDIA's the company. But then it was like NVIDIA GPUs are good and then also CPUs are good and ARM is getting into it and Intel's doing very well and it—

We're gonna make big computers.

Big computers, big computers for sure. And so I'm extremely optimistic obviously and very excited to talk to Andrew Feldman, Doug O'Loughlin, Eric Mishra, a bunch of folks who have been involved with the journey.

Yeah, Eric did the Series A. Wow, conviction. Almost a decade ago, 2016, and Benchmark are sitting on, I guess as of this morning, like many, many billions of Cerebras.

It feels like they brought a huge team to the NASDAQ. Look at this photo, the second post we have. Obviously, Andrew's there, a lot of the team members, typically the key banker, but we've been to some IPOs, and some of them have had smaller teams. This one feels extremely celebratory and feels like a very broad, inclusive crew came together. What else is important about Wine?

And the Class B shareholders have 99% of the corporate voting power. Founder's in control.

Founder's in control. Founder, no. Hoonam says, "This IPO illustrates the power of an individual partner over the brand name of the firm. Pierre Lemoine was a partner at both Sequoia and Khosla, but instead of those firms backing Cerebras, it was Eclipse, the firm he joined at the age of 84." Overnight success. "He had backed this little-known chip company multiple times in the early days. What a way to wrap up a career." He was born in 1930, the same year as Warren Buffett.

Wow.

That is an awesome story. I love that. Matthew Siegel, the recovering CFA, is giving some color on what happened to the order book. One-third of the order book, the folks that said, I want shares in the Cerebris IPO, one-third of the book got zero. And the top 25— I guess the top 25 investors took 60%. That's probably the big investment funds, the Fidelities, the State Streets, the Black Box boxes.

They have done quite well today. This picture looks wildly different than the Klarna IPO last year, in which only a handful of the team at Klarna popped over, hit the NYSE, IPO'd, and then went back home.

Yeah, it was very much just like another day at the office for the team. Yeah, that's definitely what I was contrasting it to. The Cerebras valuation every round, Series A, in 2016, $100 million. Foundation Benchmark and Eclipse. Coatue led the Series B in 2016. VY Capital led the Series C in 2017. Then $1.6 billion valuation in 2018, $2.4 billion in 2019, $4 billion in 2021. That was like maybe a little bit of a slump, but then 2025, Atreides and Fidelity come in at $8 billion. Then Tiger comes in at $23 billion. $1.2 billion. Then in May of 2026, they IPO at $48.8 billion. I can sort of just do the—

Nice work from—

I don't need any help.

I don't need the soundboard to do it.

Am I doing it for real or am I using the soundboard? You don't know.

Incredible work from Tiger. Yeah. Coming in at $23 post.

Very, very good.

And now up dramatically in just 4 months.

Very, very good. Well, our first guest is joining us in 8 minutes. Let's run through the Kevin Warsh news because he has been confirmed as the Fed chair. And we'll run through this and then we will come back to some of the other stories because we have a gap later in the show. So Kevin Warsh, who is most famous for interviewing Alex Karp on CNBC while Alex Karp appeared to have popped a nicotine pouch and then spun a notebook on his finger. Did you ever find that clip, Tyler? Is that in the—

Yeah, it should be online.

Okay, let's—

Yeah, we have the video here.

That really put Kevin Warsh on the map because this is what he's known for now. Of course, he's had a storied career.

I remember I showed up in your office once. I was dressed like this, and I think you screamed at one of the guys. You said, Kevin's here, he looks like the guy from IBM. And I was talking about, well, you know, we need like really finance controls, and you know, how are you going to sell the product, and all this stuff. Okay, but I would say you certainly built that.

He's really spinning it. I didn't realize he goes back to it like 4 times. He's spinning dirty word anymore.

He's really good at this, but somehow you grafted that onto the strange company that can produce these products. How's that transition been, if I've got it right?

OK, wait, wait. So I have so many questions. First, we have to get him to recreate that for sure. Second, Tyler, I thought we were talking about that being on CNBC, but that looks like just a podcast. Like, that doesn't have any chyron.

Yeah, no, I don't think it was actually on. I think it was from Palantir.

Like that was a Palantir.

Oh, okay. So it was just like a random podcast. And then, and then when I've seen on CNBC, they were playing the clip. Got it.

I think so.

Yeah, probably a reaction stream over there. Well, let's go through what happened because Kevin Warsh has been confirmed as the new Fed chair. The vote was 54 to 45 in the Senate. The divided vote signals challenges ahead for Warsh, who faces a Fed committee skeptical of rate cuts that Trump has demanded. Of course, we talked about the inflation news. Typically, you don't cut interest rates going into inflation and potentially economic stagnation. You definitely don't cut rates in— that's why stagflation is so difficult, because if you have stagnation and low inflation, you can cut rates very easily. Maybe the economy starts overheating a little bit, you get a little bit of inflation, but then you can you can pull back. That's what we've done historically. Vice versa, if the economy is running hot, you're seeing high GDP growth and high inflation. Well, if you raise rates, you're going to pull back on both of those. But in stagflation, you're seeing both inflation and, and economic stagnation. Harder to deal with as a Fed chairman, which is potentially the task he will be faced with. So the Senate confirmed Kevin Warsh as the Federal Reserve's 17th chair. Wednesday in a largely party-line vote that required— that reflected how tensions with the White House have dragged the Fed deeper into the political fray. I was looking back at the old Fed chairs. There's some absolute legends in there because some of them have really long runs. So very quickly you get back to the black and white portrait in the painting as you go back in time.

Who's your favorite Fed chair? Hmm.

Volcker. Yeah, Volcker is pretty good. Bernanke is great.

Absolute dog.

Yeah, I don't know, hard to pick, hard to pick. Warsh, who was nominated for the post by President Trump in January, won confirmation 54 to 45, earning support from all Senate Republicans but just one Democrat, John Fetterman of Pennsylvania. Senator Kirsten Gillibrand of New York did not vote. No Fed chair has been confirmed by such a narrow margin since Senate approval approval became a requirement for the job in 1977. Chair Jerome Powell, whose leadership tenure ends Friday, captured at least 80 votes in Senate confirmations for each of his two terms atop the Fed. Wow, Jerome Powell, just fan favorite of both teams, 80 votes in the Senate. That's pretty significant.

Pretty— Yeah, is he going to be looked at as Technical, throughout history?

Get maybe a VC fund going?

What do we think? No, but when we look back, it seems like the last 2, 3 years, he's handled himself—

Yes.

I mean, he's had a really tough situation, and he's managed to land the plane.

When did he start? When did Jerome Powell first become the Fed chair? When was he? He's been assumed office 2012. So, I think the— wait.

2018.

2018. He was right after Janet Yellen. And so, I think I'm putting him in the conversation, Jordy, but I'm not giving him the GOAT trophy.

Yeah, I'm not saying that.

But nearly, because the challenges faced, he wasn't confronted with a Great Recession, a dot-com a dot-com bubble bursting, a Black Friday, like the economy from 2018 to today—

Global pandemic doesn't, you don't count a shutdown of—

No, because— large parts of the economy. No, no, I actually don't because the economy was pretty strong in 2019. And it went into 2020 with pretty strong consumer balance sheets. Low debt. There wasn't a shadow banking economy. There was no bomb in the US economy waiting to explode. And so although we saw high unemployment briefly, and we did have to stimulate the economy, that's not his job. His job was to set rates. There was a little bit of like, I mean, maybe you put the inflation, the ZIRP era and the end of the ZIRP era and all of those gyrations on him. but those, the problems that were downstream of both the ZIRP era and the end of the ZIRP era were suffered mostly and benefited mostly on like tech companies and Silicon Valley companies that had really long cash flow horizons. And so there was not a moment where it was a dire situation that the Fed had to intervene in a meaningful way and like save the economy, like in 2008. It's a big deal, he did a great job, but he didn't, he wasn't faced with the same, challenges of a Bernanke, for example. That's what I'd say. Tyler, what do you think?

Yeah, I think that's reasonable, but also like, you know, if Powell was worse at his job and you saw some crazy crash because of COVID and then he brought it back, like, then it'd be like, oh yeah, he did face this massive thing.

But because he did, you know, such a good job, maybe you didn't see any like massive crash.

So, so like the, like nothing super bad happening is evidence that he was really good. As a Fed chairman, right?

Yeah. Yeah, maybe he's a defensive back. You know, if they don't score, there's no great plays because he's just shut down, shut down cornerback for the last couple of years. Potentially. He's definitely in the top 17. I'll give him that. Anyway, Kevin Warsh, Chair Jerome Powell, whose leadership tenure ends Friday, captured at least 80 votes in the Senate. The previous chair, Janet Yellen, was a little bit more controversial, confirmed 56 to 26. Seems like not that many people showed up in 2014 to vote for Janet Yellen, with many senators absent because of bad weather. Interesting. I wonder what would have happened. I mean, it feels like she would have cleared it no matter what. But a difficult economic backdrop and Trump's broadsides against the Fed independence have set up the central bank for a thorny leadership transition. Transition. Senate committee confirmation hearing last month, Warsh faced intense questioning from Democrats over how he would maintain the Fed's independence from a president who places priority on personal loyalty. Warsh said he would preserve the central bank's monetary independence and that he had made Trump no promises about policy decisions. Powell, citing concerns about political attacks on the institution, plans to remain on the Fed's Fed's Board of Governors, defying Trump's insistence that he leave. He says, you're going to have to drag me out of here. I'm staying at the Fed, says Jerome Powell. Warsh is 56. He's been immersed in monetary policy debates for decades, frequently as an outspoken critic of the Fed. Former Morgan Stanley investment banker, he became the youngest Fed governor in history at 35 when former President George W. Bush appointed him to the central bank's to the Fed's board in 2006. During the financial crisis that struck 2 years later, he played a key role in tying up rescue deals. He was the bridge between Wall Street and the Fed that sort of Bernanke deployed. And so, that's what— those are his laurels that he will not be resting on, but he will be drawing on from experience. So, Warsh left the Fed in 2011. He had become a critic of its direction, concerned that as the economy recovered, the Fed its ongoing efforts to support financial markets went too far. So we will have to check in with the progress on Fed Chairman Kevin Walsh soon. But fortunately, we have our first guest of the show, Amy Reinhardt from Netflix, in the waiting room. Let's bring her in to the TV Land Altar Room. Amy, how are you doing? Doing well.

How are you doing?

Doing fantastically.

It is an honor to have you here. Yes, our first ever guest. From Netflix.

I think so. Thank you so much for taking the time.

Only took like 2,000 interviews to get you guys on here, but we're excited to be here.

Honored to be the first.

Yes, yes. I mean, obviously big fans of both Netflix and advertising, but would love to start with a little bit of background on yourself, your experience, and just sort of your intro to how you found yourself as the president of ads at Netflix today.

Sure. I've been at Netflix for about 9 and a half years now in a couple of different roles. Started out first in our content organization doing both licensing and then overseeing production. And about 2 and a half years ago, I stepped into this role overseeing our ads here. And it's been a fantastic 2 and a half years. A lot of excitement. Feels like we've been able to accomplish a lot and great company.

So if you take us back to the initial, the initial push into ads, what can you tell us about the tradeoffs, the build versus buy debates that were going on at the time? The just the maybe even the cultural changes? I think we are super, you know, we love ads. We think that's a fantastic business model. It's a way to deliver great value to customers at lower prices. And there's so many benefits. But culturally?

Yeah, what was the debate like?

Yeah, what was it like internally?

Yeah, well, I think it's been well publicized, you know, that not being in advertising was a strategic bet for a long time, right? And so early in 2021, 2022, when we started to talk about the notion of getting into this business, yeah, it created a lot of, I think, fair to say, you know, angst within the company for a bit of time because it was such a big shift. To your point, culturally and strategically. So I would say, and then we made the announcement that we were getting into it. And in terms of the whole build versus buy, you know, we partnered with Microsoft to enter the business very quickly. And that got us up and running. But it's been, you know, we made the decision about 18 months ago to lean into building our own tech stack. And we launched that a year ago. So we're just a year. I keep having to remind myself how nascent our tech stack is because because we've been able to deliver so many developments and so much progress against that over the course of the last year. But I would say full circle, you know, we are past— we put to bed all notions that we shouldn't be in this business. I think everybody understands strategically now that it is important for us to be that. And we've been able to grow our user base because we have been able to get to a lot more consumers who are looking for that low-cost option and are fine with ads, right? It's been a great thing for the company, I think, and everybody's on board. And, you know, the recent news, as you heard, which we just announced at our upfront yesterday, that we're expanding that ad tier into 15 more countries around the world. And there you go.

Fantastic.

Everybody's on board. Full speed ahead.

How are you pitching the ad product today? Is this primarily brand marketing? Is there a timeline to get to more of a like performance focus? Like, what is your pitch to advertisers?

Yeah, uh, you know, as we see in the marketplace, advertisers are oriented around outcomes, right? So we know that we need to be a full funnel solution, and we believe that we have the metrics to support that. So to your point, we've been very successful with some of the brand partnerships that we've done over the course of the last year and a half, uh, but we've also seen really good conversions in terms of lower funnel and making sure that we're driving purchase intent and consideration. So as we build out more of our solutions, we are going after that whole funnel solution.

What are conversations like around how brands should— how much brands should want to associate with particular pieces of content? Because I think some brands might come in and say, well, I'm advertising, you know, strollers, and I know that parents will be watching K-pop Demon Hunters with their kids, and so, like, this is the most on-the-nose directed, and I care. I want my brand linked to this particular piece of content. But we've seen time and time again that, like, once the algorithms get good enough, once dynamic ad placement can actually flourish, you— every company tends to see better performance there. So where, where are the ad buyers today in terms of those tradeoffs?

You know, there's a full spectrum. So absolutely, we get advertisers who want to be associated with K-pop Demon Hunters like McDonald's or with Stranger Things, right? Like those big moments. Those are oftentimes the easiest to sell. I think K-pop Demon Hunters is actually an interesting case because when it came out a year ago, We didn't know that we had a hit on our hands until about 60 days into it.

Yeah.

And I think that's what the magic is of Netflix, is that we have so much variety and depth of content that we're programming and trying to hit all audiences that you never know where your next hit is going to come from. And so selling those audiences, selling that, you know, audience behavior, moods, targeting moods and relevance is really important to a lot of different advertisers. So again, we just want to meet advertisers where they're at. And some folks understand that being across a number of different programming choices is important. And some people want to tag along with those big tentpoles. And we want to, you know, provide those opportunities.

Yeah, I mean, it's interesting. Netflix has like deep, deep experience in machine learning, AI recommendations, all parts of like, you know, high-throughput data processing. But I'm I'm interested in any learnings or surprises from building the proprietary ad delivery stack. Has it been as expected? Has it been— there's new skills that you need to bring? Because a lot of companies have been successful at scaling content and then struggle to figure out ad delivery. You obviously haven't. But then also there's this AI boom going on, which can help with productivity, but also new algorithms and new ways to actually target content. And so I'm interested in where the buildout didn't match your expectations or surprised you.

You know, as a tech company, we do a lot of testing and we go into things with hypotheses. So we're constantly testing things around our member experience. And, you know, I think that's been a differentiator for us, like really leaning into reducing member friction, making sure that member User experience is a good one with lower ad loads, lower frequency apps, those types of things. But we do know that there are times when we have to pivot. So I would say you— there's not just one example of a time that we've had a wrong hypothesis. We're constantly testing things out and figuring out where we, you know, where that— those hypotheses prove out, where we need to pivot and, you know, change swiftly. It's hard to point to one specific moment where it felt like it's been a learning. I would say the bigger learning for us just as a company is, you know, this is a relationship business too. And we've never— this, you know, you talked about kind of getting into ad sales, right? We've never had a sales team in our overall organization. So I would say there's more organizational learnings. Than necessarily tech learnings because we're so used to that tech cycle of testing and learning and iterating.

How are you thinking about the, the ad product feeding back into the content production? I've noticed that Netflix has been fantastic, in my opinion, of creating more engaging content. I was watching The Rip. With Matt Damon, and you click the play button and you see Matt Damon's face in within like 2 seconds, and it clearly confirms that you're watching the right movie. And then the title card comes in, and that's a departure from 50 years ago. You watch The Shining and, you know, it's a helicopter shot of a car for 5 minutes, and they show you the full titles. And it is a different style of editing, and some people lament the old style. I particularly like the new style, and I'm I'm wondering about, we went through a period of time when television, there was the famous like fade to black and then the ad break and then fade back in and you resume. And Netflix has never had to contend with that in media products, but is that going to come back? Is there a next generation pattern for creating content that can both have ads in it and not? Are you seeing glimmers of what the future, like the impact of ads might have on, on like the editing structure and the timing and the pacing?

Yeah, a lot of it, to be honest, depends on our creators. So, you know, working with talent who, and some of those, some of that talent may be more tech-forward and are thinking through those types of things when they're writing shows. I'll give you an example. Shonda Rhimes was used to writing for broadcast and network for many, many years. So when she writes a lot of her content, she's already thinking about where those natural breaks are. But not all writers do that, and that's okay. We can still find what are those natural breaks because we want to make sure, again, getting back to the member experience, that it's not intrusive or it doesn't come mid-sentence, right? And is cutting off any of the action.

