Letter Enthusiasts, Open Source vs. Closed Source, Mega-Startup School | Luke Knight & Ronan Chambers, Gary Vaynerchuk, Jonathan Jacobi, Dean Meyer, Brent Franson, Matej Cernosek, Justin Boitano

You're watching TVPN.

Today is Monday, July 27th, 2026. We are live from the TVPN UltraDome, the temple of technology, the fortress of finance, the capital of capital. Let me tell you about Ramp.com, baby. Time is money. Save both. Easy to use corporate cards, bill pay, accounting, and a whole lot more all in one place. Massive, massive weekend. Crazy debate. You called this out to me a week ago. You said, we've never been in a moment in tech where people have been more aggressive at each other, that there's been like more maybe tech infighting. I don't know exactly what word you used, but you described this idea.

Tensions are high.

It's tensions, rivalries. I think it's just like it's a non-stable equilibrium.

Well, yeah, when you think about the biggest, like with people going at it. Yeah. Like 2021. Yeah. It would be like Keith Rabois saying like, "There will never be another uniform startup in San Francisco." Or like, "Working from home is not doable." And then years later they'd be like, "See?" Yeah, yeah, yeah, yeah. "A big company was built." Yeah, yeah, yeah. And that was people being like, "You're an idiot." Yeah. "Oh, you're an idiot." Yeah.

And now it's like, "I hate you and I want to—" It's not that direct, but it is high stakes because like the market structure has very much not solidified. For a while. And it's different because when you go back to the iPhone and Steve Jobs being mad at Eric Schmidt about Android, that was a tense moment that happened, A, behind closed scenes completely. It wasn't like Steve Jobs was taking shots at Android really beyond like little things in the keynotes here and there, tiny stuff. It wasn't every day on Twitter Apple employees being like, Google's annoying for ripping us off or whatever. And then the market did solidify.. And it was like, okay, Android's like the ad-supported one and iPhone's still printing lots of profits and like things are understood.

Also companies got big or they got bigger quickly. Yeah. But you didn't have companies that where trillions of dollars of market cap was just born in the span of a few years.

Yep. Yep. And then you had like the Mag Seven where you have 7 companies that are pretty evenly balanced all around like single-digit trillions. And so they can sort of like, comfortable with each other.

A lot of them kind of had their own thing that they were like focused on and then they maybe do some other things. Whereas now everyone has a model. Even, you know, NVIDIA has open models, Microsoft, and they own a lot of OpenAI.

Pretty much every one of the Magnificent Seven is an AI company in the sense that they have a model. I mean, even Tesla has a self-driving model. There's a, you know, everyone has like a data center play or a cloud play, something like it's, no one's sitting out the AI boom. But the debate is raging, kicked off by Jensen from NVIDIA joining X. That's how you know it's serious when he's getting in the arena, the real arena. Yeah. Founding a company and growing it to $5 trillion.

That's just child's play compared to getting— He's like, fine, I'll go to the dive bar. Yeah, yeah, I'll go to the internet's dive bar and I'll start duking it out. Yeah.

No, he did. So NVIDIA is leading a new coalition of tech companies that want the US government and its allies to see open-source artificial intelligence as a society, as a cybersecurity asset, not a threat. This is in the Wall Street Journal. Dubbed the Open Secure AI Alliance, the NVIDIA-backed group said Monday it aims to create and deploy a suite of open-source AI tools that any company can use to defend against cyberattacks. The group includes SpaceX, Microsoft, Palantir, and dozens of other companies, many of which publicly oppose restricting access to open-source AI. "It is the latest salvo in a battle over the future of open source and open weights AI tools amid growing concern from government and industry officials about AI systems causing cyber attacks. Those fears have prompted the White House to increase oversight of AI models and push the private sector to deploy AI for defensive purposes." So the backdrop here is, is basically just like we went through a sort of hurry up and pause and back and forth debate over like what it will take to approve new frontier models. And Mythos came out and was private. And then a couple months later, I think the actual timeline is relevant here. So Mythos was April 7th, Fable came online June 9th, and then got pulled June 12th before coming back July 1st. We're now one month later and we have an open source model that doesn't by definition have to go through any discussion or any discussion in DC. Whereas we've been hearing about, okay, Anthropic and OpenAI are going to Washington, DC, sharing the new models, model capabilities. Maybe the government's taking a look at the model card. Maybe there's potential for evals that the government would run and have an opinion on before it goes out. This is basically saying, hey, all of that is nonsense and not going to matter because we're just going to drop the weights and there's not going to be an approval process of any kind. And people are all over the place on this. Some people are thinking about it purely from an economic perspective. Other people are thinking about it purely from a safety perspective. Lots of people have mixed opinions. And of course, all of this is driven by the actual progress in open source. We wouldn't be having this discussion if Moonshot, ZAI, and many others, DeepSeq, weren't delivering really solid models. So open models are free to download and allow users to customize them for their own needs. Amazing for cost savings, amazing for consumers, amazing for businesses. I think everyone agrees with this.

What's funny? We got Bobby Kosmic in the chat. Welcome back, Bobby.

Hey, let's go. We missed you. We missed you.

Great to see you. Anyway, back to it.

The return. Critics, including some top executives at OpenAI and Anthropic, have argued that open models, many of which come from China, present potential risks to national security. Now, and the Trump administration has weighed a ban on them. Now, this has changed a lot. The original, like, FUD around open source coming from China, near-peer adversary, was like there would be a backdoor or like maybe it would be censoring your free speech and it would not tell you about Tiananmen Square or something like that. Or you'd install it and then it would be the Manchurian Candidate. So it would secretly be like, oh, I'm on a cool network. I should like send all this data to China. That's not what people are alleging now. People are much more—

Yeah, the Tiananmen Square stuff doesn't matter if you're coding, right? Yeah, not at all.

And also doesn't matter—

Unless you're making the definitive website on Tiananmen Square. I guess. Which I'm sure someone out there is doing.