Yeah.

We're able to adapt to any way that our creators want to write the content and fit it into that member experience.

Yeah, true.

For US markets, is this— Is there any enterprise spending? I would imagine that a lot of enterprise buyers, like, have been Netflix subscribers for a really long time, and maybe they're not getting served any ads at all, and so this is more of a consumer opportunity, or am I thinking about that the wrong way?

We've, most of our clients right now, the target segment that we're going after are enterprise top 400 clients, right? Because we think those are the ones who—

Well, sorry, I meant B2B versus like B2C companies.

Oh, we think about this more as a B2C opportunity for the most part. And I think as we expand our learning and expand our offering, may get into the B2B space, but for the most part, B2C.

Yeah, yeah. I feel like all of that, like the higher upmarket, more targeted, that's all unlocked with scale. Once more— there's more learnings on responses. I'm wondering what other signals you think might be valuable because many times, you know, advertisements shown during a TV program are very passive, harder to track. But if someone's watching on their phone, there can actually be a call to action, a trackable link. Like, I imagine that the data's messy, but how important Is it to sort of close that loop in an ATT era where it's a little bit trickier, but there's a lot of things that you can do on the signal side anyway?

Yeah, you're absolutely right. And this is an area where I talked about the testing and iterating. We're leaning in a lot on the testing. What does that second screen experience look like? You know, again, how do you meet the customers where they're at without being sort of intrusive? A lot of testing going on in this space. But the biggest thing for us is, you know, privacy safe. We want to make sure that we're leaning into, again, in that member experience and taking care of our members' data. But a lot more, I think, to come.

Have you been surprised by the return of the QR code in maybe podcast advertising? But I see it a lot because people are watching on, you know, they'll watch a YouTube video on a TV and the creator will, you know, hard code in a QR code to link out. And that was something I had completely written off QR codes and then they made a major comeback.

No, I agree with you. From a member experience, may not be the most simplistic thing. Understanding kind of the ad tech on the back end, I'm not surprised by it because it starts to be pretty complex pretty quickly.

Yeah. And there's some— we hit some sort of like inflection point where the— maybe it was in a certain iOS revision where the camera app became so easy to press a button and pull out, and then it detects the QR code so quickly that that flow— because you used to need to like have a QR app separately to scan it, and now it's all integrated. And so someone can just whip out their phone and run right to it.

Jordy, you have something else?

Nothing, nothing super top of mind right now. I mean, last question I had is, do you ever expect Netflix to serve more short-form vertical video style ads in something like the Clips tab? I know the Clips tab right now is focused on basically like content discovery. But I imagine in the future people will spend more time in a format like that, especially on mobile.

Absolutely. And that is one of the announcements we had at our upfront yesterday, is that as we roll out this vertical video content, that we are going to be offering that to advertisers along with our to-do.com coverage in 2027. So yes, we think that is a big opportunity.

Opportunity to. Last question for me. I would love to know about the intersection between games and ads. That's been a huge, huge growth driver in other— with other categories and other companies. But I'm wondering where that is in the roadmap, how you're thinking about that.

We haven't thought about that yet. Look, our roadmap— I could fill our roadmap for the next 2 to 3 years based on just some of the foundational things we want to do and a lot of the innovative areas we want to lean into. But it's an area that we're keeping an eye on. And as we watch that game's engagement increase, I would never say never. I've learned to never say never at Netflix, but it's not something that's on the near-term roadmap.

Okay.

Thank you. Well, thank you so much for joining.

This is great to meet you, Amy. Thanks for breaking it down.

We'll talk to you soon.

Cheers.

Thank you.

Moving on from Kevin Warsh, who is the newest Fed chair. We have a debate. We have a debate going on in the timeline around General Catalyst's advertisement. Some are calling it an attack ad. It's— particularly Andreessen Horowitz is calling it an attack ad. We touched on this yesterday. But if you did not tune in, General Catalyst, the large-scaled— what do they call it, giga funds now? What do they call it?

Platform.

Platform fund. But there's something else. It's a huge, huge venture capital firm. Capital firm. They launched an advertisement which we can play again to refresh everyone. Let's scroll down.

Yeah, let's play it from the beginning.

It's only 30 seconds long.

Hi, I'm GC.

And I'm VC.

Who's your friend here, VC?

This is Woof AI, an AI-native companion platform that combines robotics and machine learning. You'll never want a real dog after this.

Well, I think people like dogs as they already are though, VC.

You don't need to walk it. You never need to tell the kids you sent Milo to the farm. We're leading the seed and could probably make room for you. Well, I'd love to hear more, but we actually have a really high bar around responsibility for these things.

Is Woof AI okay?

Of course he's fine. Oh, sorry, buddy.

Easy, easy.

Stay, stay.

No.

No! Stop! Stop it!

Stop! Stop! I'm sure he'll be fine.

Stop! Okay, tons of thoughts, but you kick it off. What's your read on this? Take me through it.

So the actual ad, the way it's shot, the timing, cinematography, It's fantastic. I do— I think that— I think the ending is— I think the ending's funny, right? It makes me smile a little bit. The dog's going haywire. The robot dog's going haywire. My first thought is I actually think there's a huge opportunity for a robot dog that is 10 times better than existing robot dogs. There are robot dog toys out there. So, I don't—

Specifically in the toy market, because this is a Boston, this is probably, I mean, it seems like it's a Boston Dynamics robot dog, and those are not typically used as pets that I know of. I think that they are more deployed, whenever I see a demo of a Boston Dynamics robot dog, it's like walking through a nuclear power plant that you don't want someone walking into. It's like an industrial product for the most part, and then obviously a fodder for for viral videos. But yeah, step through it because it is a crazy back and forth from both firms. First, the stats. General Catalyst, oh, they're over 1,000 likes now, but 2 million views. So lot of discussion. Anjani Midha says, "It's a bit cringe, guys." Olivia Moore @andreeson says, "If I was a consumer founder, I would run for the hills watching this." weird take from a fund that returns so much money from Airbnb and Snap, because— is the idea that, like, robot dog is a weird idea, but so was airbed and breakfast or Snapchat, right? The biggest B2C businesses always start out looking weird.

Yeah, there's a way to do the same pitch for airbed and breakfast, which is, you know that empty couch? You know that empty room in your house? What if you were to monetize it? What if you were to allow strangers to sleep in the empty room in your house? And of course, that was a lot of the criticism at the time, was that it never was gonna work. And it was strange, but then it created a lot of positive externalities. People built businesses, people get to travel and integrate with local communities. Yeah, so the main thing here is like, I do find it fascinating, considering that when you actually look at the portfolio overlap, it is insane. Almost all of their winners, almost all of their winners, especially in the modern era, they are in both companies, right? Yeah, of course. And so they're backing a lot of the same companies. So it's super hard to counterposition. Against them. Now, granted, a16z has a very different media strategy. They're very loud. They're all saying the R word a lot. And I think GC can kind of counterposition against that. But you're not going to counterposition against, like, what companies you're backing, right? GC— I had to look this up because I thought there's no way this is true. GC's in both Kalshi and Polymarket. Big, great companies, right?

Yeah, but those have been at the center of the debates.

But that is the center, those companies are at the center of the moral debate in tech, right?

Certainly for the venture-backed controversial companies. There are some that are controversial that are not.

Yeah, and all the funds have backed various like betting, trading-related companies over the years. I think it was like GV did, what was it? Not DraftKings, but FanDuel, I think, back in the day. So it's not to say—

There's companies that start out as controversial and then they become true, like completely normalized and everyone sort of comes around to like, "Oh, that's a good thing." Like Anduril is a good example. Both of these companies, both Andreessen and GC are in Anduril. When it launched, it was like, "Wait, you're building defense technology? That's insane." And now the whole Silicon Valley community has come around to the idea that it's really important to have a functional defense industry. And then on the other side, you have companies that come out as really controversial and people are like, "This is the end of the world." And then they just sort of fizzle out. Like, Clueley would be a good example of a company that like has done okay, but it's not like, "Oh, no one's doing homework. It's completely upended education and it's so successful and it's bad." It's like, no, it's like, there was a lot of saber rattling around that. Sora went through a similar thing where everyone was like, "This is gonna wirehead everyone." It was like, ah, just like some funny memes and ultimately, you know, moved on from it. And so it's very hard to map like the controversiality to the ultimate outcome and where it lands. Like a lot of the prediction markets were not controversial when they were just predicting the election. That was not what was controversial. Tech started being controversial about it or saying it's controversial once it got into sports betting.

Yeah, so the ad is interesting a couple of ways. One is trying to counterposition against Andreessen, GC being like the cool hip fund that wouldn't back the robot dog company, even though when you look at the portfolios, there's enough overlap that you can safely say if a company is ripping or has a lot of potential, they're probably both interested investing in it, regardless of what category it's really in. And then the other thing is that historically, like, a16z is the firm that's trying to be like, hip and cool and loud and do like new media.

Yeah.

And GC is the one that I've always viewed as like more buttoned up, more behind the scenes, more traditional finance.

I think of them as like New York, East Coast, Boston.

And I love that. Yeah, we love that. We love that.

I think Bain Capital is a good example of another firm that's been like pretty strait-laced, but still has like aura behind it. And it's like, they have a private equity fund attached to the venture fund. And like Bain's done some fun stuff. I mean, Fogo de Chão, and they've leaned into things every once in a while, but they've never been like, oh, we're gonna be the craziest brand strategy. It's been like, yeah, we're by-the-book investors. We find great companies, back them, and Sequoia has done that too. Their whole pitch is very austere, and that's worked for a long time. It doesn't need to pivot the brand, which is maybe what this is signaling towards. This is the first time, I mean, it's very rare for any, VC to like take shots directly at another VC just because they're all syndicating deals to each other.

Yeah, it's way more direct. I mean, a lot of people clocked it, but because the actor could look, I guess, if you were far away enough, like Steve Ballmer. Yeah, Steve Ballmer.

I think he looks like Steve Ballmer.

Or Marc Andreessen.

But people are saying Marc Andreessen.

But the other thing is like—

So hold on. I was looking at Marc Andreessen photos, and I could not find a single photo of him in a vest. So I don't know about this, but clearly Mark Andreessen did take issue with it because he, he quote tweeted it something like 45 times, which, which, you know, really amplifies it and creates more of a conversation about it. Strategically, you might have just wanted to mute this if you don't want this to become a thing. But maybe he does. Maybe he's like, yeah, actually, I do want to fight because like, this is dumb and I'm going to fight back and I'm going to win. So that's the strategy. But it does seem like messy, rolling around in the mud. When you're wrestling with pigs, you're going to get muddy, right? Isn't that the saying?

Yeah, it's just, again, like creatively, it looks great. And it's very fresh. And I think Reggie and his team did great work. But it's funny to just take shots at a whole category of investing, basically. That is kind of the bread and butter of GC's business.

Yeah, yeah, drawing the line in the sand. What is the actual line? The really high bar for responsibility around these things? Because first off, it's just odd because like robot dog feels very, very low on like the responsibility. Like if some VC was like, okay, we take responsibility ability really seriously.

Yeah, so I, seriously, for example, the last robot—

I would be like, don't fund gambling, don't fund cannabis, like don't upset Sager and Jetty basically, is like the way I would pitch someone if they were trying to be like the responsible actor fund.

Interesting.

And robot dog would be fine.

The last robot dog pitch deck I saw was a company that wanted to use robot dogs as a replacement for, actual seeing-eye dogs.

Oh, interesting.

Seeing-eye dogs are incredibly expensive to train. Something short supply. You know, kind of out of reach for many people. And so the opportunity for a robot dog that can maybe travel with you, like on a plane without barking and needing to eat, right? That's actually a very world-positive kind of endeavor. I don't know if it'll work or what the business will look like, but in general, like there's a, you know, anyway, so.

I'm extremely bullish on robot dogs. I think it's very complimentary. You think about, I mean, especially if it's a toy for kids, think about how many vehicles kids have these days.

Yeah, Gabe says service dogs can reach $60K.

Whoa.

And that doesn't count lifetime of all the other costs associated with dog ownership.

Next time I see some of the robots, with a service dog. Bow! Flex like, "Whoa, buddy!" I didn't know you had to flex like that. No, the 60, the, yeah, I mean, kids have like, you know, a bicycle, a tricycle, an RC car that drives, and then another one. And like, it's like, you throw the robot dog in the mix, I feel like that's going to be in addition to a dog. Robot dog rips. Maybe 1,000 robot dog startups will flourish, and the most ironic scenario is because the industry will become so deniable, the general catalyst and Andreessen both have to back it, and they're like duking it out for allocation at every round, and the future robot dog founder out there is like, I just do nothing win, or build a great company and win. Was there anything else going on with the robot dog debate?

I, selfishly, I think it's bad for the industry if you have two of our platform funds just making these somewhat petty videos that are basically ad hominems. That, basically, throwing stones from glass castles, as one could put it. So I think it's generally bad, but for entertainment purposes, if they want to keep going at it and turn this into, you know, proper Drake versus Kendrick situation, this might be a good use of A16Z new media. Yeah. Although much of the talent there is now over at—

I mean, I would— I think I would love— yeah, I would love like a, like a response video from Andreessen shot in the same way or something like There is an opportunity for like a rap battle here that's a beef that's like very, very entertaining that I would absolutely love. The flip side though is that I agree with you. Like if you're a VC, you're much better picking a villain that is something around like a lack of technological progress. Like Thiel did this very well with like stagnation is the boogeyman. Like if we don't get the robot dog, we don't We don't get the seeing eye dog. We don't cure cancer. We don't do the big thing. We don't visit Mars. Like, that is— you can still have like a villain, but the villain needs to be something that the industry and America and all the constituents can align around instead of like picking fights where like these two firms are obviously aligned on like 99% of like where the future goes and what the goal is of building a business. A business that delivers a value that consumers enjoy.

Yeah, it's just so funny because if you look at cap tables, they're almost always probably touching. They're effectively holding hands on the cap table, right? Because one of them might have more of one company, slightly higher. One of them might have more of the other, right? But in general, they're just hanging out together.

I mean, Cap Oil did Inderal at GC and then went over to Andreessen. And so there's more overlap than— than differences. Plenty of other things to take shots at, but I mean, to Andreessen's credit, they've done a good job of that, focusing on geopolitical competition, focusing on China, focusing on stagnation and nimbyism, all sorts of different things that they've been aligned with more of an abundance view as opposed to punching down, although there's memetics all over the place.

Over the place, it happens. Yeah, it seems like GC has an opportunity to be the buttoned-up platform.

Yeah, I think so.

You know, they don't need to say the R word. Yeah. They don't need to be super loud, right? They can just focus on the craft of investing.

Yeah.

I wrote about it in the newsletter, but, you know, run some ads.

Yeah.

You know, you've been saying this forever, run some ads in The Economist, Financial Times, Wall Street Journal. Just there's a way to counterposition yourself without attacking your rival and often partner.

I mean, do you remember the last time General Catalyst went viral? Hamant, CEO, was on Harry Stebbings' show, 20VC, and he said like triple, triple, double, double, double is no longer good enough. We wanna see 10x-ing every year because that's what's happening in the AI space.

Era.

And it was a brash statement, it was bold, it's smart to debate.

But actually looks—

But look at the results, like, you know, like there are companies that are doing this, like we talk to them every day. And sure, and he clarified it on our show, he clarified it on other shows, he was not saying that like, you shouldn't build a business that only triples revenue every year, only doubles revenue every year. He was just saying that the reality of the market right now is that there are power law companies, companies that are growing exceptionally fast, unprecedentedly fast, and so you as a venture capitalist have to adjust your benchmarks and think about how you're allocating funds, what companies you're investing in, make sure you're in the best company in the category that's actually going to win, and it might be the company that's growing 10x a year, not 2x a year. And so that was something where it was like thought leadership from Haman It sparked a conversation. There was some debate around it, but it was from a position—

And so far, that take was controversial, but entirely correct when you look at the growth of a lot of the most exciting companies in the industry right now.

Yep, totally, totally. And so that sort of more narrow, staying in the lane view, it got a lot of attention. It did break through. It caused a conversation, but it still didn't it didn't feel like, oh, he's taking shots at someone specific, right? It was just like a market analysis from someone in a position to give that exact analysis. So anyway, Jensen Huang is over in China. Jason Calacanis has a photo that looks extremely real, zero AI detected, but he's bringing two huge boxes of GeForce RTX 5090s, which are—

This is a picture from when he was in Alaska too.

Jason says, never stop selling. I agree. There is some news, which we will cover later in the week, around the dynamics around H100 sales and Blackwell's, what's actually happening. It's all in flux as the Trump-China summit plays out on the front page of The Wall Street Journal every day this week, because it is headline news. High-stakes US-China summit kicks off. There was another drama in the tech world yesterday, but we'll come back to it after our next guest, Ben Hylak from Raindrop, joins. I believe he's in the waiting room, so we'll let him come in. He's the co-founder and CTO. We've had him on the show before. Welcome back, Ben. How you doing?