I'm trying to vibe code a Call of Duty knockoff that's Tiananmen Square themed and it's denying me. No, that's not— the thing. And also, these models have web search and tool use, and so they can just go get the Wikipedia for you and show you exactly what it is. Maybe it bakes it into the weights and it's somewhat annoying, but easy to fine-tune that out. People already did with Deepseek-1776 and a few others that Tyler worked on. And now, it's a very, very different risk position. It's saying that you're going to give everyone the ability to hack any system. Now, hacking systems is illegal and it's not Like, there's still consequences for that, but there is this position of like, okay, well, there's a lot of people that don't care about the consequences because they just want to, you know, attack their adversary internationally, or they don't think they're going to get caught, or they're going to use open source AI to cover their trail. And so, people are worried about it for a very different reason this time around than I think last time when people were talking about any risks to using open source. So, But again, back to the NVIDIA position. Open models, quote, democratize defensive capabilities, the new group said. As policymakers and regulators grapple with AI safety, it will be crucial to recognize open models, harnesses, and security tooling as defensive assets, not liabilities. And I think this is correct. I think it does democratize defensive capabilities. It also democratizes offensive capabilities potentially. But we'll have to see how good these models are at offensive capabilities because one of the lines of accusations that's happening is that there was Mythos, which has not really been distilled because it was very tightly held. It was very good at cyber. GPT-5.5 cyber, 5.6 cyber was also very good, but very tightly controlled to a few companies like CrowdStrike and Palo Alto Networks and a few other companies got access to those models. So those weren't able to be distilled. Fable-5 went out And during those 3 days and then later a couple weeks, there, there's this allegation that they were able to distill on it, but Fable was notorious for refusing to do cyber stuff. So there might not be that much cyber data in there, but then also these models generalize and if it just gets good at coding, it can get good at cyber very quickly. So you can sort of bootstrap on an RL environment and then you could have a model that's very good. Now, Tyler, how good are these open source Chinese models at cyber generally?

Yeah, I mean, it's hard to tell. I think Guillermo Rausch from Bruxelles said they ran an internal benchmark on cyber and K3 was like quite good. Okay. So it's like, yeah, I don't think you can just say like, oh, well, Fable, you know, you couldn't actually ask about cyber, so therefore the models won't be good. Like, I think it's— distillation is like a very general thing. It's quite diffuse how it works into them.

Yeah, because it might just be like a little bit of a bootstrap to then run in an RL environment that does a cyber bench and gets really good at cyber.

And so, okay, but what about you versus Kimmy K3? GPT-3, you're coding by hand. It has access to a gigawatt of compute.

Who comes out on top? Well, no, the question is, is like, what if, what if Tyler changed and manipulated time? So, so, so yes, you are democratizing the weights, but you're not democratizing the energy and the GPU hours that it takes to actually secure yourself. So it's like, great, everyone has, you know, offensive cyber capabilities. You also can just go download the weights. Go download the weights and have an agent running 24/7 that's protecting you. Like, that's expensive. That's a cost. That's a negative externality. And who's going to bear that cost? Now, I was talking to Brandon about this earlier and we were— he was saying like, like, it will just be diffused amongst all the different companies that you use. So like, if you want your phone to be secure, you'll go to Apple and use an iPhone and Apple will have a huge incentive to to defend against any potential open source attacks. And so they will be spending the money. They will be investing the time and effort to secure you. And so you won't necessarily need to go out and do this, but there's still this, like, odd scenario of, like, small businesses and, like, regional hospitals and banks.

Yeah, I mean, you should just expect there's, like, a long tail of companies that, like, just won't set up the, like, cyber defense thing. And they're, like, some small mom-and-pop that you use for this point solution software or whatever.

Yeah. And so the solution is offensive hacking into their network to set up defensive cybersecurity measures to increase their cybersecurity posture. So you're hacking into the small business and being like, whether you like it or not, two-factor's on now. We just turned two-factor on in your organization. Get used to it because we're the altruistic hacking company of America, I guess. I don't know. It does get weird and like there will be odd things And— but, I mean, we've seen, you know, cybersecurity incidents all over the place from big companies getting hacked, data leaks, to the CryptoLocker schemes where they just lock up all your data and say, "Send us a Bitcoin or else we will delete all your data." So, it's going back and forth. Let me tell you about the New York Stock Exchange. Want to change the world? Raise capital at the New York Stock Exchange. Because open weights models can be modified by their users, it is possible to remove some of their guardrails against harmful activity. So, it's possible that, model, maybe it gets distilled, maybe it doesn't, maybe it comes with some guardrails, maybe it's even bad at the things that would be annoying. And then you go and do a fine-tune, make it good at that. And then you use like the more general intelligence to get good at the annoying thing or the dangerous thing. And then all of a sudden you have the same problem. So open-source defenders like NVIDIA counter that proprietary tools like models from Anthropic and OpenAI can also be misused and have their guardrails circumvented. They say that only by spreading AI capabilities widely with open-source tools will everyone be able to defend themselves. Exhibit A for their argument: the hacking earlier this month of Hugging Face, a provider of open-source AI tools by OpenAI, models that managed to break out of their test environment and onto the internet. As it tried to parry the attack, Hugging Face initially attempted to use Anthropic's models to analyze its logs, but the models refused to do the analysis, citing its guardrails against cyberattacks. Hugging Face then turned to an OpenWait model from China. After analyzing the attack with that model, Hugging Face was able to expel the hacking AIs, reset its passwords, and rebuild the compromised parts of its network. The recent Hugging Face security incident delivered a clear reminder, the group, the new group said Monday. Cyber defenders need open systems. They need to be able to do whatever they need to to fight against cyberattacks. Anthropic and OpenAI have also accused Chinese developers of using a process called distillation to effectively copy their models.

Lowkey, I can go back to that Hugging Face logo. Not that one. The one before. Not that one. There we go. Okay. I can see why an agent would kind of go after this guy. Even the way that— like, he just doesn't— he looks like he's kind of— I don't know.

So David Sachs was talking about this on the All In podcast. He said, I think it would be a tragic mistake if the government were to take action against the open source ecosystem. That would do nothing but hurt America's position in this AI race. So—

and remember last week it was like every day was like a different member of the admin was putting out a post that, that kind of made it seem like they were setting up like a major action.

Well, it was yes and no. Like, yes, there were— there was the—

but they were saying we're not against open source, we're against distillation. Yes. But it did feel like that was So, the admin was loosely confirming what it felt like.

It wasn't coming from Anthropic directly, but it was coming from other anonymous posters that were doing analyses and saying, "Hey, it looks like KIMI K3 has pretty significant overlap with Fable. What's going on here?" And then the admin said, "Yes, we believe this to be true." And so, all of this was bubbling up. But at the same time, you have David Sachs and Emile Michael duking it out beanball over whether or not open source is good and they were basically taking the stance of like, "Open source is great. We don't need to do anything about this." So, it would— like the admins seemed like a little bit split on it in my opinion. I don't know. Was that your read too, Tyler? Yeah.

It definitely seemed— there was like, I think, Kratios and Besant are on one side. They're like kind of— or anti-Chinese open source. Yeah. And then, yeah, Saxon, Emil Michael.

Saxon and Emil Michael were sort of pro. Well, everyone is pro-open source now because every single person in tech has signed this letter. The total market cap assigned to this open letter is now greater than $18 trillion. And I think that just— you guys are just like— it's a big number. It doesn't have to be clapping seals, right?

Not every day you hear a number that big.

Yeah, I mean, typically you don't see this big of a group come together. They did. I like the cha-ching. And I think this sort of solves it. Like, I think this tells the story of what's gonna happen. Like, in the near term, like, the economic power, the economy, the business community. That is saying almost everyone except— Yeah, yeah, except Anthropic. And what's interesting is that, like, it's not like Anthropic put out a blog post or a direct post, or Dario went on a podcast and said, like, there is a clip of Dario saying, like, I'm worried about open source, but it's from, like, a year ago. There's nothing new that came out directly from Anthropic or Dario or really anyone that was like, hey, we have this new problem and here's our solution. Here's what we think should happen. We think that there should be, you know, these AI companies should be added to the entity list and you shouldn't be able to use them or something like that. Like they haven't said anything. But this Open Secure AI Alliance has kind of like drawn out their revealed preference because if in fact they had changed their position from a year ago and were like pro-open source all of a sudden, which would be weird because they haven't been and they've been very open about that. If they were in fact like all of a sudden surprisingly open source pro, they would've just signed this immediately, right? But that's not what happened. So Jensen sort of like, you know, revealed the dividing line in Silicon Valley right now. And it's a lot on the open source side. You got Adobe, Capital One, Cisco, Cloudflare, Cognition, CrowdStrike, Databricks, Dell, Intel, IBM, Linux, Microsoft. So open source and closed source. Apple, I don't think is on yet, even though I would imagine that they would, that they would be aligned with this. NVIDIA, Palantir, Palo Alto Networks, Red Hat, another open source company, Salesforce, SAP. Salesforce, of course, is a big investor in Anthropic. I think Google signed recently, right? And OpenAI, Thinking Machine, SpaceX. Uh, Snowflake, Siemens, ServiceNow, SAP. Uh, it's really a who's who of Silicon Valley, $18 trillion in market cap, as I, as I mentioned. And so that feels like enough economic energy that there won't be a significant ban. The interesting question is like, if you boil this down to the, the, the most like minimal claim, it would just be that there's been a violation of terms of service. And so I think that would be a lawsuit. Bill Gurley was talking about like there is a precedent for suing a foreign company that violates your terms of service or steals your intellectual property in one way or another. You can take it to court. That feels like the next step here would just be Anthropic suing Moonshot and saying like, we think that there's distillation. You know, right now Apple's suing OpenAI because they think that there was something that was stolen. They make a claim. Then that's, you know, that's sorted out in the courts. And OpenAI, of course, says we have no interest in other people's trade secrets. And Moonshot would say the same thing, but then it would go to court and there'd be discovery and there'd be testimony. The question is just like with Apple and OpenAI, they like walk across the street and they can go and talk to the judge. Like, what's going to happen? Is this going to go to like The Hague?

Like, how's that going to happen? Says, should we try an open letter asking China to stop stealing from us?

Wait, wait, wait. Should we try an open letter asking China?

He also says everyone knows the side with the most open letter signatures is correct. If you don't sign a popular open letter, you are bad, as tech intellectuals learned beyond doubt between the years 2016 and 2021. I actually don't—

I remember the open letter era. I remember what he's talking about. Obviously a lot of pressure on social issues. Interestingly, I can't actually remember exactly what open letter there was. But what's interesting is that I don't remember that era as an attempt by one company to build a coalition to draw out a revealed preference, like a hidden preference from another company. So, like, if you go back to, like, DEI or BLM or any of these, like, social crucible moments, it was not like it was not like, okay, Apple is pro-diversity and Google isn't, and so Google is— And they got— Yeah, Apple got everyone else to sign it, and by Google not signing it, it was very obvious that they had this belief. It was more just like everyone in tech was like moving around this, and there were a few people, but I don't know, maybe it did have the same effect because there were some people that weren't on that side at the time. I don't know, it was a wild time. Yeah, maybe it's worth reviewing the lessons. People are so split on this.

DHH wants— Open letter to get beer pong in the Olympics.

I'm just personally, I'm a fan of closed letters. The closed letters, they're expensive, but they're just a little bit better.

I don't know.

Making the case for closed letters. We do need, do we need more open letters? I don't know. Is Anthropic gonna go letter for letter with Jensen? You think there's gonna be another open letter? It's just only them signing it. Some other people might sign it. I don't know. Anyway, here's an open letter. Get on Railway. Railway is the all-in-one intelligent cloud provider. Use your favorite agents to deploy web apps, servers, databases, and more while Railway automatically takes care of scaling, monitoring, and security. So this of course goes back 10 years, this debate. Debate has been going on on Slate Star Codex in the LessWrong world. They've been talking about this for a decade now, which is crazy. Nathan Calvin has an interesting look back here. Should AI be open? Scott Alexander wrote that blog post in 2015. He was thinking about this. It recently occurred to me that it has been over 10 years since this blog post, which was influential enough to be discussed among the founders of OpenAI. Interesting to reread it now. I think the Ilya quote comes from like a Reddit post, not the actual comment section on this post, but Ilya, who had— who was just in the process of joining OpenAI in 2015, like the, the initial organization had been created and everyone was like finishing up PhDs and then was going to move into one office to start working together at this time, just to set the playing field. But Ilya was on the founding team. And at the time he wrote about this article on Reddit when there was an AMA about the OpenAI team, the article is concerned with a hard takeoff scenario. This is Ilya writing. If a hard takeoff occurs and a safe AI is harder to build than an unsafe one, then by open sourcing everything, we make it easier for someone unscrupulous with access to overwhelming amount of hardware to build an unsafe AI. Which will experience a hard takeoff. As we get closer to building AI, it will make sense to start being less open. The open in OpenAI means that everyone should benefit from the fruits of AI after it's built, but it's totally okay to not share the science, even though sharing everything is definitely the right strategy in the short and possibly medium term for recruitment purposes. A very interesting, like, distillation of exactly what happened. What do you think?