Doing well, man.

How are you?

Fantastic.

Great to see you.

What's new in your world? Uh, reintroduce the company quickly and then tell us Tell us the news.

Sure. Raindrop, we make observability for agents. So the main thing we do is self-healing agents. So what that means is that when your Raindrop hits a problem in production, we detect it, we fix it.

How do you do it?

That's a good question. So at the end of the day, like, we consider ourselves like the intelligence for your intelligence. What that means is that we are the best, fastest way to essentially look at anomalies. So what that means is that, like, let's say you make a change, right? We're able to very, very quickly find out that, like, oh, users all started complaining about something, or the trajectory, the traces are kind of starting to evolve into a different pattern. And so it's kind of a combination of agents, but also more like classic ML techniques, a lot of like custom trained models for every customer.

Walk me through the shape of the agent market right now. Like the way you're talking about it, you know, sort of illustrates the broad diffusion of agents and custom agents. I think that a lot of people think Claude Code and Codex, and I don't know if you're doing enterprise deals with those firms or that's the goal, but I imagine that every startup, many legacy companies have built some sort of agent, some sort of harness. And I'd love to know the shape of how broadly diffusing like custom agents are in companies versus is it the domain purely of startups that create an agent for legal or an agent for sales and then they vend that into a company?

Yeah, so I would say that there's two kind of categories of customers. We started with super high growth, startups at the time, startups. So those are companies like Klaviyo, for example, Framer, Speak.com, some of the fastest growing companies in the world. And those were some of our earliest customers. We're lucky they grew a ton. So, you know, that has helped our growth.

Always helps.

It always helps. Yeah. Someone once mentioned that, like, you know, this kind of business is a lot like early stage seed investing, actually. It's kind of interesting. Like, you know, we have— you have to be pretty, pretty picky not to work with companies that are going to die. Because if, like, especially analytics, like all these sort of things, like they, you succeed as a company when your customers succeed. Like if all of your customers are terrible, it's like everyone's like, well, why do I, why are you interested?

Yeah, I had a portfolio company that was working on like agent infrastructure, like roughly 2 years ago and pivoted because he was like, okay, this is clearly going to be a big thing. Someday, but right now he's looking at all the underlying companies and he's like, I don't believe that any of these agents in their current iteration are going to work. Now, maybe they're trying to.

I think it was very counterintuitive at the time, but I think we chose to find companies like Clay.com, right? Which are, were clearly on an insane trajectory, but at the time weren't necessarily as large. And so I think a lot of our customers that now are pretty large, but at the time weren't necessarily as large. And then in the last few months, we've been moving into Fortune 50s, Fortune 100s, and a lot of amazing things happening there. And again, it's kind of like two shapes of a product. One is in our bread and butter is companies that are redefining the way people interact in different verticals. But then, yeah, there are Fortune 50s, Fortune 100s that are also deploying agents internally, I think the shape of that looks very interesting. And like, it's something that, like, being on the forefront of like understanding how these companies are deploying things, like there's not that much I can talk about right now, but yeah, always very interesting.

What do you think is a generally underhyped agent category right now? I'm sure you're seeing the future a little bit.

It's a really good question. I think that, I mean, you know, I, so this is a tough question. I, what I want to do actually is pivot the question a little bit because I want to talk about, I want to talk about our launch today, if that's okay. Is that okay with you guys?

Yeah. Yeah.

I'll tell you my questions. You tell me your answers.

Okay. Okay.

Sounds good.

Does that mean that you want me to not answer this?

No, no, no, no. I'm just messing around. Go for it.

Okay, cool.

The joke is—

yeah, I botched it.

The joke is, what questions do you have for my answers? It sounds good. Like that, where it's like you could ask him anything and they're just gonna— sorry guys, but it's fine. I want to hear about the launch today, so just tell us about it. Let's talk about it.

Okay, so guys, there's been this crazy thing that, that has been missing for a very, very long time. That's why I want to talk about it. So like, people have been building agents You're building them locally, like you're using some sort of SDK. It could be OpenAI, it could be Vercel, whatever SDK it is.

And what do you mean locally? Like it actually has to run on like just on your laptop or—

Right, like before you push to production.

Sure, right, sure.

It's on your laptop.

Yeah, yeah, yeah.

There's no way to see what it's doing, like no standard way, nothing. Like so people will send those traces out to like a server, like Raindrop is like one of those, you know, and there's a bunch of others.

Yeah, but they might also just like drop the logs in a non-relational database, for example. Sure.

They'll just print it to console.log, like, oh, here's what was happening. It's like that bad.

Yeah, yeah, yeah.

The other problem there is you can't see a nice trace or you're sending it to some server and it takes seconds to see everything and whatever, it looks terrible. But then also your coding agent can't see the traces either. Then when you hit a problem and you're like, hey, this response was wrong, flawed code will just make shit up. It'll just be like, oh, I think that maybe this tool was wrong, or I think maybe this happened because it doesn't have any of that data. It doesn't actually know what the coding agent did. So I think that as someone building agents, as our company building agents, we're like, it's actually kind of embarrassing how long it took us to solve this problem.

No one else solved it either.

But yeah, that's what we launched today. Free local open source tool. Raindrop.ai/workshop. And it's completely free. It's just open source.

Why open source?

That's a really good question. I mean, I think the genuine answer, and I think part of why competitors haven't done it— I mean, there's probably other reasons for that as well. But I think it's that it can be, right? Someone else can do it. You know what I mean? I think that it running locally is the best experience for people. And to be clear, like, there's still things that it enables, you know, if you connect it to your production Raindrop, which is like you can pull in a remote trace and replay it. And then Cloud Code or Codex can just keep doing that loop until it works. So there's still benefits for us. But also the truth is that we want people to hack it. We want people to, to meld it into whatever works for them. So I mean, we use a lot of open source things here, right? So it makes sense to contribute back as well.

Yeah, that's great. Yeah, I'm wondering about other just predictions about the next breakout category of AI agents, what you're seeing. Feels like we're so close to being able to book a flight, but maybe no one wants that. I don't know.

I mean, I'm not sure if you guys saw, I had a little bit of a thing Ryan Chesky earlier about Airbnb.

Oh, yeah.

Oh, yeah.

Talk about that.

Talk about that. No, I think like, you know, I use Airbnb a lot. I love Airbnb. I think if I had to guess, I would say Ryan Chesky knows a lot more about Airbnb than I do and probably a lot more about being a founder than I do as well. And so I think there's probably a lot that I'm not considering. That being said, I think it's fresh. Like if Airbnb had an API, I would use it and I would book Airbnb with it, like through Clock Code. Right. So it's like, I know I would do it. I find Airbnb very, very hard to search. And I think that there's a lot of— I think the tough part and what I see industry-wide right now, everyone's trying to figure out is you see companies almost reducing themselves into an API with absolutely no moat. You look at Photoshop, Illustrator, etc., they're like, oh, we have a cloud code integration now.

MCP.

At the point where people are just using Photoshop, Illustrator, etc., as like an MCP, they've sort of lost the game, right? Like if no one's actually touching the UI anymore. I think that right now companies have to do that increasingly because they have no other choice. I think that there will be a point where the incentives don't make sense anymore. Like I can give an anecdote from, from when I was at Apple. You know, do you guys remember App Clips?

Yeah, yeah.

Where did those go? I only see them—

where did those go?

With like parking meters sometimes.

Yeah, right. So like one of like the hero ideas there was like, oh, you know, like imagine you're in line at Starbucks, you don't have the Starbucks app downloaded, like, well, why not just, you know, scan something, have an app order your drink? And it's like, turns out Starbucks doesn't want that, right?

Sure.

That's the last thing in the world Starbucks wants. Starbucks wants you to download the app, they want you to have They have an entire— like, there's a reason why DoorDash and Uber Eats and whatever, God knows, other apps exist. It's not because they need to, but because each have companies and money and goals. So why would they reduce themselves into an easily interchangeable API? It doesn't actually make sense.

Yeah, but I think it's using— I think it's important to be careful around using a tool like Photoshop interchangeably. Interchangeably with like a retail store like Starbucks or like a marketplace like Airbnb or DoorDash because I really think that these marketplaces provide, you know, an exceptional amount— all the value is not in the UI, right?

I agree.

I agree.

And like the value of Starbucks is not that it's a pretty app, it's because they have specific drinks that they can make pretty much anywhere, you know, someone would be.

Yeah, I think of that company Buy, the drink company. They got started during the direct-to-consumer boom. Obviously, they would have some beautiful Shopify website. They didn't. They just went direct to retail, and they had Amazon. You could order it online. And if you went to their website, it would just say, go to Amazon. And they did fine, a billion-dollar company. And because the value is not in the e-commerce experience, they didn't play the stars game. Of course, Starbucks is maybe sacrificing a piece of that business model, but it's not giving away the whole cow. I don't know what the phrase is.

There are going to be ways to monetize this, right? Like there are going to be successful business models built on top of this sort of layer. And to be honest, as Raindrop goes into the future, like that's the future we're building towards.

That's the future we want.

I mean, we're going to be announcing a partnership with a really large— one of the large coding companies as soon as far as like integrating with them more where it's like I don't see Raindrop as a company that's going to submit PRs and production to people's codebases. Someone else is going to be doing that. We're going to be the layer that's really good at finding those issues, diagnosing them, and tracking them. So I just think it's going to be interesting into the future how much companies are willing to sort of just be the API without all those hooks, without knowing everyone's email, having the mailing list, all that sort of stuff. So that's a very interesting trend.

I feel like you're generally on the frontier and cutting edge of like adopting all these tools. You mentioned your cloud code use and I'm wondering about— give me a reality check, a health check on your experience with computer use because you're lamenting the fact that Airbnb doesn't have an API and—

Yes.

I imagine you could create a scraper or download the HTML and interact with it like treat the front end as the API effectively, you know, puppeteer the computer through computer use. Like, where are you on like the AGI moment in computer use from what you've seen? Like, where does it work? Where doesn't it work? Where would you recommend people get started if they wanna play around with it?

Yeah, that's a really good question. There's definitely places where it works. Like, I think that Codex has done a very good job of implementing like browser use actually, both for like debugging applications that you're working on and in general. Like, this is something that Claude Godefroy just doesn't do, creates a really, again, that I think the next couple of months, the thing you're going to keep hearing from me, but also everyone in the world is self-healing loops, loops, loops, loops. How do you create loops where it's, how do you close the loop? How do you have Claude code, make a UI change, see that it sucks, and then just keep going? A lot of do we have AGI or not is how many loops in a row can you do?

It's all loops.

Before things just end catastrophically, right? Because there is sort of this, like, it gets worse, right, in many cases. So yeah, I think there's a lot of ways to answer this. I'm a fairly security-conscious person. I think that I'm not like an Open Claw guy. I'm not going to give all of my cookies to some agent, et cetera, et cetera. But yeah, I think TBD.

Yeah.

Cool.

Well, yeah, new challenge: book an Airbnb with an agent. Can it be done? Is that where the goalposts need to be set? Let's figure it out. Uh, anyway, thank you so much for coming on the show.

Great to see you, Ben. Congrats to the team.

Congrats.

Of course, last thing, if you want a hat, uh, you can— we have a new CLI. You can run raindrop-drip. You can get a hat, an umbrella, a couple other things.

That's a fun— that's a fun way to give out merch. I like that. That's very creative. Uh, thanks for coming on the show.

Great stuff.

We'll talk to you soon. Okay, back to the debate around Figure. We've had Brett Adcock on the show before, and he had a livestream. We talked about it a little bit. Watch a team of humanoid robots running a full 8-hour shift at human performance levels. And Brett Adcock said this is fully autonomous running Helix 2.

All right, pull up, pull up this post from Pete.

Yes. Yes, and the stream did fantastically. It was 24 hours, it got 3.4 million views, but at a certain point during the stream, there was some questions about whether or not the humanoid robot was in fact—

All right, back to the beginning, back to the beginning.

Okay, let's play this.

All right, so it's cooking. I mean, the speed is actually insane.

And we were extremely impressed by this. This was remarkable, remarkable.

Even if it's teleoperated, it's extremely impressive.

Yeah, yeah, yeah. Yeah, like the robot's clearly working. This is very, very clear.

But they're saying that it's not tele-op.

Okay. So then the robot starts missing things, being a little bit like an inch off, and then reaches up and touches the robot's head, the robot, which is something that wouldn't normally be necessary. It doesn't have like a logical explanation or conclusion. So a lot of people are asking.

Well, it does have a semi-logical conclusion, which is that Brett is claiming when it reaches across its body to go to the right that it puts its hand up here to get the hand out of the way.

That's what I was thinking, was that if the hand is, is halfway up, it might be blocking the sensor, the camera sensor. And so even though, like, you might— the robot might reach the hand up further to move out of the view so that the robot can look at the next package. So that's one possible explanation, but a lot people are asking even harder questions, saying that potentially, was there a human in the loop? Was this teleoperated? Which is something Brett has said it's fully autonomous. I feel like that means no humans in the loop, but Teor Taxis has an artist's representation of Helix 2, figures in-house neural network running entirely on board. And it of course is a human in a VR headset. Very debatable. We'll see where you stand. But there is a third option, which I have shared, which is potentially no humans involved. I don't know if you'd call it autonomous, but you would call it no humans in the loop because you have—

Well, it is an autonomous system, right? It just sort of runs.

Yeah. I would consider this autonomous. It's the, it's the image that I shared in the production chat. It's not of a human and it's not quite robotic, but there's no human in the loop. And so this could explain it's the system is running with no humans in the loop. If you make that claim and you follow this, I think this qualifies as no humans in the loop. If you have a giant orangutan in a VR headset puppeteering the robot via teleoperation, you could say that this system does not have a human in the loop. And you could make that—

And I could make the argument that it's autonomous.

Yes.

The chimpanzee is running its own. It has somewhat of a neural—

Neural network.

Almost like a neural network. Yes, yes. No, no, no. No one knows.

Andel says, I think there was a human physically inside.

Ooh, physically inside.

Yeah, I mean, the thing that I'm so, I wanna talk with somebody at a place like Amazon who I imagine does this kind of thing all day long. And are they asking for a humanoid to do this process? Like this seems like something that e-commerce fulfillment and logistics companies have been doing for, many, many, many, many years. Is there not a purpose-built robot that sits right there and makes sure that the packages are in the right orientation? Does it have to, you know?

Yeah, if you watch an episode of "How It's Made," you'll see every variety of custom-made machine for flipping around, sorting packages, that type of activity. There are custom-built machines that run at scale They might cost like $10,000, but they last 50 years. And anytime you see a, you know, a Diet Coke factory or gum manufacturing line, all these things, like the gum that you have there comes off and the gum rattles down and is sorted into the pockets of the packaging. And then the sleeve is wrapped around and glued and all of that is done autonomously, but just with, you know, a bunch of machinery that was built in probably like 100 years ago, honestly. If it works, don't fix it. But you can clearly see how this type of task package sorting would be like on the curve to a more economically valuable humanoid robot. And like if I was going to buy a humanoid robot to do my dishes, and you showed me this video and it was in fact fully autonomous, that would be an encouraging demo to me. That would be something that I would look at and say, oh, well, like if it can do this successfully for hours and hours and hours, I'd probably trust it to put some laundry in the washing machine. That doesn't seem well beyond the scope of capabilities. I am interested about the power.

It's so interesting how, quick it is when it's just sorting packages there, and then it doesn't bite and walk on the way off.

Hmm.

Like, if you rewind for a second.

Yeah, yeah, yeah. The walk is not—

I only use that terminology because that's the terminology that—

that is what Brad used.

Yeah, like, look, why does it look like that?

If you were able to shuffle like this so fast and so fast, you'd think that you'd be able to hustle a little bit. But maybe that's V2. Maybe that's less, less relevant for this particular task. You know, there's a lot of different options, but we will dig into it. Brett launched day 2. I mean, regardless, putting up views, sorted 32,000 packages. Day 2 is live, and he shared more details on what's going on. The original goal was an 8-hour run. After zero failures yesterday, we decided to keep going. We're now over 24 hours of continuous autonomous operation. Without failure. This is uncharted territory. The task is small package sorting. F.03 detects the barcode, picks up the package, and reorients it barcode face down onto the conveyor. Humans average around 3 seconds per package. F.03 is now around human parity. The robots are reasoning directly from camera pixels. Uh, the robots are fully autonomous. Using Helix-2, our in-house neural network running entirely onboard F03. There's no teleoperation. Every action comes directly from Helix-02. Okay, well, I feel like that rules out the monkey business. I think teleoperation would fall— if you had a monkey puppeteering this thing, I think it would count as teleoperation. So he is denying that allegation from the timeline. Yeah, but the timeline Everyone seems convinced. YouTuber commenters started naming the robot Bob, Frank, Gary yesterday, so they added name tags to each robot. And if the robot gets stuck or the AI policy goes out of distribution, Helix triggers an automatic reset. You'll occasionally see this happening during the livestream. If a robot has a software or hardware issue, it autonomously leaves for maintenance and another robot takes over. We run our labs and figure this way to maximize uptime. If we have had a failure yet. We haven't had a failure yet, but statistically we probably will at some point. So very, very fun going back and forth. Who else is chiming in? People are— Dar says, I'm the last person I expect to rush to Figure's defense, and I'm looking forward to hearing Brett's take here. And in any and all cases, I stand with PBD. King, but IMO, this demo seems authentically autonomous. And could see this being learned behavior from teleoperators that collected the data for this model with their VR headsets. And PBD Sucks, who broke the story or went viral first time, said he actually has a pretty reasonable sounding excuse, but doesn't give me tons of confidence on the model's brittleness. For cross-body research, for cross-body reach, the policy lifts its arm to avoid hitting the metal shoe. "Shoot, nice try." I wasn't sure if he was going to reply to this and sort of engage or just sort of let the timeline run wild with it. But the metal plate does seem like a piece of what's going on. But people are still hungry for teleoperation bombshells. It sort of cuts both ways. I remember Jason Karmen did a video maybe with 1X, and everywhere in the video they said, "This is teleoperation." We're doing teleoperation, we're bullish on teleoperation, put it at the bottom in the text, in the description, like, so told everyone, and still people were quote tweeting it being like, "This is teleoperation." And so people, you know, are sort of grappling with like, what is real, what is fake constantly? Well, is there anything else on the Figure story that you'd like to dig through?

No.

Switches hands after working more than 4 hours straight.

Huh.

Well, there is some news. A newly released OGE form, Office of Government Ethics 278-T, discloses that President Trump filed 3,642 trades involving stocks of public companies between January 1st and March 31st. Transactions include hundreds of stocks and ETFs such as NVIDIA, Microsoft, Broadcom, Amazon, Apple, Alphabet, Meta, Goldman Sachs, AMD, Airbnb, Palantir, Netflix, Costco, Walmart, JPMorgan, DoorDash, and others. Individual purchases of NVIDIA, Microsoft, Broadcom, Amazon, Apple.

Individual.

So he's averaging around roughly 40 trades a day.

40 trades a day.

Check my math there. OK. That is in Q1.

Trading activity. We talked about this. Like, should you just give Jane Street right access to the federal government? Should they just be able to change the laws to optimize for max GDP growth? And it feels like we're one step closer, one step closer to the economic singularity of the hedge fund running the country. Anyway, we have—

Any other—

Well, yeah, what else?

I'm trying to find the history of presidential day trading.

I don't know if there is one. Jimmy Carter famously divested from his peanut farm because he was worried about conflicts of interest. But we are in a new era anyway. We'll have to figure out if Trump is long or short the Cerebras IPO. He's probably watching right now to hear Doug O'Loughlin's take on it to understand what George W. Bush with Cerebras said.

He says, yes, not a day trader, but had a famous controversial stock sale. He sold 200,000 Harken Energy shares in 1990 before bad news came out.

Okay, interesting.

Well, and there's no other evidence that we're finding of presidential stock traders.

Well, we'll dig into it. But we have Doug O'Loughlin from Semi Analysis in the waiting room. Doug, how are you doing?

Welcome to the show. Good, good, man. You know, pretty busy day. Another day, dude. Honestly, every day is a busy day.

Every day is a busy day. Take it, take us through it. How do you think the market reacted to the Cerebras IPO, to your, to the Semi Analysis deep dive on the company? What is the overarching story here?

So I think the market was obviously positive. I don't think we're quite as positive as the market, but it's a bull market, baby. I think the takeaway is that Cerebras got to IPO, which at one point in time we didn't think that would happen. The semi-analysis world.

We're very—

we've historically been very bearish on SRAM.

Okay.

But I think there's a path forward for them to be a disaggregated pre-filled chip or maybe even an AFD chip.

Yeah.

Meaning attention feedforward disaggregation.

Okay.

So yeah.

Yeah. Unpack sort of the, the competitive dynamic, like what the, the, the fear around Cerebras as far as I could tell years ago, it was like, will this ever be useful? Will they ever actually be able to make it? Will it have defects? Then it became certain applications, demand side, customer concentration. But where do you think they are now? How has that journey evolved?

So first and foremost, Cerebras is about SRAM. SRAM's like the fastest possible memory and it's kind of done on a logic process. But the problem is SRAM scaling is dead, meaning that you can't make smaller and smaller SRAM scales. So pretty much they like kind of committed to this dead-end, uh, process by having the biggest scale-up world, uh, as a wafer size, but then the models got much bigger than just a single wafer. And so they have really, really fast inference but only at a certain size. And I think the real capability problem is, can they inference models larger than a trillion parameters? And I think the answer, as we think right now, it's pretty unlikely in the near term.

Yes. So I, I, so I understand all that. I, I'm just wondering about the world where should I view it more like a CPU? Because when the AI boom, the ChatGPT moment happened, the obvious buy was NVIDIA because we're gonna need a lot of GPUs. No one was really expecting a chip shortage in CPUs, but then agents wound up using CPUs for a bunch of stuff. You have to keep the GPUs filled and so So CPUs are now in demand. And I'm wondering if there's this world where there's this, yes, you are gonna move past the trillion-parameter models, but we're going to keep using them forever, just like we use relational databases forever, even in an agentic AI world.

Or you have a scenario where you have a big model that is giving sort of—

Yeah, orders.

Orders, workloads.

Delegation or something.

Delegating to smaller models.

Model. Yeah, I think, I think in a perfect world where there's no silicon constraints, that might be true. But obviously there's silicon constraints, and I think, um, Cerebras is really well optimized for a certain problem, and we think they do a great job at answering that, which is fast inference at a certain size of model. Maybe that, that, that market's going to be large enough. And I mean, honestly, I don't think I was ever bullish Cerebras the entire time, but now that we're here Like, not ironically, 1% of a very large market works. And I think they got like 1% of a very large market. When it first started, I was like, oh yeah, what are you gonna do? 1% of a very large market? That's gonna be a few hundred million dollars.

And that's like the classic seed pitch too. 10 years ago, they're—

Yeah, is there any, for a long time, a lot of fear around ASIC companies around architecture changes. We're going to move past the transformer and they're all going to be locked in the past. Is there any optimism around there's an architecture change that actually is to the benefit of Cerebras and makes them more relevant in the future?

Do you think that's possible?

No, that's—

I mean, might be outside my pay grade. That's too GigaBrain for me, right? But where I'm at and the understanding, there is a narrow path for them, I think. And I think they're going to be able to inference maybe 1 trillion parameters at very small context window sizes or smaller models at very, very fast speeds. But I don't know, man. Maybe the true gigabrain take is Mythos is so good or whatever that it makes compute efficiency super easy. And, oh, yeah, your model is inefficient in AGI. I understand. Yeah.

Distill yourself so you can run on a Cerebras chip just as effectively. Okay, now we're talking.

That's, that's the gigabrain thesis. But I think, I just think that there is, um, there's demands, right? Like, clearly we're in a shortage, and, uh, ironically, in a shortage, it's not the best company who wins. I mean, you can look at NVIDIA's stock chart and that tells you it's the second, third, fourth best companies where the demand overflows, right? And so we're seeing all that today.

Yeah.

And I think the reality is the market's big enough for a lot of demand and Storyverse is in that space.

Okay.

So they've done a really good job.

Yeah.

And I mean, it's a cool engineering problem.

Yeah.

But we think it's kind of a solution looking for a problem because the world of LLMs blew up at a much faster scale than anyone could have ever thought of.

Yeah.

The size, I think it's really the difference.

Yeah, yeah. Give me a little primer on Groq, how Groq fits into the SRAM machine market, what the view is, because it felt like that, that NVIDIA's move there with the license acquire, as you put it, was defensive against Cerebras. Is that the correct framing? Like, how does Groq fit in on this?

Okay, so let's talk about exactly where Groq fits into the architecture. So in a transformer architecture, you have like the multi-heads of attention and then there's a feed forward network. That's a portion of essentially the entire transformer block. And what's become really hot in the last few years, or not even two years, like probably a few months, man, is you've been disaggregating all the different parts of inferencing into subsequent specialization. So we're talking about GPUs and ASICs being a specialization over CPUs, but now we're actually starting to break the, essentially the constraints of inferencing into different different, I guess, compute and memory-bound pockets. And so for example, we're finding prefill ends up being— prefill being essentially loading all the weights ends up being compute constrained. So you don't really need a lot of memory bandwidth. So why don't you just use a very FLOPS-heavy portion and you disaggregate the memory onto the decode portion, which is extremely memory bandwidth limited. And so this is Groq, where this fits in the strategic thought process here is in the GV200 rack, What you can do is you can pass the activations over to the SRAM in the Groq LPU rack. And that is an extreme speedup. And so that's like a perfect example of another break apart of the transformer architecture. Yeah, pretty technical. But that's like the thought process here is that the memory is so fast, the memory band or the speed of the I/O doesn't really matter. And you don't need a huge scale-up world size because you're less streaming the activations. Yep. Um, that problem wouldn't work with, uh, the Cerebras chip because you're kind of— it's, it's an island, right? You think of it as an island of compute. It's really, really good at everything in the middle, but moving anything off the island is really hard versus moving something off the island onto a Groq chip because there's a plug at the end of it. It's a lot easier. And that's kind of the, the calculus, I guess. Yeah.

So Cerebras lower memory bandwidth, lower interconnect speed.

Off the chip.

Off the chip. But on the chip it's as fast as Cell.

That's great.

Yeah.

Okay. Yeah, yeah. So what does that mean for the Groq NVIDIA ecosystem? Because is this something where the default configuration is going to be a Blackwell and a Groq chip, like in, you know, 50% of racks, 80% of racks? Or is this like still some sort of niche application where Groq is going to be deployed, you know, sort of sparingly sprinkled into specific use cases? Do you have an idea?

Yeah, I think I don't have an idea with high precision. I think you'll find that a lot of these things, there's a lot of different ways to split up and serve your model. So expert parallelism, pipeline parallelism, tensor parallelism, right? And so the correct optimization per hardware rack is going to kind of depend on the shape and architecture of the model. Model, and we don't really know with high precision what is what. And there's been kind of like different road— roadmaps along the way in terms of what, uh, what they wanted to do for speeding up inference. A perfect example of this is the CPX rack, which was mostly built for expert parallelism. It kind of remains to be seen if this is like— if the Groq GB200 speedup is going to be like the way forward, but it's definitely a technology tree that I think Jensen is excited about. So I mean, we'll see.

What about Lisa Su at AMD? Is she excited about this technology tree? Can you give me an update on how AMD fits into all of this?

So AMD is mostly just trying to get the last thing to work, which is the rack scale-up.

Yep.

And I think they're gonna do a good job of 450. I think what's gonna happen is that like, you know, it's compute shortage, right? So you're talking about overflow demand. I think Lisa's gonna figure it out. But on the inference serving side, I think there's definitely some demand or desire to probably match the NVIDIA roadmap. And I wouldn't be surprised to see if there's some kind of fast SRAM offload FFM chip in the, in the next 12 months. But the thing is, the number of candidates there is actually like pretty low. I think Intel's really— Intel's going for Samanova, which is a little clever. There's like HBM2. There's a few other players out there too that pursued SRAM scaling. But I think that in this specific case, Lisa's mostly just figured— focused on the last thing. And I think AMD is definitely good enough right now.

Okay. On Intel, what is the latest there? It feels like the roundtable has been assembled and sort of everyone has held hands and decided to maybe jump across the transom at the same time, take the leap of faith. But It also feels like, you know, lithography machines are majorly backlogged. Like there's a whole supply chain that they have to answer to that's backlogged. And so really high expectations. But also, what are they— what is the next milestone for them after they actually get these deals with Apple and Elon Musk?

Amazon and Elon Musk.

Yeah.

Yeah. And the gigafab. Sort of like once they get those signed, like what does the next couple of years look like?

I think it's about execution. It's kind of crazy to me that I think the stock price is ahead of the technical turnaround. And I think that— I think Liputan clearly has like righted the ship and gotten the right people onto the party, if that makes sense. And I think— I really do think the government intel deal was a stroke of genius because Pat Gelsinger spent, you know, 3 years trying to build a bottom-up demand to essentially come to the fab. And Trump's like, yeah, none of this. I'm going to sign the deal from the top. And what's going to happen is you're going to come play because we're in the United States government or else. And so I think, I think people are there. I think the customers are there. I think the process is good enough. I think 14A will be also good enough given how much of a shortage N3 at TSMC is. And it's all execution risk from here. But the historical Intel has quite a bit of execution problem. So we'll see.

Okay. Before we move on to TSMC, which I want to go to next, are there any other interesting ASIC projects on the horizon? We've talked to a few of these companies, but I'm interested in like the shape of the differentiation. Like you explained a little bit of the divergence in strategies between Groq and Cerebras, but there's Etched and a bunch of other companies that are are working on new chip designs, and I'm wondering if any of them stick out to you as particularly differentiated.

You know, I'm not going to go too into the details because I feel like some of them are even like still figuring out their roadmap. Yeah, I think Maddox is kind of interesting, uh, the way that they're, they're kind of trying to pursue, um, the memory problem. I think, um, I think Etched— I'm excited about the kind of YOLO bet, if it makes sense. Just make a big cystolic array. Yeah, but I think there might be like niche cases. I think the problem is like at the end of the day, Nvidia's big bus is still really good for the majority of cases and you're going to have to like start to make really opinionated bets on the ASICs to find what niche market ends up being all like a, like a diverter of demand into their ASIC. And so the ASIC specialization from here, I feel like you have to make some pretty big make big brain bets in order to make your bets come, come pay off. And I think most of the bets that I would have guessed when you like, when you originally did them, didn't really, wouldn't have paid off. And the ones I didn't expect did, like, it's kind of crazy.

Yeah, it is. It is a very weird market dynamic where a couple of years ago we saw ASIC and new chip companies, new silicon companies raising hundreds of millions of dollars or $500 million. And it was like, well, Well, for that, you're gonna need this massive market. Are you really gonna flip Nvidia or something? And then the market grew so much that the 1% of a huge market sort of potentially maths out for some of these companies now. It's a fascinating development. Jordan, do you have something? Yeah, I do.

China trip.

Yes.

Oh yeah.

What are you tracking?

On the H100? Oh, so honestly, do you guys see the parade? You know Trump loves a parade.

Oh yeah. They're winning him over.

That was a good parade.

Good parade.

I was like, dude, I was thinking, man, I'm not much of a parade guy, but I was like, dude, if they showed up and that parade was for me, I'll be like, these guys could be friends. Yeah. My impression is that the executive branch really wants a deal. And I think, you know, you saw the H-200 list, the verified H-200 list. I expect probably more lightening up on the executive branch. Something that's really interesting is if you look on the legislative branch, there's actually more expert control bills going through the House than like ever in history time. So there's kind of this tension. But I do think, you know, Trump's a businessman. He loves a deal. I expect, I expect a deal.

So, yeah. Somewhat related, TSMC. Ben Thompson was writing that potentially they weren't ramping CapEx fast enough. What are you, what are you tracking on TSMC being a potential bottleneck for the AI buildout? Just as more and more— Cerebras is now trying to get allocation. It feels like a particularly sharp-elbowed place to do business.

Yeah, so I think at the end of the day, TSMC is kind of a kingmaker in terms of supply, and there's no reason for them to really let the market go out over its skis. And I think they're happy with the pace of what they're expanding out because like, hey, they're growing their CapEx like whatever, 40%. But in absolute dollars, these are big numbers. Yeah, we're going to run out of TSMC engineers in the island of Taiwan pretty, pretty soon here. So I think, I think this is all kind of good on the margin for overflow demand, which is actually it's Intel. Intel's, you know, definitely reflecting some of that. But I think the shortages specifically at TSMC is driven by cleanroom. It's a long lead time item. It takes 3 to 5 years or let's just say 3 years to bring a clean room up. And so in order for them to have like figured out and like perfectly matched demand 2 years ago, they would have to been like, we have a 10,000 square foot house and we need to buy a 50,000 square foot house with conviction. Right. It wasn't that clear 2 years ago. And so I'm going to expect supply to kind of lag over and over and over. But demand signals will continue to essentially command premiums, move up wafer pricing, move up orders. And that's what's going to make TSMC invest more next year and the year after, but they're going to do it in a, like, in an incremental, not a revolutionary way, but like an evolutionary way. They are very like methodical and do steps one at a time.

Okay. Cleanroom fungibility. When you say it takes 5 years to build a cleanroom, I immediately go to SpaceX. I imagine that Elon can build big things quickly. Is there some world where where that partnership accelerates Intel regardless of your timeline for the mass driver fab on the moon, all the crazy long-term stuff, but just having Elon around the table to say, oh, we need to build something big and it needs to be, you know, capable of operating as a fab, like, is there something where he brings more to the table than just dollars potentially?