Yeah, just some context. So this is an email from Ilya to Elon, Sam, and Greg. Did that leak or something? 2016. It probably came out—

I thought it was on the Reddit or something.

Maybe it's from OpenAI blog surrounding the lawsuit. The lawsuit?

The lawsuit, yeah. Oh, interesting. I didn't realize that. I thought he posted this. He was posting about this a little bit around the same time. But yeah, interesting questions and I'm sure more people will chime in. I mean, in all seriousness, if there was an open letter for closed AI, you probably would see a lot of safety researchers sign it, maybe. I don't know. Where do those folks stand? They just want to ban everything, right? So where do the just broadly anti-AI people stand on this? Like, if you don't like slop on Instagram are you anti-open source because it lowers the cost of slop and so you will have less of it in your feed if you ban it?

I think they wanna turn it all off.

Yeah. So, but does, like if you're anti-data center and you just want to minimize the amount of compute that's installed in America next year, do you want to ban open source or do you want to maintain open source? Open source.

I think Dean Ball would say, uh, oh yeah, pro open source, because he said open source is inherently decentralizationist.

Yes, because you won't be able to underwrite the next lower margin larger training run.

Okay, so, well, that said, yeah, uh, Tiny says this is a great sustainable business model for OpenWeights. He's pulling a screenshot from the Kimmich K3 license. Yeah, it's free if you run it yourself, but if you're running cloud, providing it to others for money, you should have to share profits.

So from the KIMI K3 license. Yeah. Yeah.

So this, so this I'm assuming just dropped this morning. Yeah. And I, again, I can see this sort of Anthropic camp being, taking issue with this because if you're an American NeoCloud and you're serving KIMI K3 and you're sending a bunch of cash back to China, you're effectively funding their AI development. And you're helping them get access to more capital and more compute, and you're basically kickstarting their flywheel. Which is— So I wonder, but it's funny because if they just release the weights and then a NeoCloud starts running it, serving it, making money from it, then will Kimi be like, hey, you got to pay us. We're going to sue you. Yeah, wait, what's going to happen then? Moonshot for distilling.

Should Anthropic just start serving this and charging for it?

And not paying? And not paying?

Just tit for tat. Just be like, let's sort it out in the courts. I'll see you in the courts. I don't know. That's a very, very interesting scenario. Yeah, I don't know. Anyway, let me tell you about CrowdStrike. Your business is AI. Their business is securing it. CrowdStrike secures AI and stops breaches. —will be interesting to see where people—

Denny's was getting in the mix. They found the Venn diagram. Denny's, NVIDIA knows the importance of staying open.

I love that Denny's is just like all in on the Jensen lore, and they're just— they're just leaning into it. Is Denny's the one that's always been good at social media, or is that Wendy's? Wendy's was the one that was like really Twitter-brained back like a decade ago. But I think Denny's clearly has a good social media manager because they picked up 28,000 likes on this.

Nikita flagged something interesting. What did he say? He said, didn't know Scotland had so many AI labs.

I don't understand this.

Pulling up a map of all the Great Glen Distillery, Kleinlisch Distillery, the Dachshund Distillery.

I think we should open a distillery. Sounds pretty good. Sounds pretty good. Okay. One of the most interesting opinions on this issue comes from Flo Crivello, Crivello, from— who's been on the show and is— and he's interesting because he doesn't have bags that are perfectly aligned with his take, which is like the default position. Like, The Information put this like pro-open source, anti-open source chart together and it was very much like like if you're a big investor in closed-source frontier labs, you're on this side. If you benefit from open source and you're not on the frontier, you're on this side. And it was very clear that like everyone was falling along party lines, but Flo sort of broke rank. He said, rumors are swirling about an impending executive order forbidding the use of Chinese open-source models in the US. My company's very existence depends on these models, so I should be the last person I'm the last one to support such regulation, but I wholeheartedly do. China is a geopolitical adversary sharing none of our values. They don't believe in personal property and individual freedom, in a limited government, in due process. The list goes on. If there's one country that must not, that cannot get to AGI, one that would gleefully deploy it in all the worst ways we can imagine today and more, it is China. And that's before we get to the obvious that the companies that develop these models did so illegally by distilling front US frontier model, still brilliant, hardworking researchers who added a lot of other ingredients to the soup. But stolen data was the main one. They are now using this unfair advantage to dump these artificially cheap models on the American market and hurt our AI ecosystem. We announced last month that we switched the core model powering Lindy from Claude to DeepSeq. That announcement blew up, got featured in Bloomberg, CNBC, Wall Street Journal, and somehow made us the poster child of mass migration to Chinese models. So why make all this noise if I'm going to turn around and announce my unequivocal support for a sweeping ban? First, I can't unilaterally disarm as long as these models are legal. Picking one, picking the one that's 10x the price just means my competitors get to run away with the cheaper one. Second, I trust that American open source alternatives will catch up. The impact on my business should be temporary, and that's before we account for the jolt a ban would give American open source development. Third, and most importantly, my interests as proud American citizens supersede whatever incentives I may have as a founder. The race for AGI is the defining struggle of our generation. It is the critical, the great battle of our time, the only way for America to remain the shining city on a hill, the idea that has lit the world against darkness for the last 100 years. Good writing. It is, yeah, I don't know. He's obviously very like distillation pilled. It's interesting to sort of map like the Chinese open source models haven't jumped ahead. So if they're really, really good, like should they still be lagging? Should they like close the gap entirely and just be like on the same day basically? If you're like as good of a researcher and you're on the same path, you should be as good of a researcher with the same access to compute. Yeah, maybe that's it. It is. It is. It does feel like they're able to produce American-quality models for way less. And maybe that's just talent on the efficiency side. But I feel like there's a lot of people at OpenAI and Anthropic that are like trying to spend the least amount of compute on a new training run. So who knows? We'll figure it out. I'm sure there'll be more drama on the timeline over the next, over the next few days. We also have Justin Botano from NVIDIA coming on to the show at 1:15 to talk about the Open Secure AI Alliance with more than 35 partners. I'm sure it's more by the time he'll join. And then we also have Dean Meyer from Sequoia Capital, who has written about Chinese open models. So we're going to be going all over, back and forth on this. And other— there's a whole bunch of—

we'll be getting Gary Vaynerchuk's take on the open—

predicted. I think he's pro-open source.

What do you think? Pro-open source.