So I definitely think Elon is the man to do it. I forgot who said this, Elon makes the impossible late. I don't expect it to be on time. Talking about the cigar in the Terra lab, I'm really kind of doubtful. I guess from first principles, it's easier to just clean the entire room than to make really hyper-concentrated pockets. And that's what I would guess the bet is. But I still think by the time Elon figures it out, the supply response will have reacted already. We're still 2, 3 years out and there is some cleanroom fungibility. But, and you've already seen this actually, Micron bought an old power fab. I think this is the TSMC deal. People are buying display fabs. Essentially every bit of cleanroom that is not accounted for in the world is being snatched up and retrofitted to kind of meet the supply demands.

Interesting. Yeah, I mean, that's happening all over. Didn't Ford just announce some sort of AI play today? Stocks up on something. It's all over the place. I am interested in terms of like 6% getting, getting powered shells.

Ford is worth more than Figure now because last year, around a year ago, I remember Figure Robotics was worth more than the Ford Motor Company.

I won't have time, but now they're both AI companies, I guess. But what are you tracking on the American data center buildout? Domestically or terrestrially before we move on to space capabilities?

Yeah, basically how— Yeah. Oh, go for it.

No, no, no. Just— No, go for it. I'm just curious about, I mean, we're starting to see glimmers of pushback at the municipal level, different data center bans, and I'm wondering about what are the big levers that need to get pulled to actually continue to bring capacity online in America?

Yeah, I think that's a good question. And you're already seeing the first level of this is the delays. Yeah, my favorite clickbait is 50% of all data centers in America are delayed or canceled, implying 50% is canceled when it's really just everything is delayed. That's like my favorite clickbait. I got to steal that in the future. But I think that— I think it's going to be local, municipal, and people have to really believe and demand and desire the jobs. And I think one of the ways that we're seeing this is like, you know, capitalism works. And effectively, the dollar per megawatt has been going up. It's like a one-way train in the same way that, like, you know, the power per rack has been going up. The cost of making these data centers have gone up. And one of the ways that happens is it leaks into labor. Right? So essentially you're super against it, but all of a sudden it offers 3,000 new jobs to your home and you're like, well, maybe, maybe I'll take it. And I think that with enough economics, oftentimes, um, you know, money finds a way. And that's kind of, that's kind of how I would guess. But it's going to be like, uh, it is a, it is like a county-by-county fight, right?

Yeah.

And some places are just going to say hell no.

Yeah. On that note, we were debating this earlier today. There's been a couple of examples in like viral photos and articles about like, they— I bought a beautiful house in the countryside and then they built a data center right next to it. And, you know, no matter how pro-AI you are, it sounds annoying to have a huge building that's an eyesore and maybe noisy, maybe smoky next to you. But have you been tracking like how, how feasible is it just to throw the data center like truly in the middle of nowhere? It feels like America has a lot of land, but what goes into selecting data sites, these data center sites these days. Do you have something else?

So, yeah, so I think pretty much two fiber pairs is the big, the big desire.

Got it.

Essentially, it's like you're, you're more than willing to go to where the power is because you have to go to what the biggest actual bottleneck is. And power is the biggest bottleneck. So you can just— in the past, you're talking about like, hey, having these inference or rather like, let's say, point of presence near local cities, right? But power was never constrained in that world. It was just, you know, the biggest constraint was getting this video from TikTok to your phone as soon as possible. If the biggest constraint and the largest part of the cost is going to be power, why not move the data center to power and then, then like, you know, essentially hook it up with fiber? And so I think that we're going to put them in the middle of nowhere. That's just how it's going to work. To a certain extent, there's going to be more densification in some of the inference near the population. But I still think the ROI makes the most sense to kick out in the middle of nowhere.

Yeah.

Has the political backlash, pushback, updated your thinking at all around the viability of space data centers? I remember, you know, we talked as this idea has like gained popularity, you guys have like consistently said, yeah, technically you can do that, but like maybe it won't be viable.

Takes a long time for the—

There are space data center players now that are kind kind of loving the pushback against terrestrial data centers because they're like, the more pushback there is, the more it could make sense for us to put this, put these up in space. But what's your view?

I still think economics is gonna win out. You know, something, a pound on Earth is probably 10 times more expensive in space. And it's really hard for us to go to like essentially beat that out new completely specialized supply chain for what's going to be a smaller market in the near term. It's a real adversary against the adoption in, let's say, the short run. In the very long run, because I'm sure you saw the Anthropic Colossus thing where it's also interested in space, the biggest maxi vision of this is AGI. We have 1,000 terawatts of GPUs on Earth, and we're like, we got to put a terawatt in space. right? So like in that world, I think space data centers work where a small percentage—

1% again, it's 1% of the market again. It's just like just 1% and it's a trillion dollar—

is trillions vindicated.

Yeah, yeah, VCs are vindicated.

Tam pitch deck slides vindicated.

Yep, yep, yep.

Yeah, it's, it's literally as big as the galaxy, bro. Just there's no end to it actually. Think about how big the TAM is. Is.

Yeah.

So I think what is more likely is if it continues to be painful to do it from a zoning perspective in America, we— it will essentially slip into other geographies, probably in the Western Hemisphere. There's a lot of power in space in Brazil, and I think that that's probably good enough, right? There's definitely ways to make this work. I definitely think the only way you do it is by paying more and finding someone who's like, you know what, I'll hit the bed. And so that's the important part. But, you know, apples and pines away.

Is that sort of the bull case for sovereign AI initiatives? I was always super skeptical because like Europe didn't get like France's Google, like they just use Google. And for a lot of consumer aggregator type consumer internet companies, it's like Spotify is from Sweden. But it could be from America and it wouldn't matter. YouTube is from America and they use that over there. And you didn't need a, like a national champion in every consumer category, or there were certainly like returns to scale and a lot of the American companies just won. But so I never really bought the whole idea that like, oh, the French need like a locally trained LLM and the Germans also need a locally fine-tuned something or other. But if every company, every country has some sort of, you know, excess supply of energy or space or regulatory capacity for data centers, sort of bringing that online and just operating like a Neo Cloud could just be economically valuable for that country regardless of whether or not they're vertically integrated to the point of the consumer or the business that's running an AI agent.

I think that's probably the case where at the end of the day, economics is going to kind of push it through. And there is FOMO. And Europe did do a lot of investment in the internet really late. If we're going to use 1999 as an example, I think the thing I keep thinking about is that this AI thing is going to be a big deal. I continuously am shocked and surprised by the magnitude and scale That's a narrative violation.

I don't think it is right now. I feel like we are in a particular moment where— no, there's just the people calling the top in the bubbles. Like, they're awfully quiet right now.

And that makes me even more scared.

Yeah, that is— okay, so to be clear, yeah, you know, the true top— there's no— everyone's bullish, right? Everyone's like, dude, it's actually gonna be bigger next year. It's actually just gonna be a bigger bubble. Shut up.

So yeah, I was not concerned about— I was not concerned about a bubble when everyone was saying it's about— it's a bubble.

Yeah, exactly.

I am— I mean, I'm a little concerned it's a bubble, but at this point in time, I think if you look at the big—

I've been—

I've been reading a lot.

Honestly, here's my view. Here's my view.

Please.

It's not a bubble until you guys are spending 120% of revenue on tokens.

Yeah, our gross margin goes negative.

Yeah, you're just like, we're raising a major— we're not going to be investing in it. We're going to be burning it.

It's actually not a bubble until Semi Analysis goes public and trades up 600%.

There we go. I like that.

That's the, that's the real talk. No, no, I think there's a few things that have to happen. I think OpenAI or Anthropic, someone has to go public and it's going to be this year. Like, we have, like, we have to hit that keystone before, before it's all over. But I also, I also think, I keep thinking about this is like, dude, this is a big, a big technological revolution. Yeah, I think it's bigger than the internet. And I firmly believe this. I don't think I believed it would be bigger than the internet when I, maybe even 2 years ago, but I'm pretty convinced this can be bigger than the internet. And if you look at the past, these big technological changes are often sometimes bigger than, I don't know, everything else. It reshapes the entire world. For example, on the sovereign AI thing, maybe you're like, yeah, you don't need to fine-tune LLM. But what happens when AI becomes such an important fundamental, almost like society-level institution that a government can't control it? That becomes really uncomfortable and weird where it's like, hey, Anthropic can just put 5% of the compute of Mythos and run a really effective government whenever you want it. And you're like, whoa, what does that mean? For us. Yeah. And so this, this wave is so big that I think people are going to, out of fear and concern that they're going to be left behind and that the institutions that, that AI will bring is going to be bigger than like the original thing that we're doing. I think that that's like the problem, right? Like the Industrial Revolution changed everything.

Yeah.

The other thing that we, we were, we were joking about in, uh, like Q4 Q4 of last year, it was like, John was like, great, like the bubble pops, like, like the bubble inflated and then it popped, but then we got agents and then you have this sort of like reacceleration of every metric across the board. And so the other thing that we're like, we're trying to comp the AI boom to the internet, but the problem with the internet boom is that we didn't have the internet. So everything just took like, or the internet was coming online and people were getting access to it. And so the entire buildout and all of the capabilities companies took a lot longer to sort of grow, right? And now you have that core infrastructure, and so when you're layering on more infrastructure, that accelerates all the underlying trends.

Yeah, yeah. I mean, the labs— the lab revenue multiples are like an order of magnitude or two off of dot-com peak multiples. And in the public markets, Google, Amazon, Apple, all the hyperscalers are at like pretty reasonable price-to-earnings multiple still, even with all the CapEx and stuff.

And so the pushback would be it's on free cash flow that can make earnings look good instead of free cash flow. I think the revenue continues to be real, the demand continues to be real. And until you just like see demand evaporate, like, yeah, it's hard for me. It's hard for me to sit here and be like, GP prices are up a ton. Cloud Code is really valuable to me. I still think I'm an early adopter and this is all going to end tomorrow. I envision myself using it every single day more for the rest of my life, which is kind of crazy. I think I'm an early adopter. I just think it's hard for me to envision this not being a ginormous deal. It's kind of like we just got the— I really— I wrote this whole thing like, "Engels pause or whatever." It's going to change everything. The amount of net output that's going to increase is going to just blow our minds. It might be bad for GDP, ironically, because GDP will be unmeasured. Like, we're going to— like, GDP might be broken as a concept. GDP got invented in the 1930s to measure how much output you could make to not screw over the domestic economy for World War II. Like, it was a way to essentially organize the American economy. And it's a statistic. It's an estimate. Like, I think all of— I think we're going to like attack and like a lot of institutions and ways that we're doing things and ways we measure are going to be attacked by this because it's like such a big change. We have to rewrite the playbook over again.

And people— and it's funny, I think it wasn't Ben Thompson was talking about this in a recent interview of like people are comping this like, okay, Silicon Valley, like, you know, brought crypto online and then it wasn't maybe as big as some people had pitched it to be, even though it's been—

Yeah, self-driving cars, super powerful. VR.

And then even the way you're talking, you're like, you know, we're still early, you know, which is like a classic crypto line.

But the problem is, if you are early, you have nothing to sell.

Or you're saying like, you know, in crypto it was like, well, like a community could have a DAO, and that DAO could be worth a billion, that community could be worth a billion dollars, but there's just, no way to measure that. But now we have tokens. And you're saying GDP. But anyways, I'm trying to unlearn, I think, some lessons from that cycle because there are a number of things that are quite different.

It's also—

What about—

What about—

The reflexivity that people do have a little bit of an immune system to just running away with everything. Because you could believe this and then bid Nvidia to 10,000 times earnings or something. And at a certain point, you have to start grappling with the reality.

What about robotics? Has Figure had a major breakthrough?

I mean, one, I have not been following the feed as close as I should be. I just think robotics feels a little further out than the hype would let you believe. I feel like robotics is much more akin to the driving car paradigm where it's like, oh yeah, it's definitely going to come and automate everyone's jobs. And then it takes a lot longer and it's a lot unsexier. I think the scary or positive thing about AI is since it's information work and it's already been distributed and it has the perfect network to run on, which is the internet, it can disperse very quickly. And that's what we're seeing right now. So yeah, I'm just not anywhere near as bullish robotics as I am the fundamental core.

Well, I'm bullish on the next Semi analysis. I don't know, what are ClusterMax and InferenceMax? What are those called? Dashboards or analyses or rankings?

Dashboards, dashboards now.

Everything's a dashboard. Everything's a dashboard. Well, you need to make a new dashboard. GTP. Gross token production. This is what we're measuring now. This will be the output of the United States. Gross token production, GTP.

We need to— I mean, I think more on this soon, actually. This is like a place we're doing some research on. But I think the real bubble metric is if we're like, how many tokens? What's the token replacement cost? That would be some really good bubble math, where it's like, Yeah, yeah, software company has really low token replacement costs per market cap. But like a hardware company has an extremely high token replacement costs. And then it's like, oh no, no, it's just enterprise value divided by token replacement costs.

Well, the real, the real bubble one will be to go to the full Mary Meeker, uh, like eyeballs metric, eyeballs multiples.

Yes.

So you will value companies purely on token consumption. You'll say, oh well, they're consuming 10 trillion tokens, so they must must be worth $1 billion, and then you'll get really weird gyrations.

That'd be great for Semi Analysis. That'd be really good for Semi Analysis. We are consuming a lot of tokens.

Well, you're also putting a lot of good stuff. I really enjoyed the—

Would you guys ever make a sort of political-style attack ad against another research firm for having AI psychosis?

Hmm. Is that a reference to the GC?

Yeah, sorry, it's a reference to General Catalyst attacking, uh, Mark Andreessen. Andreessen.

Um, you know, life's pretty long.

I actually think XRP Analysis is just peerless. I, I, I don't think there's like a neck and neck with someone else. Like, it's just you guys.

I'm not—

I was gonna say, I don't really know who our competitors are. Yeah. Um, I don't, you know, I don't really think about it, Mark Andreessen, or, or, you know, another research firm like that. Maybe one day, maybe we will go through AI Honestly, you guys need a rival.

You guys need a— you guys need a— you guys need an arch nemesis. You need an opp.

Moody's.

I guess it'd be Gartner. Gartner had to say they're like— but this is not a good opp.

You need a sell analysis hype cycle and it's up only. No, no trough of disillusionment.

Straight line.

No axis.

And it just— it actually goes backwards.

It's a straight line on a log graph. That's what it is. Semi-analysis hype cycle. I love it. Gartner doesn't stand a chance. But thank you so much for coming on the show. This is fantastic.

Always full analysis.

Full analysis. Yeah. No more semi-analysis.

Those guys would kick our ass. If they had full analysis, they would kick our ass. It would be so over, man.

It would be over.

Anyways, take care, guys.

Have a great day.

We'll talk to you soon. Cheers. Goodbye. Next, we have Andrew Feldman from Cerebras joining in 20 minutes. We'll go back to the the timeline because the OpenAI Elon Musk trial is in its final day. The trial is ending. People expected 4 weeks of trial. We only got 3. They're cutting it short. What are the prediction markets saying about who's going to win? I want to know that. And I want to go to Mike Isaac, the Rat King, because he has a breakdown of what's going on. He says, good morning. Closing arguments of Musk versus OpenAI with special guest Microsoft are happening today, Thursday, May 14th. Again, Mike Isaac, of course, he kicks it off with what his lunch is. He's got an epic bar. He's got the bison snacks. He's got a La Colombe latte. He's got a couple other good things. He looks like he's prepared. He's got a bunch of snacks. I feel like he's in a better position today, learned his lesson 3 weeks of complaining.

Sort of like recursive self-improvement.

I think that's what's going on here.

So the call sheet. Well, Elon, when his case against OpenAI, it peaked at a 58% chance.

Okay. Where is it now?

28th. It's now sitting at a 30% chance.

30% chance. Okay. So right now the judge is instructing the jury on the criteria by which they should be judging the outcome of the case. Important because if the jury listens and carries this out, it is a very, very specific lens through which they view all the evidence. Ostensibly, it's where theater ends. Listening to this and being read out in court for the last 23, 30 minutes is very helpful because it's clarifying on how high the bar is for the plaintiff's side of proving some of these claims. Sort of feel bad for the AV guy during this trial. There's been feedback, there have been mic drops, but not in the good way. The mics have been dropping out. Funky video feeds. They need to revamp this place, says Mike Isaac. LMAO. The first joke of the tweetstorm. He says Musk counsel is going after OpenAI execs. Altman and Brockman and has the mugshot-style photo of Altman on the screen. Again, battle of Photoshop of executives in this trial has been entertaining to watch. You want to depict your opponent in the worst possible light. Musk counsel going back and forth, hammering the point they made over and over. The argument essentially painting a picture— Sam Altman, liar. Chipping away at witness credibility has been a core strategy for the plaintiff's side. And we're back to everyone. Hey, Molo is using Larry Page, who they claim doesn't care about humanity, as a foil to the noble Musk, whose only care with respect to AI is the future of humanity. Musk counsel is painting the Don't Trust Sam picture in a bit more detail for the jury. Also, Musk's side has a picture of Elon and Altman on the screen now. Sam's looks like he's about to be processed by a US Marshal. Musk's looks like he's getting ready for the Met Gala. LOL. Lots of Musk closing side arguments, semi-populist track of pointing at OpenAI and saying these billionaires are making gobs of cash while running a charity for the supposed good of the world. I'm curious if jury can register this argument even if it comes from Elon Musk, the world's richest man. Ouch. OpenAI counsel begins closing argument with a broadside against Musk. Even the people who work for him, even the mother of his children can't back his story. Oh yeah. Back to The War of the Photoshop. So closing remarks now in, in the digital displays on the monitors for exhibits. All the OpenAI executives look like Olan Mills photo shoots. Do you know who Olan is? He says it's complimentary. I need to get up to speed on my photographers.