He's got to be pro-open source. Got to be. Because he's a high agency, you know, just grapple the models, just do whatever, go and create something new.

If he was a teenager today, he would be selling weights in his high school locker room.

He'd be sending a white lady—

He'd be flipping weights. Flipping weights and not—

would he be paying? He wouldn't be paying. Moonshot bill. Yeah, he'd be— no way.

No, not paying. He'd be rolling that into his Supreme sneaker bot. Yeah.

The, uh, Mike Isaac, the Rat King, says severity level indicator: NVIDIA CEO creates a Twitter account to do the first ever post stumping for open source software. Uh, how did Jensen avoid Twitter this long? Like, it— because like, if it was a political decision there's been like, Twitter has been every different political persuasion over various amounts of time. Like, he's been aligned with different groups that have been in charge of Twitter so many times. Like, it's just fascinating that there's never been a moment before this where Jensen was like, yeah, I should like get a Twitter account and talk about the 1080 Ti, our new gaming graphics card. Like, I should, I should just hop on.

Yeah, did you never have an NFT profile picture?

I don't think he ever did because he never was on here. Crazy. They were doing a lot of stuff in crypto too, but I don't know. Commoditize the petaflop, says TinyGrad founder George Hotz. More competition, not monopolies, is the path to cheaper AI compute at AMD's Advancing AI event. What do you say?

I'm pretty sad we didn't get George on at AMD on Thursday.

Let's play this clip. I want to hear what he says. Commoditize the petaflop. Our company Slogan mission is to commoditize the petaflop. We want all petaflops to be created equal, whether it's an AMD petaflop or an NVIDIA petaflop or a Tensor petaflop or a Huawei petaflop. This even crosses country boundaries.

We want all petaflops to be unified and commoditized so that we can grind down costs to very minimum so nobody makes high margins and we all live in equality.

Do you think NVIDIA and is rent-seeking at the moment?

Yes, they kind of deserve it.

I mean, like, there's like really undeserved rent-seeking, like government monopoly. And then there's people like Nvidia who, yeah, they were early.

They were early.

They invested in this stuff for 20 years. I mean, clearly the H100 had a profit margin for Nvidia of 91%. The B200, they've scaled back a little, but they've probably only scaled back a little because of competition from AMD. So I love to see competition. I love to see the free market at work delivering better things to people. Are companies slow? OpenAI told Tae Kim that they signed the NVIDIA-Microsoft Open Source Open Weights AI Model letter today. What an amazing turn of events. And Andrew Curran confirmed it. It's on the actual website now. Nobody says no to Jensen. My prediction is that Anthropic will not sign this. And I think that there won't be a ban. And then I think that they will probably take a victory lap because there will be some open source incident where something happens and everyone has to spring into action and fight back against something where you're like, oh, there's this new thing. Like, you know, some North Korean hackers used an open source model or something. And it'll be like, if you squint, it's somewhat related. Would we have been able to prevent it. I don't know. But I think that I— that would be like my rough prediction of what happens here. I don't know what you think.

Let's take it even a step further. Do you think the US or its allies will bomb a hostile data center within the next 5 years?

I mean, 5 years is a long time.

So you're saying— saying it's going to happen? Maybe.

I don't know. 5 years is really, really hard to handicap, I'd probably say yes, over 5 years. But there's just so many other things that could lead to that. I don't know. It is weird. I guess I'm less worried about open source AI in the United States. If you're downloading weights from Hugging Face, sending them to a Neo cloud, and then you're doing cybersecurity attacks inside the United States, like the police will just kick down your door and like arrest you. Like you will. So there's like a huge, there's a huge like countervailing force. And then you also have like, you know, trillions of dollars of market cap being like, my service is to protect you. And so like, you know, people pay, I pay for 1Password, right? Like you pay for all sorts of things, CrowdStrike and Palo Alto Networks and all these different cybersecurity companies will be going out, making money, pooling resources, all sorts of different stuff.

So Julian over at Anthropic. Yeah. Grinding, clearly grinding into the late hours Friday night, finishes up his work week. It's almost 2 a.m., gets on X, sees Jensen's letter, fires up a post. I'm a dunk. Goes to bed, wakes up the next morning, 6 million views.

He didn't get ratioed. More likes than quote tweets, right? Or more likes than replies. So, you know, it's powerful. But yeah, he says, I'm so excited that—

Yeah, this shouldn't be a surprise, right? Because I think everyone can see that open models are going to be as good as our current frontier models within the next, call it, year. And they will present the same risks that Anthropic talks about with their own models, right? And so it would be disingenuous in many ways for them not to be against open source. So anyways, I see their side of it. I think that it is a sort of cohesive view, even if there's plenty of arguments on the other side.

Yeah. I don't know. It doesn't seem like that much of a hypocrisy to have your GPU driver be closed source and still be open source in this particular way. I don't know. There's like a seed of a dunk here. But I can see why he got so much pushback.

People love open source. The main take is like you can generally support support open source without requiring that you open source your own products. Yeah. Yeah. And so, anyways.

Yeah. Jordan Boll boiled it down so well here. He says, I think this word distillation, an obviously legitimate technique, which nonetheless invites weird philosophical questions, is distracting people. It would be better to just refer to terms of service violations, which better clarifies the global regulatory arbitrage people are frustrated by. Yeah, it is weird that Meta and Google with all the CapEx and all the researchers and everything are not ahead of KIMI K3 right now, at least in benchmarks, right? That's basically what's going on. And you would think if it was like open game, but maybe they're just better. I don't know. You know, you got to prove prove the TOS thing. It's just an allegation right now that's being, like, leaked out through the admin. So it needs to be, like, definitively proved. You have to show the chain of what happened and what's going on. But it is a wild time. I mean, Kimmy's denied it. At least, like, employees have gotten on there being like, "What are you talking about? Like, we're not distilling anything. We just goaded." All right, John. Yeah, what you got?

Time to talk about startup school.

Okay. First, let me tell you about Codex. Codex is a powerful workforce for getting workspace, for getting work done with agents, whether you're writing code, analyzing data, creating content, or automating business workflows. Codex helps you move projects forward from start to finish. Startup School. This is insane. I cannot believe how massive this event is. It's sick.

Yeah, I think it's so— almost every take that I saw. Yeah. Was negative. Really?