Olan Mills is a portrait— offers portrait photography.

Oh, does look very nice if you pull up the Google Images on Olan Mills. Uh, anyway, short summary of the closing Musk camp: all these OpenAI executives are rich as hell and lying all the time. OpenAI camp: all of that is a sideshow and literally all the claims Musk is bringing cannot be stood up by actual law. The Microsoft camp disappears into bushes. Dota got mentioned again. They love mentioning Defense of the Ancients. Incredible Photoshop from the OpenAI camp of a calendar of events complete with little characters and a timeline of events. I wonder if they're using image generation 2, or if they're doing it the old-fashioned way. I can't wait until it's entered into evidence this afternoon so he can show us. Uh, sort of want to buy this meme guitar, but I also have 2 Telecasters. That's just completely side, side note. Uh, gamer has entered the blog. The Dota moment has been mentioned nearly every single day during this 3-week trial. AI researcher— we gotta have Mike back on the show, it's so good— uh, saw as a true breakthrough in the technology.

So Mike Isaac says, "I played with it in the past." What is the timeline for the jury to meet? Is this something they're doing today?

They're getting a 30-minute recess, most they've had in a month. I might actually be able to go outside and get real food. There's a Popeyes across the street. Is it a bad idea to get a bucket of red beans and rice? That's what he's thinking about doing. So not much news. On when this will close. It is 1:10 Pacific time. I imagine that they will wrap up by— what did he say, 3 PM, 4 PM? So 30-minute break that happened 40 minutes ago. Um, so I imagine that—

and but they've been taking Fridays off is kind of what I'm getting at.

Oh yeah, because this could— so maybe this happens to Monday.

This is just closing arguments, not necessarily the end of the trial or the jury might get the results Arguments for the jury might, might make a quick call, but—

well, there was an update unlikely 11 minutes ago. A lawyer for OpenAI on Thursday defended the company's chief executive, Sam Altman, from withering character attacks by Elon Musk's legal team as both sides deliver their closing arguments in a trial with potentially seismic implications. The stakes are high. Mr. Musk, who was not in the courtroom on Thursday because he was in China with President Trump, is asking for more than $150 billion in damages. He is also asking the court to remove Mr. Altman from the startup's board and to stop a shift the company made last year to operate as a for-profit company. They pushed back. Sarah Eddy, member of OpenAI legal team, tried in her closing argument to dull the attacks on Altman's credibility and to argue that there was never a firm agreement among the founders that could have been breached. Not one in this case other than Elon Musk has testified to any commitments or promises that Sam Altman or Greg Brockman or OpenAI made to Mr. Musk is what she's saying. And there is a new update that just dropped in after the recess. William Savitt, OpenAI's lead counsel, told the jury that Musk does not have a claim against the startup unless there was a specific agreement between Musk and OpenAI describing how his donations to the nonprofit should be spent. That agreement does not exist, said have it set. So that's where I guess OpenAI is leaving it for now. We will continue to cover the story.

Doug says, is the jury allowed to use Codex slash Gold? Be down in 1.5 hours.

Uh, there's other tech problems going on. Max Zeff over at Wired has been covering the story as well and says Musk's lawyer brought a big monitor, maybe 36 inches, into the courtroom. OpenAI's lawyers asked use it. Musk's lawyers said no. The judge told Musk's lawyers that they have to let OpenAI use it. Then OpenAI said it might not be possible to connect their laptops to it. AGI is here, but we'll still need a dongle, I suppose. A dongle has entered the courtroom, Max Zeff says. There's about 50— 15 lawyers standing in the middle of the room right now talking how to use— talking about how to use this big monitor. This is wild. They should have talked to OpenAI about sharing their monitor. What I always do, I always tell you when you come in here, talk to the other side. We don't have the technology available right now, so we don't want to use the TV. We think we should just get rid of it, says the OpenAI lawyer. Sam Altman just walked into the room, by the way. So that happened 4 hours ago. One of Musk's lawyers carried the big monitor out of the room upside down, wire dragging behind him, defeated, defeated liar. Retreats. That is a very, very funny story.

Uh, anyway, in other news, bring it down. Tim Draper says, I think I broke a record. I took 52 pitches in 52 minutes at below 40 degrees. Welcome to my office. Hashtag Draper University. Hashtag survival training. What do we think about going in the ice tank?

How cold are ice baths typically? You've done ice baths. I feel like I did one, and it wasn't as insanely difficult as people said, but then I checked the temperature, and I don't think it was 40. I think it was closer to 50.

Yeah, you can totally get closer to—

I—

'Cause there's a couple companies that sell—

Personally, when— if you're going surfing and the water is below 45 degrees, it can just be very painful. To, so even in a wetsuit, you're not putting anywhere that's not covered. A lot of people are putting gloves on.

What do you think, Tyler?

So apparently Joe Rogan's at like 34.

34?

Yeah. Wow.

So that's like the cold plunge, you know.

He's the top of the mountain when it comes to ice baths.

He's the final boss.

Yeah, this is just a crazy picture. I did think it was AI, but turns out it's real. It's just funny because it looks like, like, what is this set? What is this setup?

Yeah, what are all the trash bags there? And the wall is like sort of decrepit.

It looks like kind of like a prison ice bath.

Yeah, this is not what you'd expect from— I mean, isn't he a billionaire investor? You'd expect some sort of palatial, you know, you see the, the, the, the, the properties that Mark Zuckerberg's acquiring, that big investors are acquiring. You would expect something that would be much more regal, but he's doing it the old-fashioned way. Whipped this up himself, bought some trash bags, and took some pitches. Yeah, and, you know, who knows? Maybe the next founder of Cursor, Figma, or Ramp is sitting right now.

Yeah, also 52 pitches in 52 minutes is crazy.

A minute is crazy fast for a pitch. I mean, we do 10-minute interviews, 15-minute interviews, barely get to the meat of the— and this one you got 4 minutes for the founders, 4 founders jumping in, 1 minute. That is remarkable.

Um, I am no stranger to controversy though.

Yeah.

Um, Joe Lonsdale says, I am not a humble man, but this is legit legitimately beyond my capabilities.

Absolutely wild. Uh, well, Vercel Guillermo Rao, friend of the show, is apparently running an ad campaign on Lyft by buying custom license plates and deploying them through Lyft drivers. Is that what's going on here? No, you think it's random?

The guy, Peter, the driver, was like, "He must love Vercel." Love Vercel or work there or something?

I don't know.

He was the 8th employee at Vercel. I don't think he'd be driving Lyft.

Hopefully not.

Unless he just loves, truly just loves the game, loves driving.

Or he's just super illiquid. He's just like, pay me zero, actually. I'll drive Lyft.

Please.

I want all equity. I'm super bullish on Versa.

That's a possibility.

That's a possibility. That's a possibility. Well, Alex Conrad says, is your startup even sponsoring Lyft license plates yet? It's an outside-the-box strategy. Energy. Someone should pick it up. Someone should do it. Get a bunch of license plates for cars, rent them out to Lyft drivers, get those impressions. Wix is down a bunch. This seems like a very logical company to suffer in the age of vibe coding. People are vibe coding websites all the time. And Wix is a supplier, a service to build websites. Sites based on templates. But Wix was buying 30% of its shares at $92 6 weeks ago, but the stock is now down another 45%. I was wondering about this. I almost asked Max Levchin about this yesterday. When you're going through this world, it seemed like he was very confident about the SaaS-pocalypse and did not feel the need you don't need to respond or take any dramatic actions, just sort of wait and let the metrics do the talking. But I was wondering about, you know, are you tempted as a CEO when your stock trades down on a narrative that you know does not apply to you, but you're just sort of collateral damage, are you tempted to do a quick buyback and just sort of, you know, get a good deal on your stock? Stock, even if it's just 3 months down, then right back up.

Imagine being a public company CEO and buying back your stock and then getting a return on it. It has to be one of the most euphoric experiences.

Yeah, totally.

Not actually getting a return, but obviously, you know, decreasing or increasing everyone's—

Well, Wix is a $2.9 billion company now.

Yeah, they acquired this company Base 40 Base44. Remember, this was like a one-person—

Oh, yeah, that's right.

One-person company, and they were growing revenue. I think Base44 has been growing revenue quite quickly. It seems like pretty much any of these vibe coding tools, just the experience is so magical for people that a lot of them have grown revenue really quickly.

It's fascinating when you look at the stock chart, if you zoom all the way out. So during COVID COVID-2021, ZIRP era, stock was at $300 a share. It's at $52 today, by the way. It traded down after ZIRP era ended all the way to $50 a share, $60 a share. And then post-ChatGPT moment, 2024, fantastic for the stock. It gets back up all the way to $250, $200 a share. But then since 2025, as AI has gotten better at coding, vibe coding websites, doing frontend design, there has been a significant sell-off that continues today.

And so, rough goings over there. I was looking to get a comp. I looked up Squarespace. Squarespace is no longer publicly traded. It was traded on the NYSE, but it was delisted after being taken private at $7.2 billion. That is tough timing, taken private on October 17th, 2024.

Oh, interesting.

And at the time, there was not a SaaSpocalypse narrative. You couldn't one-shot a beautiful website with a single prompt. It's going to be so hard for this firm to make money on this deal.

Yeah, it feels like a new customer problem just because it's not the hot new technology that you're hearing about. Like the podcast ad conversion has to be a lot worse. But I would be very interested to know. What is retention like? Because I know some people that have built these—

Yeah, it's one of the most predictable businesses, basically.

Yeah, I know some people that have built these website generator companies, and then they just keep growing and growing and just sticking around forever because once someone has the magical experience of building a little website for their company or their personal brand, and then they just let it run forever, and they're like, ah, $10 a month, I'll just let it keep going.

Well, you see— Yeah, so SquareSpace Squarespace had done around $1 billion of revenue in 2023. I'm assuming they grew into 2024. We don't have the full year numbers because it was taken private in Q4, but is pretty reasonable revenue multiple.

Yeah.

But if they lose out on a lot of those new customers because there's every single company in the world, every single company in the world, it seems like is trying to make a box that will make you a website.

Yeah, everyone. Anyway, you know what very few companies are making? A nightstand that turns into a bat and a shield for defense. I like this. It looks so unassuming as a nightstand, very believable. No one would guess, but then something happens, you grab your bat and shield and you're ready to rock. Would you pick one of these up? It has a little bit of a hotel vibe to it. It doesn't—

Don't.

And also I like a nightstand that has—

All I would say is don't. Don't bring a nightstand to a gunfight.

Okay. Yeah. Well, people are having fun with the AI-generated videos showing that yes, in fact, if it's not bulletproof, it has some trouble. If you disable, Ben Thompson says, if you disable Open@Login for the Gemini app launcher that the Gemini app installs in the background without asking, Gemini app launch will immediately re-enable Open@Login. I will now, needless to say, delete the Gemini app and don't intend to install it ever again. And so this is very, very odd. Gemini login. Oh, so it automatically logs in no matter what. He says, I'm actually struggling to remember a bigger middle finger to a user from an app ever. It's bad enough to install a helper app, but to immediately undo the user's explicit setting change. Incredible. And Josh Wood, Woodward from Google chimed in and said, "This is a bug. It will be fixed in the next release aiming for right after Google I/O. More if you're interested." So, that's good. They did receive the feedback. Well, we should talk about Nikita Beers. I was reminded of this because he screenshotted and posted it. The greatest growth hack of his career for one of his projects. This happened— was this a year ago? Gas or explode? Explode app. This was, this was about a year and a half ago, pre-joining X and working with Elon Musk over at X. He launched a company called Explode, or an app called Explode, and he had a very interesting growth hack where he incorporated the company as TapGet Inc. And so in the iPhone App Store description, under the name of the app Explode, It would say tap get and then right below it would say get because it's a call to action and it doubles down on the call to action.

He made the entity a call to action.

It's genius. These little, these little things really add up and you've seen them all over X. Um, and he's done a good job of creating re-engaging areas. And I just feel like the UI of X has been improving significantly. I'm really, I'm really enjoying the latest UI feature where if you're watching a video and you want to speed it up, you can hold on the right side of the screen, which is fairly common in video apps these days. Doesn't work in the iOS native video player. I don't even know if it works on YouTube, maybe. But what's really cool is that if you press and hold it, you will temporarily be in 2x speed mode. But now in X, if you press hold and you're in 2x speed mode and you drag down, it fills draws a little circle and keeps you locked in that 2x speed mode, and it actually changes the speed of the video permanently until you change it back. And so that little delightful touch is something that I'm seeing more and more of from the X team, and I'm a big fan of. Well, without further ado, we have Andrew Feldman from Cerebras in the waiting room. Let's bring him in to the TV Van Allstrom. Andrew, great to see you again.

Looking sharp.

Feeling sharp. How you guys doing?

Congratulations. How has the day been? I would love to get just your reactions from the day. It seemed like there were a lot of people there. Take us through your, your emotions today.

Well, you know, this was better than, than we'd hoped for. I think a chance to celebrate. We did bring a lot of people from the company and we brought families and to, to share with the team. We brought everybody who'd been at the company for longer than 9 years and their families. We, you know, when you do a startup, the family is a meaningful part. It takes patience from them and a great deal of it. And so they came and we celebrated it. It was really an extraordinary day. We opened up, you know, we did, we priced at $185, we opened up up $3.50 and we settled at about $3.20. What an extraordinary thing. We're just so proud.

Yeah. Take us through some of the history of Cerebras. Has it been a straight shot? Has it been an overnight success? How do you characterize it? What were the darkest moments? What were the highlights? What are the good old days to you? What does that mean?

Well, look, I think in the hardware business, business, if anybody tells you it's a straight shot, you can call BS. I just don't think that's the way our business works. I think the first time you build a chip with a new architecture, it's a little more than a prototype, a little more than a proof of concept. The second chip, you iron out your challenges and you begin to show it to customers en masse. Third one often that really takes off. And so it's a long long road in innovative hardware designs. And so we were founded in 2016. We're more than 10 years old. We sought to solve problems that others— that's right, overnight success. Exactly. Like a decade. I was 15 pounds lighter in weight class.

As most overnight successes are.

Yeah, that's right. I mean, they're just overnight because most people sort of weren't paying attention. But we tried to solve some problems that other people thought were impossible. As we showed you last time, you know, we tried to build a chip that was the size of a dinner plate.

Yeah.

And everybody told us it was impossible. And the truth is, for a while it was. And, you know, we didn't solve it until August of 2019. We built this extraordinary chip. We were faster than everybody, and absolutely nobody Cared, nobody. And, uh, AI wasn't ready and it was still sort of a novelty, and nobody cares about how fast you are when it's a novelty. But, but starting with, with GPT, uh, and in 2025 the models got so darn smart they became useful, and suddenly everybody wanted to use AI, and you use it with inference, and, and business was rolling.

Yeah.

What were those early rounds like? I'm thinking the Benchmark round, Coatue, Eclipse, a bunch of others.

You know, we had the advantage of the founding team had been together at our last company that had paid pretty well for the venture capitalists and the team. And so we had some wind in our sails when we went out and raised money. It's not like today where we're 4 guys in the the word lab and you're raising a billion pre for your A. That's not us, but we went out, we made 8 calls, we got 8 term sheets, we chose Benchmark and Foundation and Eclipse.

That's amazing.

And we got going, you know, less than a year later.

I was expecting you to say like, yeah, I mean, it was a slog.

We were so long.

Those rounds were a slog. Other rounds were a slog. At the beginning, not so. You know, Thomas Lafont at Coatue came in shortly thereafter and we did a round with them. I think the truth is between about 2020 and 2023, it was, it was much harder.

Yeah.

AI was sort of in this situation where everybody was saying, oh, that's cool. Look what this model can do. Look how big it is. Is, but it wasn't being used anywhere.

Yeah, right.

Nobody was using it. Yeah, they were pointing at it. They were saying, wouldn't this be nice? And they went back to whatever they were doing before. And it wasn't until really sort of 2025 when the models got good and you just saw this tidal wave of people using AI and demand for AI compute. And that's been exceptional. It's just been an amazing thing to ride.