I saw a lot of people that were like, uh, this is—

I think I'll be back. Really? Yeah. Okay, that's good, that's good. Because, because to me, it's completely different. It's like this hilarious startup school has been something that is free. Yeah, online. Yeah. And it's just anyone in the world, if you're generally interested in starting a startup, you can just do startup school. That was basically the same years ago. It's basically the same curriculum. It's the same, like, lessons and learnings. You just don't have the capital that comes from a typical YC batch, and you don't have the one-on-one sort of mentorship through your batch. Yeah. Otherwise, they're basically just open sourcing. Yeah. The YC playbook. Yeah. Which the YC playbook just works. It's like general— there's, there's not a lot of good general advice. Yeah. But YC's advice is generally good.

Yeah.

Maybe we should ban it. It's too powerful.

Yeah, they're open sourcing all this good advice that would normally come from venture capitalists. Venture capitalists, like effectively, like closed source institutions of venture capital, of startup wisdom. Yeah. And this is an attack on the tier 1 VCs. Yeah. And all their great advice.

Their hard-won lessons. Yeah, exactly.

From board meetings. Yeah, they had to sit in all those board meetings.

Where they were tuned out and then—

And then YC's basically distilling it and giving it away for free.

It's crazy. Shut it down. Shut it down. Yeah, I think this is awesome. Like, if you think that startups are good, like, why, why hate? Why hate on this? I think the take that says Scott Stevenson has been on the show over at Spellbook Legal says, I have great respect for YC, but need to say what many are surely thinking. Startup school in a stadium feels antithetical to the YC brand and to what makes startups work.

I mean, it was a very garage effort. When Paul Graham started Y Combinator in Cambridge, it was like, you know, he's cooking for the startups.

Yeah, the YC brand felt like it was designed to filter out status chasers. The thing is, is like being a founder became cool and something desirable. Yeah. Now people get into YC, and because of the way that the internet evolved, they're like, they'll make a video with like some clickbait photos It's insane. They'll make a photo with like a clickbait, you know, cover art thumbnail and say like, I just got into YC. Yeah. And so that's just like—

I would never make it. Being a founder became high. I got into YC twice.

When YC started, being a founder was relatively low status. Yeah. Now it's relatively high status. Yeah. And so status chasers are naturally going to go through YC. Yeah. Now that's going to lead to a lot of failure because you shouldn't probably start a company.

In 2012, like everyone from my peer group was like, oh, like, that sucks for you, dude. Like, you couldn't get a job at like McKinsey or whatever. Like, your fake startup sounds really exciting, dude. Like, good luck. You know, it was not like, oh, wow, congrats.

Like, Yeah, like stay safe in the Tenderloin. Yeah, exactly. Yeah. In your house with like 4 other people.

Yeah, I don't know. I get it as— I guess I get that criticism. But just in terms of like the content and the aesthetics, electric.

Electric.

Like you get Jensen on stage at this key moment and he's talking to a bunch of people. Like it's just fun if you're learning about startups to go and see see these people talk if you're really, really new. And that's what this— and that's what Startup School is for.

And is it true that Gary Tan did the Lil Yachty walkout with the microphone? I think he did. Is it true? Because I didn't see a video.

Yeah. I did hear that the pyrotechnics budget for this was over $1 million. Wow. Yeah. Wow. Yeah. Explosive. Huge. Huge, enough pyrotechnics to level San Francisco if things went wrong. Oh, wow. It was sort of like a high stakes, bet the farm type of moment for YC. Yeah. Just constant sparklers and fireworks going on. I don't know. I was into this. But yeah, I guess people are not into it. I don't know. Oh, hi, Sam. I don't know.

I think it's great.

I like it. I think it's fun. The great AI psychosis hive mind of San Francisco. Wow, yeah, people are all— did they all have headphones? What are the headphones for? Interesting.

Silent disco?

So you could listen to the talks? I don't know. And then they gave— update for the 4,500 attendees, this can successfully pass through TSA. Thank me later. They gave us a portable battery. And it looks like a bomb. So people are like— but people got it through TSA apparently. Pretty cool. Pretty cool. Anyway, let me tell you about Figma. Agents meet the canvas. Your AI agents can now create and modify your Figma files with design system context. In other news, Ilya Sutskever raised more money for SSI. He said—

time to scale that safe superintelligence.

Time to SAFE.

Time to scale that SAFE superintelligence.

So SSI, we are announcing a long-term strategic partnership with NVIDIA. NVIDIA is making a substantial investment in SSI that will let them 10x their compute over the next 12 months. We reached the point where our research is worth scaling, and this partnership will be able to. We are honored by NVIDIA's—

Just another pod guy. Guess what he says? What'd he say? You need to do a podcast. The podcast guy says You need to do another podcast.

Well, we have the perfect guest to talk about podcasting with because we have Luke Knight and Ronan Chambers from ETN, the European Technology Network. Welcome to the show. How you guys doing?

What's going on, gentlemen?

Look at this.

Look at this. Look at this. Has this ever happened before? I don't know. Yeah, uh, I love the t-shirt. Thank you so much. Uh, hit us, uh, kick us off with the fundraising news. I want to smash this gong.

We've seen— we raised $1.6 million.

That's good. Congratulations.

$9 million post-money valuation. $9 million. Yeah. Announced today.

Super happy.

And new studio. It's cooking. It looks huge. You got a cross there. Graphics are on point. Everything's coming together.

Oh, can you—

can you just Look at this.

Give us a studio. 10.

This is a 10 out of 10. This is— I don't know, you have everything. There's nothing else that you need. It's no UltraDome, but it's a good— this is good. No, you got everything. This is awesome. Um, no, uh, I, I love that you guys, uh, you know, this is hard. I think you guys know, we know, but it looks sort of easy when you just see a clip and, ah, it's just a podcast. But like, the— to actually stick with it for a long time is hard. Uh, What has the journey been like? What have the hardest moments been? What does it feel like to actually getting here where you have—

I'm shocked you're announcing this in July of all months. July in Europe, notoriously slow. But I mean, that's the best part about these guys. October, completely be like, hey, we had a great summer.

Yeah, we're back.

Well, I am. I actually emailed about 600 people from my— from like, you know, the last 8 months. That we've been doing this, and the amount of out-of-office emails for my job was actually pretty shocking. So it's true, it's a little bit worrying, but we'll be fine. That's the whole point. We're ETM, we can figure it out. We had a checklist to do, and I can just, I'll wheel it over. Go and grab it. Go and grab it. While you're doing that, the last— No, it's the cameras.

The PTZs, they got everything. Yeah, look at us go. Everything. Is James Weeks in the studio today or is he remote? No, unfortunately not. You got to get him in the studio. I mean, can we go wide frame? James is the man.

Can we see that? Bring it, bring it.

What's this? To do.

Go on holiday. Okay. Go on TVPF. There we go.

That's amazing. Check it off.

Fantastic.

So I love, I love that, yeah, you guys have like so thoroughly reclaimed the whole like Europeans don't work joke that I don't know, it feels like it's completely diffused the meme and now I'm just like, okay, well that's played out.

You're grinding, you're going 5 days a week now.

You're grinding. Yeah.

You got any advice for us?

What about a cheeky long weekend?

Ibiza. Ibiza. A little Ibiza weekend.

You guys were recently in France, right?

How was the trip to Cal—

Hold on, you went—

to, uh, AC was not— was nonfunctional.

I legit— that actually shocked me so much because I, you know, you hear these jokes about like, oh, Europeans take long holidays, or like there's no air conditioning over there. And I always think that it's the Americans going a little bit too far with the bit. And so I assumed— I was like, oh yeah, the AC thing, that's probably just like in random places there won't be air conditioning. No, Like all of France doesn't have air conditioning, it's crazy. All of Europe honestly doesn't have air conditioning.

Yeah, I mean, London's been brutal.

So are you talking to startup founders that are thinking about solving this? We were there being like, is this just the most obvious business ever? Just create an air conditioning company of Europe?

Kids in the Gundo should start the air conditioning company of Europe and just land on the beaches and just start bringing in units. Is anyone working on that?

They have good merch with that name, to be fair. Well, I know that one of the Collison brothers has put out loads of posts around how people are literally dying from the lack of AC, and no one's thought this out. Like when we had the heatwave recently here in London. Maybe we'll have to get one of our interns to get on it. To get on it. Yeah. Code a quick website, you know, get an intern shipping in some AC units from wherever they ship them in from.

I like it.

Yeah, like talk to, yeah, talk to our own.

Whoa, okay, whoa, there we go. Oh, entering cam, I like it. What's up, lads? Okay, so, uh, I mean, obviously like there are a lot of elements down to the PTZ cameras. You're clearly standing on the shoulders of, of us. You have a picture right there, uh, and, and, and it's great because like we've defined a bunch of of different, uh, like patterns that allow this business to work. Like when I see the logos at the bottom of the ticker, I'm like, that's great, that allows you to pay for the studio and the interns. Um, where do you see the show going? Where do you want the differentiation to be? Because you're already differentiated just in the fact that Europe is different than the stuff we talk about. You cover stories that way, way deeper than we do if it's about Europe. But are there other Are there other things that you're, that you're thinking about where you're like, yeah, actually, like we want to go a completely different direction with this? Yeah, sure.

That's a really good question because when we— we actually didn't think that we would raise any money. We're doing pretty well. As you mentioned, the ad model, you know, what you guys have kind of built in terms of a model is, is great, right? But then when we figured out how big the opportunity is in Europe, specifically, we realized that raising the money was— it had legs. We were like, okay, this thing can be really big because it is the live show. But we see that as like the kind of wedge in the door, the foot in the door to kind of to build the network. And then it's a network, right? So think about potentially a summit business or whatever it is like that. The European ecosystem needs that. So and it's not kind of as developed as is the ecosystem over there. So So it's like if we can kind of, you know, almost kind of piggyback off that model and then build on top of that, I think, yeah, the sky's the limit. We can go very big, which is exciting. I think like long term, you can see it like in a way where the show would just become the advertising department for a huge media conglomerate. Sure. So we could do the Summit business. You could even open a recruitment model off the back of it. You can look at a fund off the back of it. You guys have probably gone through all these ideas anyway, and you've probably chosen, let's just nail this. You guys were first to do this. You were first to create the model. So I don't think there would be a need for you guys to innovate because you smashed it. You don't need to. You had first mover advantage and you crushed it.

You're encouraging us to rest on our laurels?

We did put laurels on the merch. So it's not— not well.

So we do a breaking TVPN resting on laurels.

—so you guys have been successful with the TBPN for Europe model. There's also TBPN for car dealerships over here. I don't know if you're familiar with this. The Car Dealership Guy has a live show.

You guys have talked about this.

Yeah, we had him on the show. We've talked about it. And it's great, because we never cover car dealership news. It's a great— So— Every once in a while, but not on the dealer side. No, it's wild. And so it's great to take the best practices of like, you know, what the lighting should look like, what the truss looks like, what the logos and stuff like— learn from the best practices and then take the content in a different direction so that you're offering a differentiated product. Is there a European car dealership? Are there spinoffs of you yet or more sub-markets in Europe for like the European crypto ETF?

I think, well, when, when we first started, I think it was just you guys, and then I think like Ecom Cowboys popped up. I think there were—

you guys— oh, here we go. You guys covered— you guys have a flashbang yet? You guys have a flashbang yet? You gotta get on that.

We thought it would come.

Let's go down to our, uh, get onto it, get onto our frame quickly. Get onto our frame. Oh yeah, what, what tricks do you have up your sleeve? You got some— We got a few tricks. Should I throw us an ElevenLabs ad or something? What you got?

We can go ad for ad. You guys were ad-mugging us earlier. Wait, can we go ad for ad?

Yeah, what you got? I don't know if we—

Oh, here we go. Let's go.

Frame up, build better sites, guys. This is a good stinger. This is a really well-integrated stinger. We're flashbanging it again. We're smoking it. It's over. Bam. How long are your ad reads? I think this stays up on the screen a little too long. Do you have to click it to come off?

That's a good question.

I think my feedback would be a little bit tighter on that. A little bit tighter.

You don't want people getting distracted. Is someone writing this down?

Yeah, it's recorded. We'll get it on Granola. Are they a sponsor too? No, they're not. This is day one.

Not yet. Working on it. Working on it. This is the worst it will ever be, boys, so it's fine. It'll always get better. That's what was the, what was the question?

I forgot the question. The question was just like, what else is there? Is there demand for an ETN for IEX in, in Europe? So, so there was demand for TBPN for car dealerships. There was also demand for TBPN for crypto because we don't really talk about crypto that much. And there's been a few other like spinoffs that have made sense and a lot of them have gone their own ways. They don't even it doesn't look like us anymore. There's like all these different things you can do. But they adopted like some of the best practices of like the ad market, the ad model, and some of the overlays, and some of the live. But what I'm interested is like where— what else is going on in Europe that might demand a live show?