Yeah. You mentioned like if you have 4 guys and your company name ends with lab, you can raise $1 billion. There's a little bit of that going on in the market with just like chips, semiconductors, AI. There's not that much that needs to be explained. But what were the key ideas or thesis that you needed to explain in the roadshow to investors that wanted to go a layer deep, but be a layer deeper deeper than just AI chips?

Yeah, I think there were— there's the first, the market size and dynamic. And I think Jensen said some time ago on, on Brad Gerstner's podcast that the demand for, for inference will grow by 1 millionx and nobody believed him.

Yeah.

And, you know, at the same time you saw Sam Altman, you know, displaying real vision and going out and trying to lock up huge amounts of compute and memory and data center and power because he saw it too.

Yeah.

And I think trying to share what that means, what an exponential demand means, and that we're still so early and yet the demand for AI compute is overwhelming. I think sharing that was interesting and I think helpful in educating the financial community. The other thing is that there are lots of ways to do this. The GPU isn't the only way. You've got a TPU, you've got Trainium, you've got us. There are lots of different ways to build a solution here. And finally, that maybe the notion that CUDA is sort of this grand lock-in is overplayed and that, you know, the Gemini 3, which is an excellent model, model was trained on TPUs with no CUDA, that Anthropic's models were trained on Trainium with no CUDA. I mean, that lo and behold, some of the best models, some of the most interesting things are being done without CUDA, and that, that lock-in might be overplayed. And I, I think these three factors were really important in, in educating the, the financial community.

Going forward, how do you think, uh, how do you and the team think about sort of calling your lot and sort of trying to predict where and how inference demand will look in, you know, 2030 and beyond versus like working closely with the labs that now have, you know, product lines with billions of dollars of revenue and their own roadmaps that you can work with?

Yeah, you know, like the Babe, I'm going to point out to left field and just say, wait, this is where it's going, baby. No, I don't think that's the way it works. I think we're calling our shots every day by making big investments in data center capacity and collaborating with the leading visionaries in the field in working not just with OpenAI to serve as sort of the cutting edge and deliver their extraordinary models, but also with AWS to make to make sure that we can get access to the largest enterprise customers. And instead of having to work with these enterprise customers, procurement sort of organizations who provide master purchase agreements that are the size of a Bible, you know, you can say, look, why don't you buy us through AWS, and it'll count against your annual commitment. And so I think those are really important ideas and ways we get access to the market. And then we're taking huge amounts of data center capacity. And so that's the other bet we're making.

Yeah, makes a lot of sense. How do you think the year will play out in terms of just broader consumer awareness of what fast inference feels like? I had a really magical moment using Suriya Jubris in GPT-5.3 Spark and Codex. And even outside of coding tasks, just talking to the model and having it respond instantly was sort of, it felt like a new breakthrough or new paradigm. And I feel like this hasn't fully diffused, but it also feels like when it does, there will be potentially like entirely new ways of working, entirely new paradigms might emerge. How are you thinking about actually diffusing the technology?

We think that's exactly right. And we think that the experience of engaging with a real-time AI. Yeah. Will, will encourage people to do more things, to stay longer, to work on harder problems and to invent new things. I mean, if you remember, you know, when Netflix started, they delivered DVDs in envelopes. Right? And when the internet got fast, they became a movie studio. They didn't get better at DVD delivery, they became something completely different, something that had never been in existence before, a movie studio that delivered directly to your home. I think that's exactly what's going to happen. And you can just sit back and you can ask yourself, how big is the market for slow search?

Zero.

How big is the market for dial-up internet? I mean, how much would I have to pay you to swap out broadband at home and bring in dial-up, right? Is it $1,000 a month, $1,500 a month, $2,000 a month? I mean, no way. I mean, it just wouldn't be worth it. And so the community is going to engage with inference in the same way. And that fast inference is going to be all of the market.

Yeah.

So you make the chips. I believe you also make cooling infrastructure as well, cooling units. Is there— are there other products on the roadmap roadmap that you think will be required to roll out and scale Cerebras over the next couple of years?

No, I don't think so. I think right now we build the chip and the system, and the system includes— it's about the size of a dorm room fridge. There you put two of them in a standard data center rack, and the cooling infrastructure is built into the system.

Sure.

And I think And that's where we want to focus. We want to be measured on our ability to build AI computers that are faster than anybody else.

Yeah. How are you thinking about scaling on-chip memory? It feels like there's some, there's some concern about, well, what if the models go to 10 trillion parameters? What if it gets too big? How are you thinking about that challenge? Or maybe it's an opportunity.

It is an opportunity. I think a 10 A 10 trillion parameter model is hard for everybody. It's actually easier for us.

Okay.

Right. There's a reason we're not a 10 trillion. It's because it's really hard and expensive to serve for everybody. I think one of the things that we've been able to do for the larger models is to tie together a bunch of these systems in parallel and run them as a pipeline. And that way we can train and do inference on trillion, multi-trillion parameter models in ways that I think are much more intuitive than, than on GPUs that have much smaller compute. They have off-chip memory, but their problem is the compute. They don't have enough compute per chip.

And then how are you talking to, to, to customers about potentially bringing Cerebras in not as a full replacement to their entire semi- semiconductor supply chain or stack, but as a complement to everything else that they're running. Because I have this vision of like the next generation of AI agents. You get this genius model, but it needs to use a small model over here, an open-source model over there, a super fast model for a certain thing if it's looping through something.

The same way you hire, you have a superstar employee. Yeah. You don't necessarily want them doing every single task themselves. Is like, yeah, you should be able to delegate.

Yeah, delegation. How are you thinking about that?

Yeah, I think that is sort of a notion of a confederacy of models, right? That there's a collection of different models. And one of the things we thought about early on was how to interoperate in that environment.

Okay.

And we connect in via standard 100-gigabit Ethernet. Nothing fancy, nothing proprietary. We are deployed in, in many places where they've got GPUs from, from NVIDIA or GPUs from AMD. They've got x86 compute from Dell or HP. And so that's not a problem at all. We're eager for those environments.

Yeah.

How— what, what do you think the company would look like today if you guys had had access to today's frontier models when you started the company? Like, are you feeling And how do you think about just like the speed up in, you know, at the company today due to how good the models have gotten?

We use frontier models every day in coding, in running our G&A. I think if you start a company today, you build a very different organization. I think there are whole departments that look different in the next 9 to 18 months, I think. I think much of what HR does, much of what training does, is solved by, by some form of AI. I think a lot of the work in finance, right, closing the books, a bunch of what they do is, is checking. And those are all done by agents. I think what it is to be selling or doing recruiting, those change. I think for a long time what recruiting was, was hunting through or writing scripts for LinkedIn. I think that changes substantially. And so when we look out, we see sort of fundamental changes. The obvious ones, of course, are, you know, a year ago engineers were using approximately zero tokens and now they're using, you know, $10,000 worth of tokens a month. And the rate of change and the rate of new PR requests, new pull requests is just extraordinary. And so AI is having fundamental changes. Obviously, it usually starts in Silicon Valley and sort of works in waves to, to other areas. But that's what we're seeing right now.

Since the last time we talked, there's been a ton of movement in the space data center market, a lot of energy. Just yesterday, SpaceX and Google eyed a launch deal in the Wall Street Journal. Have— has any of your thinking changed? Like, what is your current thesis on space data centers and how it might fit into your business plan over the next decade even?

Well, one of the hardest things in a space data center is communicating across chips from one chip to the next. And we solve that, right? I mean, one of the great parts about a big chip is that you have to communicate from one chip to the next less frequently. It's a huge advantage for us in space. Space. I think that this is an idea like self-driving where the last 10% takes 80% of the time.

Sure.

Right? And that we're not 3 or 5 years away. We're 8 to 12 years away. That doesn't mean we shouldn't be working on it or thinking about it or making progress to it because if you don't do that, it's 25 years away.

Yeah.

But I don't see data centers in space in the next 3 or 4 years.

Yeah. And arguably you've solved the key problems that you would be asked to solve. And so you'll be ready if demand shows up. We're ready. There's not that much for you to do individually to advance that. That makes sense.

That's exactly right. Yeah, that's exactly right.

Well, we're hoping for it. It'd be exciting, but plenty, plenty of work to do here on the ground.

Congratulations to you and the whole team on this incredible milestone. We're honored that you would spend time with us.

We really appreciate it.

Historic day for the company.

Yeah. And let me hit the golf.

Incredible to watch your progress, and I look forward to your next appearance.

Yes.

And enjoy the rest of the evening.

Enjoy the rest of the evening. We'll talk to you soon.

Thank you guys. It's time for a cocktail.

Be well. Fantastic.

Enjoy.

You deserve it. Goodbye.

What a fantastic— I love that analogy. He's like, I'm Babe Ruth. I just point.

Then he's like, no, I'm not going to do that. It's way more complicated. I've been working on this for a decade. Yeah, what a fantastic story. What a fantastic performance. I'm very excited that we're bringing in Eric Visria from Benchmark, who was in that Series A that Andrew Feldman just mentioned. So we will talk to him about that in just a minute. We're going to bring him into the room. The waiting room. But there are some other posts that we can talk about in the meantime. One, someone is using Runway ML to create— let me see this— a full hurricane inside a TV studio. I want to watch this clip, and it's 1 minute, and we will see how convincing is this. Are you going to be turning off the news in order to watch this? But wind is only the beginning.

The real danger is when the storm starts moving.

This is very cool.

As the storm builds, ordinary things stop feeling ordinary. Roof panels.

So you think audio also, like fully AI-generated? Because the typical workflow for this is, uh, the, the new— the host would, uh, stand on a green screen or LED volume, and then, uh, all of these effects would be, uh, added in post or live through like a traditional visual effects pipeline. This feels fully synthetic. I think that you'll probably use some sort of hybrid approach, but the ability to prompt something like this on the fly for a small news organization that maybe doesn't have the budget for a huge VFX team, you're just gonna see a lot more VFX like this. You're gonna see stuff all over the place. There are so many small news channels, local news stations that just don't have the, the, you know, access to Digital Domain or some huge visual effects house. So looks pretty good.

Before our next guest, yes, Dylan Field has a quick update.

What do you say?

Have their Q1 results. He says, quick update, not dead and putting up some insane numbers. 46% year-over-year revenue growth accelerating for the second straight quarter.

They're raising 2026 revenue guide 6.8% today, up 8.6% after hours. Congratulations.

Founder says design matters more than ever. The Figma team continuing to execute incredibly well.

Fantastic news.

Let's bring in—

bring in Eric. Welcome to the show, Eric. Congratulations on the progress. Thank you so much for taking the time on such a busy day. Great to meet you.

Great Great to meet you guys. Excited to be here.

Long, long overdue.

Yeah. Crazy that this hasn't happened already.

Well, you guys like Ev, you know, so you have Ev on.

You don't have to have me.

He was a former colleague, but everyone is welcome here. But I would love to just hear the story from your perspective. We just heard it from Andrew's perspective. It seemed obvious, but was it obvious to you? Yeah.

Where was it the most obvious deal ever? Because I was— we were talking with Andrew Andrew, I was asking him for the story of those first couple rounds, expecting him to be like, you know—

It was really hard.

ChatGPT wouldn't come out for almost a decade. It was a slog. We kept getting— we walked up and down Sand Hill Road. We got nosies. Like, yeah, we got 8 term sheets. So clearly, it was a deal you had to win.

But take us through it.

Well, you know what? The hilarious thing about it is, in venture, it's very useful to be naive. And certainly I was so naive about how hard hardware actually is. Like I just, like I can't even describe to you guys how naive I was and we were. You know, it was 2016, deep learning was clearly gonna become a thing, which would obviously evolve and empower the AI that we have today. And I was looking at all of these different applications. So I was looking at like deep learning for radiology and security and other things. And it was really hard to figure out where it was going to work, like which application was going to take off. And you guys have to remember, this is, this is 2016, right? The TPU hadn't been announced. The transformer paper hadn't come out yet. LLMs haven't, hadn't been born yet and obviously not ChatGPT or anything else. And so it's really early, but there was clearly something there. And when I first met Andrew, he came in and I was like, we hadn't— we're not hardware investors typically. I think our last hardware investment before that one was Ambarella, which was 10 years earlier. And he came in and he said, you know, it was like the team slide, very impressive. And then, you know, the slide 3 was GPUs actually suck for deep learning. They just happen to be 100 times better than CPUs. And as soon as he said it, it's just like a light bulb went off. Like, of course, of course. Like, why would a graphics processing unit be the right solution for deep learning? And then, you know, of course, he proceeded to explain like why GPUs were so much better than CPUs for training and also what the ideal ground-up solution could look like. And they had their idea of the wafer scale and everything else. And as soon as he said it, it's kind of like, oh yeah, that makes sense. And like, I should, you know, like we don't know what application's gonna work. We should invest in infrastructure. This is an amazing team and a really provocative idea. Fast forward, like that was 2016, spring of 2016. You fast forward like 6, 7 years and like we're still slogging it out. Out and have raised so much money and have very little revenue. And, you know, and it's just, it just hadn't all come together yet. And then, of course, over the last 2 years, inference is exploding. It turns out Cerebras switches from training to inference and, and really focusing on inference and making inference speed where speed matters. Coding explodes where speed really matters. And so all these things kind of came together. And so, you know, a lot of luck, a lot of naivete on my part. But for the team, just relentless grind, never giving up, always taking feedback, but being persistent, being open-minded about where the market was going. So, yeah, I'm so, so proud of them.

Yeah. What was your role as an investor like over the journey of the company? —because obviously, Andrew and his core team, deep engineering bench. Were you focused on how you position the company, the private markets, fundraising, or management? What were you focused on in terms of value-add or just helping build the company alongside?

I'm really the algorithm specialist. I go in there and I do that. No, I'm just kidding. I don't know anything. You're in the fab.

You're the one that's making the chips. That's right. I was making it up. Clean room. Yeah.

It really changes a lot over the course of a company. This is, I think, the 4th company that I've worked with for more than 10 years. Wow. And so when you work on them a long time, the companies evolve a lot, right? You start out, it's just 5 people. It's just the 5 founders originally. And, and so at different points in time, it's a lot of fundraising help. At points in time, it's like really helping build out the broader management team. And a lot of it is also just being someone for the founder to talk to. You know, this— there's— being an entrepreneur is very— the highs are very high and the lows are very low. And, and so someone you can talk to and be really open with that, like, helps moderate that. And I think that's a part of it. So it's just, it's an evolving, you know, conciliary kind of role. And I really love it, actually. That's the part of the job that I love the most. And it's very, it's rare and special to have these kinds of relationships. I've had a few of them. I'm very lucky to have a few of them where I just feel really like a lot of chemistry with the founder.

And just feel like we have a really productive relationship. Where are you excited to invest over the next decade? Because, you know, it feels like we're still in a semis boom. There's a lot of opportunity there. You could go deeper into that side of the business, but then there's so much software.

I'm sure you've gotten pitches that look like what maybe would be the next gen and, you know, maybe like, "I already got my horse." Yeah, well, yeah, that, but then, You know, talking to these teams that don't necessarily know what it'll actually take, right? Sure. They don't. They didn't learn the hardware is hard lesson.

Yeah, yeah, totally, totally. Well, you know, one of the funny things— and I ask myself this question all the time, obviously— is, you know, this is a 20— for us as early-stage investors and looking for, you know, really big outcomes but willing to take big swings, you really do have to kind of look many years forward and try to see like what's going to ripen at the right time, right? So in 2016, you make an AI hardware investment and, you know, Groq was, I think, 2017, for example. So like, you know, there were several contemporaries of them. And of course, Groq and Cerebras have ended up doing really well. And so you— but you're trying to say like, okay, this fruit's going to ripen in like 6 years, right? And so there's kind of some mention of projection. You know, right now I think I'm really excited and continue to be really excited about a lot of the AI applications. We're investors in Sierra and Lagora and a number of others that where like they're obviously booming, they're selling magic to their customers and the companies are doing great. We also have these like infrastructure investments like Fireworks, for example, which is also riding this enormous inference demand. And then there are kind of things that are a bit more forward-looking. We invested in Star Cloud. My partner, Chafin, led our investment in Star Cloud, which, you know, space data centers. And we also, you know, we led the initial round in Sunday Robotics, which is a home robot. And so, you know, I think those things are gonna take longer. You know, they're, they're not going to be, you know, massively scaling revenue like next year. Like, that's not, that's not what they are. So you kind of have a combination of these different things, which are— but it's kind of trying to figure out when they ripen.

Next time you come on, we got to have you debate Deleon because he came on and was debating EV and hardware versus software. But you got space chips, you got everything Deleon likes.

Yeah. Well, it's nice. It's nice to have a portfolio. And I think, you know, one of the beauties of Benchmark is each of the partners is attracted to different things and different types of founders. And so we, you know, you put it together and it works out really well.

Yeah. Yeah. That makes sense.

Walk us through funds 7 and 8 because there's chatter on the timeline as those funds being some of the best in venture history. And although this is Cerebras day. This is your first time on the show. So we do have a big gong here.

Yeah, well, I, you know, Fund 7, Fund 7 has or had Uber, Snapchat, Elastic, Stitch Fix, WeWork. I mean, there were so many things. It was like, it was such an embarrassment of riches. And I had nothing to do with that fund, just to be clear.

Like, I I joined in 2014.

That fund was already deployed and invested in, but the team, you know, that team at the time just did such an outstanding job with winner after winner. Discord is in there. I mean, it's like really—

wow.

Like when you have, you know, guys like in venture, if you catch the trend right and obviously work hard and get lucky, but you have you know, the 6th or 7th company in the portfolio delivering a multiple of the fund or something like that, like that, you're in such rarefied air. And that's— there's— it's really special. So that's Fund 7. You know, Fund 8 is a very enterprise. It's our 2014 vintage, I think. And, you know, it has— it's a very enterprisey fund. And so, you know, we had Confluence, which returned a bunch, and Amplitude had returned a bunch, and, you know, and then we have Cerebras, obviously, which is big, but Chainalysis is in there and several others. And so it's kind of interesting how these, how they switch. I think that's actually more interesting to me, which is Fund 7 was very consumer mobile, and Fund 8 is like very enterprisey, and they're like back to back, but they turn out to, you know, they both work. And so, you know, I think that that tells you a little bit about what, what venture is and how we, we all have to be really open-minded about what's happening and what's the right timing for these various ideas. And then, you know, fast forward and our 2022, I think 2022 vintage has, you know, the first round of Sierra, the first round of Fireworks. The first round of Agora, you know, Mercore, Reductor, Mercore. Yes, absolutely. LangChain. And so, you know, all of those are in there. And so obviously that's a totally different fund and has, you know, a different set of things. But also, you know, looks, looks pretty interesting. So it just, it evolves. And that's what's so hard and tough about this business is staying on your toes. When you're in a very, very dynamic world.

Yeah, well, it's interesting. Something that, you know, this has been talked about on plenty of podcasts, but it's worth bringing up. You guys have stayed true to the strategy, and you can count on the market changing and evolving. But a lot of funds are, like, having to deal with markets changing and evolving while having a fund strategy that is changing and evolving. If you keep one of those things true, it seems, at least from Benchmark's track record, that it gives you some advantage in that you're playing a very specific kind of game and not having to evolve your own game while dealing with changing technology trends and markets.

I've been at Benchmark 12 years, and I've thought about this a lot. You're watching your peers do all these different things and swimming in fees and all these amazing things. And so you're like, wow, that's pretty, that looks pretty cool. Um, so like, you know, you kind of like look at this stuff and, and, um, but I'll tell you, you know what I think it actually comes down to? It's what it actually comes down to is what do you love doing? And, um, you know, we're obviously in a very fortunate position and, and I inherited an amazing platform. And so, um, and very fortunate to have done that. And so you know, we're in this amazing position where you get to do what you really like doing. And at the end of the day, we really like partnering with early-stage founders and working on these companies for a decade plus. And that's kind of what we like doing, you know. And so I think things have definitely evolved. The opportunity set is changing and evolving. And, you know, more recently, I mean, just in February, we raised an SPV, which we've never really done before, and to invest in Cerebras. And that was unusual. But it was, you know, you can also— we've actually a few years ago we did public market investing when, when COVID first hit and the Nasdaq tanked. You know, all of the early stage stuff just disappeared. We were like, wait a minute, like these publics, like there's interesting stuff in public. So we started deploying a little bit in the public. So, you know, Well, yes, we're really focused on the early stage and that's what we love doing. And then also occasionally, like, we see these special opportunities and we try to jump on them.

Well, yeah. Well, thank you so much for coming on during a business day.

Had to sneak in that, the SPV round of $23 billion. So congratulations on that investment. Fantastic. Another little cheeky 3x.

You deserve a drink. Hopefully you can find Andrew and cheers. I'll definitely have some drinks tonight. Have a great time, Salvador. Well done.

Great to finally meet you and congrats to everyone.

Yeah, let's do it again soon. Thank you, guys. We'd love to do that. Thank you.

Cheers. Goodbye. Up next, we have Steve from Foundation Capital. He's Cerebras' first term sheet investor, also the first investor in Solana and a bunch of other great company. So we will bring in Steve from Foundation Capital from the waiting room. Steve, how are you doing?

There he is. Doing great.

Sorry to keep you waiting. Congratulations. Thank you so much for taking the time to come chat with us. How are you doing?

It's just another, another, another day. You didn't— are you at the Nasdaq or you're calling in from home?

Yeah, exactly. No, just another day. No, I am. I'm at my hotel on my way to the dinner that Eric's also headed to.

Okay, we won't keep you too long, uh, but I would love to hear the story of you meeting Andrew Feldman in 2007, how things, uh, matured from there, how you wound up working together.

Yeah, so, uh, I showed actually Andrew the email last night over dinner, but, um, yeah, he and I and Gary met in, uh, October of 2007. They were raising money for the company that they started prior to, to, uh, Cerebras, which is called C-Micro. Yeah, uh, and it was kind of broadly in sort of new server architecture. So these guys have been thinking about these kinds of problems for a long time, but I passed on the investment but stayed close. We really connected in that meeting. And then when I saw them get acquired by AMD, it was about 4 or 5 years later. I was like, guys, Andrew in particular, you guys are not going to stick around this company for too long. So let's start riffing on some new ideas. And that began basically a 2-year conversation. About a whole bunch of ideas. Actually, it all started really in kind of this concept of warehouse-scale computing. We were looking at companies like Mesosphere, ended up actually doing a small investment there, and CoreOS and a whole bunch of others. And Andrew came in in November of that year of 2014 and shared his ideas with our enterprise team. And then basically we riffed on ideas and in the spring of 2015, 2016, so it was like March timeframe, we started telling them, look, we want to be your first term sheet. We've been courting each other for a while here. And yeah, we got them a term sheet to lead that first financing. And then Eric stepped in, and we changed the terms a little bit to make room and co-lead along with Eric and Pierre from Eclipse. Yeah. Yeah. And then they started it right in our office. That's amazing.

Can you talk to me about— there's, you know, crypto and AI feel like two wildly different technologies, but there's a ton of overlap everywhere you see from, you know, crypto miners pivoting to Neo Clouds. There's a lot of movement back and forth. And I'm wondering like what in your mind the similarities, differences are, like why you've been drawn to both over your career, where the gap is, where there's similarities.

So what I would say, the similarities, which are probably in retrospect somewhat obvious, I would say the hardest problems of software and systems live in the area that we're working on in AI. So the AI infrastructure, the Frontier Labs as well, all the work they're doing there. And the same thing is also true at the bottom of the stack, the layer ones and the very hardest technology technologies over in crypto. The folks that are attracted to both of those areas tend to be very technology-driven. They love distributed systems. They love the hard problems around cryptography and elliptical curve cryptography. They love low-latency computing. They're quite similar in terms of being systems thinkers. And so, those are the ways in which I would say that the problems are quite similar. And in fact, In fact, here's a funny anecdote related to this. So, Anatoly Yakovenko, co-founder of Solana, part of the reason why he chose to work with us back in March of 2018, so about 2 years after we invested in Cerebras, was because we were investors in Cerebras. Oh, no way. He was like, you guys take hard problems seriously. He had spent 12 years at Qualcomm.

That's right. Yeah. Distributed systems.

Running the operating system. Exactly. Was at Dropbox and understood those challenges. And so he said, wow, you guys care about these kinds of hard problems and that matters to us. So we ended up doing a fair bit more diligence and writing actually a larger check into that very first Solana financing.

So yeah, can you take us back to earlier in your career, pre-investing, obviously fascinated, hard problems, but like where does all that come from? Does it start in high school, college, early career? Walk me through some of the early days.

So I studied robotics and embedded systems, sort of the intersection between mechanical and electrical engineering in undergrad, and then came to graduate school and did more of that. And then my very first Friday at Stanford, I met David Kelley, who's the founder of IDEO, which is a product development consulting firm that worked with the very best kind of Fortune 1000 companies when they would hit a snag. A hard problem, or want to invent a new product, and they didn't often know how to wrestle those challenges to the ground, they would call us. And so, we did a lot of work for Apple. We did a lot of work for Cisco. We did a lot of work across every industry from healthcare to consumer devices to really hard problems in systems. And so, worked there for 5 years designing products. In fact, saw one of my other products, uh, earlier today on the desk, uh, at the trading floor at NASDAQ. It was Cisco's voiceover IP phones, which I worked on now 28 years ago. No way. So, so just working on cool, cool things, hard problems, um, mostly, uh, where it feels like if, if you solve that problem, it was worth solving. There's, there's a, there's a real prize at the end. Okay. I want to— that's how I got started. Yeah.

I want to take the bring this full circle then, because robotics is sort of having a moment, but it still feels like it's early in terms of, as a consumer, as optimistic as I am, I just don't think I'm going to have a humanoid robot walking around my home this year. Most people we've talked to have said, yeah, it's maybe 5, 6, 8, 10 years away, but that's like the perfect timeline for a venture capitalist to start getting involved. You don't wanna be trying to build custom AI chips today, you wanna start 10 years ago like Cerebras did. So how are you thinking about the, like pulling your experience from robotics into the modern era? Because if the boom isn't already here, it's probably gonna be here in a decade, if not a decade, two decades, like it's coming, robots are gonna be real. So how are you thinking about it?

So we've done a fair bit of work in embodied intelligence in terms of research and, And as I'm sure you're familiar, it's always a little tricky to invest in an area that you have some operating experience. It tends to bring some scar tissue. And so you might be more circumspect than if you'd had kind of a beginner's mind. I would say I am generally not a big believer in the humanoid approach. I think there are use cases, for example, in the home companionship. And even in that case, it's a bit of a stretch. I think you need to think about robotics more broadly and think about industrial automation and then look at the problems that are not necessarily kind of the consumer-level use cases. But you walk the factory floor and you see people moving around pallets and the human form factor is not good for moving pallets around. And so you wouldn't actually build a humanoid robot if you were trying to deal with that use case. I think when I zoom out and I say, what are robotic systems? Robotic systems are basically ways of automating, automating human labor. And so, and in fact, the greatest compliment for most of these systems is when you stop calling them a robot, you actually call them a forklift or you call it a washing machine. Oh, that's a great— And it's when that technology diffuses into the background and you just focus on what is the application. So that's how I look at it through kind of the product lens as opposed to the technology lens.

Yeah, yeah, I was, you know, you see these demos of humanoids loading washing machines and I've been thinking in the back of my head every time I've interacted with my washing machine, like, is it time just for a ground-up, first principles rebuild of what a washer and dryer stacked is? Like, if you constrain it to like, you have this dimension, but now you have all the modern technology and your goal is to just take in dirty clothes and put out clean clothes, Can you do something better than just a big tumbler and then another tumbler, one with water, one without? And I'm excited by that. Is the implication of that that almost you would be open to talking to entrepreneurs who are maybe thinking a little bit narrower, thinking a little bit smaller, at least in the interim? And then how would you guide someone towards long-term messaging around their company if they are finding finding a wedge, but then they want to grow at some point.

Yeah, so I think it is exactly what you just described, which is, and again, the sort of the applications do matter here, but the notion that you would start with something that is, let's call it sort of big enough to matter, but small enough to win. And in hardware technology, being more focused is actually a huge advantage, a huge point of leverage. And then as you continue to build, you want to be able to access larger opportunities in markets. And so I really do believe that that is the way you get started with hard technologies and hardware in particular. I think there's another thing that we do, and I will just say this kind of brings it to Cerebras again for a minute, which is we look at workloads. And so one of the reasons why we backed Andrew and Gary and Sean and team back in 2016 was it was quite clear, and we saw this through the lens of our portfolio, that the AI workloads at that time was more ML. They were ramping very, very steeply. And whenever you see computing workloads that are doing something new and different, and this, you know, you're talking about in the robotics context, and we'll get to that in a second. But when you see a workload that is spiking hard, there's often an opportunity to basically replace the, the compute layer. In other words, there's often sort of purpose-built silicon that should exist here. And so in the case of personal computers, very clear serial programming, uh, and, and you were very well suited to the x86 platform. It was actually something we saw go on and on for decades. As soon as you started to see the need for much better graphics, of course you would build a graphics processing unit, uh, that's really good at, you know, at rendering graphics, at doing floating point math, at managing lots of multiple cores. And then of course, take the mobile era and then you say, okay, wait a minute, what's going on here? I need low power, I need a smaller form factor. And so when you look at these workloads, oftentimes there is this sort of transformative opportunity. And that's exactly what we saw in 2016 was, wait a minute, like there should be purpose-built silicon for this ML and AI workload. At first, of course, we started with training, back to your point around how do you start small? And then 7 years in was actually a board meeting when Sean, one of our co-founders, said, "We gotta go after inference. It's just, it's exploding." And so again, to this point, you start small and then rotate towards the much larger opportunity.

Yeah, I mean, we talked to Andrew about all the ups and downs, a classic overnight success with tons of moments on, of, you know, intense tumult. But I'm curious about, were you ever worried or hesitant that the company might narrow down too much? And because you've heard like, you know, YouTube has custom silicon for video encoding and there was probably an opportunity at some point to narrow the focus even more to do chip development for one specific company, be less generalized. And maybe ramp the revenue a little bit faster. But was there a tension there that you were observing and like, how did you get through those moments?

I'd say that the, the primary tension that relates to your question was probably around making sure we would not silo ourselves into use cases that were traditionally just high-performance computing use cases. Sure. So those workloads are valuable and those markets are actually still relatively interesting. But they're, they're not growing anywhere close to the rate of the inference and specifically the reasoning part of inference where you start chaining workloads together. So we worried a little bit about that being, you know, a niche that was not interesting enough for us to build, you know, a really nodal company. If I zoom back from that and you ask sort of what are the things we really worried about in those early scary days, I mean, there were— I don't know if Andrew shared this— and there were like 5 startups were the hard problems for us to go after. I mean, it was absolutely, there were moments, I was joking with one of the other founders last night, where you would come back from a board meeting and you weren't quite sure whether we were gonna figure out our way through a very fundamental thermodynamics challenge.

Okay, so when you say 5 problems, you're not talking about fundraising, hard negotiation with TSMC, talking to your supplier. —All of that too. All of that true.

I'm talking about the actual hard problems, meaning hard technology problems. And, you know, the ones that are sort of more physical, where you have laws of physics and thermodynamics to obey and you don't get to negotiate. Andrew's a very good negotiator, but he's also learned that he can't negotiate with the second law of thermodynamics. So, no, these were— this was how do you yield a semiconductor that's the size of a dinner How do you power it? How do you cool it? How do you maintain continuity across thousands of connections? How do you put it in a system and integrate it and then in a data center and then put 65 of them or 64 of them in a data center together? So it was those kinds of very hard challenges where I say 5 startups in one. And they were of course also stacked, which means that the risks are now combinatorial. So even more dangerous.

So you've been through taking companies public, being involved with public companies several times. A lot of times the founders that you're backing, it's their first time becoming a public company. What are you telling them? What advice can you share with a founder, not Andrew specifically, but any founder who's going public? How will the company change? What are you telling them as they become the CEO of a public company?

Yeah, so there's a few things that come to mind One is buckle up because it's going to be— particularly in markets like the one we're in right now where, I mean, you see the headlines change every, every few days. I mean, there'll be another drop of another model tomorrow that could, you know, whatever, upend the public markets. And so you don't have a lot of control over what the world thinks about your share price. And so you've got to coach your teams and your engineers and in particular, to know that when the share price is moving, it very often has nothing to do with what you're doing in the day to day. And you just need to steel yourself against that. I think there's also a piece which is you just have to grow up. There is a cadence to these businesses, quarterly, unfortunately. I wish they were longer, where Andrew and Bob are going to hop on an earnings call very soon. And they're going to have to start talking about about the business of the business, not necessarily the technology of it. And that requires a level of discipline and planning that oftentimes founders don't, you know, don't have their stuff together well enough in order to be able to sort of manage through that, that transition. And then the last thing I would say is actually the flip of it, which is don't forget what made you special, because when you get into this quarterly cadence, and you start to think, well, how do I meet the next quarter? You oftentimes lose sight of the long horizon that was the larger opportunity for you to go after, you know, not just, you know, the opportunity right in front of you, but there's much, much larger opportunities. And we're, you know, building systems for the next gen and the gen after that and the gen after that. And so you can get tricked into being in a kind of quarterly mindset. And it's one of the most dangerous things one of the most toxic ways to kill a company that's built around innovation. So you just wanna, you wanna make sure that, you know, there's that horizon that's still calling. That's where we need to go.

I love it. Thank you so much for coming on and breaking it down. Sorry for running long. I'll let you get to the celebratory dinner. Say hello to everyone and have a great day.

Awesome. Thanks so much.

Great having you. We'll talk to you soon. Have a good one. Bye. That's our show, folks. Leave us 5 stars on Apple. Apple Podcasts and Spotify. Another one. Sign up for our newsletter at tbpn.com. See you tomorrow at 11:00 AM Pacific time and have a great rest of your day. Goodbye.