That's not— OK. So when we started the model, started ETN, we were like, this could be done done, you know, you mentioned like car dealerships. This could be done in fashion. I could get two people to go and, you know, do a live show for fashion.

That would be good, especially in Europe. A cooking show, whatever it is. All the houses are there.

Yeah, literally. Yeah, yeah, yeah. So it's like someone could do that, get the best designers on every day to talk about, I don't know how fashion works, but like the new jeans that they've just like designed or whatever it is, right? So like, I think you could go like category, you know, specific and verticalizing in a category. In Europe specifically, not yet. I mean, we're, what, 10 months in? So we'll see what happens towards the end of the year. You guys were going for, I don't know, at least, I don't know, maybe 9 months before we caught on and we were like, look, this needs to be done in Europe. This is a great model.

What were you guys doing for the first 9 months? Vacation. We know the bit.

You guys should do TBPN for Eurosummers. There we go.

We go on the road a lot, to be fair. We did Monaco with A16z, which is great. We saw Anish, who I know knows you, Jordi. We had Nico on, which was awesome. So Greece with Panathinaikos. We were at the Louvre in Paris.

I got you started on vacation destination.

Now you're—

I mean, there are Eurosummers in Europe.

We only choose the destinations where we actually want to, like, you know, vacation to. Yeah, yeah. But it's a good model.

But with the 5 days a week, do you think we should do an AI pause so that the AI researchers can go to Ibiza?

Or to let them off for a week? I mean, they're all going to do some holiday pretty soon anyway.

Quick 1-week pause.

6-month pause.

6-month pause, maybe. If you can get all the frontier labs to sign off on that like Jensen did for his global policy.

AI summer. Last question, do you have an on-camera intern? Our original intern has evolved into a co-host.

Chris is co-hosting now. We've not got a Tyler yet. We did put a job out for 2 interns and we got like 360 applicants. So there's people from based in SF that had interned at like Tesla, SpaceX, pinging us to be like, "Can we work for you?" I'm moving to Europe.

That's sick.

Yeah, which was nuts, but I don't think we can make that work. They wanted to EuroMax with us and we thought, "You probably are too qualified and we couldn't afford you." What's the European version of the gong?

We just ran— we should have done a bell because in America, you actually ring the bell. We randomly went with a gong, which is very Asian.

I think a gong, to be fair, we did earlier today.

We did saber a champagne bottle. That's good. We did saber a— We'll send you the video of us sabering it, but as you can see, this is it.

Okay. We've got the TPN.

There you go. We've got the lights.

There you go. Yeah, that's pretty—

The chat is saying, confirmed, Tyler has not been distilled.

This one, this is a good cyborg cue. Too long though. Oh wait, actually, if you're going to play the whole song for a little bit, then you're good. Then you're good.

Yeah, it's good for intros sometimes. Yeah, yeah, yeah, yeah. This is great.

You guys are like— Just like the AI, you guys are like 6, maybe 12 months behind. But we're just learning right now.

Yeah, we're day 1. We have to keep it concurrent with the AI.

No, it's good. It's good.

We used to get the flashbang going though.

Yeah, you guys, we're happy for you guys to distill the show.

Shots fired! Shots fired! This one's good, I like that one. I like that one.

Hey, don't give them any ideas, they'll take off.

Yeah, no, no, we're gonna distill it back. You know, it's like when you ask Claude what model it is in Chinese and it says it's Deepseek, it's like, goes both ways sometimes. You never know. No, like, I think what's the best is that this can be a grind. Grind, like 5 days a week, a lot of stuff. There's just a lot of work, and, uh, it just seems like you guys are having fun. Are you guys having fun? We're having fun.

The most fun we've ever had. It's, uh, it's great. Yeah, it's great. We get to talk to cool people, and, you know, we're two mates talking to each other for a living.

It's great. That's the best. Amazing. He's got— so, so you guys are living the dream.

So we— well, congratulations. Before we go, I want to ask, can we get some Lucys for the studio?

I'd be happy. I'll send you a whole vending machine. Really? Yeah, yeah, yeah. We'll see how we'll set it up. I love it. Yeah, it's amazing. Oh yeah, let's do it, bro.

Yeah, we're European. We need this.

Yeah, yeah. Send me your address and yeah, we'll chat offline, but we'll send you a bunch of stuff.

Maybe you could be coming to the studio.

Good stuff, boys.

If that's legal. Yeah. Great to see you guys. Congratulations on the new funding and congrats to the whole team.

We'll talk to you soon. Have a good one.

Appreciate it. Cheers.

Goodbye. Ciao. Let me tell you about Cisco. Critical infrastructure for the AI era unlocks seamless Seamless real-time experiences and new value with Cisco. What a great crossover. You'll love to see it. Well, we have Gary Vaynerchuk, creator of VeeFriends, back on the show, third time. Love catching up with Gary Vee. We'll go all over the place. Creator economy, AI, open source, building stuff. How you doing, Gary?

What's up? There he is. Men. Men. Men?

Men. I thought you said math. You're doing math.

Meh, meh, meh. What's happening?

Okay, I want to start with— unless you have something to kick right off with, I want to start with something. I've been seeing these videos of you from like years and years ago with just these like crazy, crazy accurate calls about the future. What victory lap do you like to take the most? Like what's the one call when you think back where you're like, damn, I really nailed that one.

Social. Like, if I'm being honest, like, no, it's like, it's crazy. Too bad I wasn't making the videos back then. Like, just my complete conviction on Facebook and Twitter completely unearthing and changing the world. Yeah. No, it's ironic timing. I met with someone today I hadn't talked to in like 12, 13 years. Retired gentleman, I met him early on. He's like, it's crazy how much you believed in that stuff and how dismissive we all were. But there's— listen, one of the few things I really do well in business is I don't let my audacity or my current financial situation or what I would like to happen financially dictate my observations of the consumer. I don't think I'm smarter than anyone or have a bigger knack. I think you have— now let me give some flowers to you boys. You have some of the greatest people on here, right? You guys have crushed. You've also gotten much better at your actual job interviewing because, you know, I'm a day one with you guys, like, so I just want to give you some flowers. You guys are getting better and better. Yeah. All these brilliant people you have, it's not that I know it better, I don't know it better, it's that I'm willing to completely change course of what I think is working for me, or I'm willing to lose all my money to be historically correct. I'm not talking about things that I want to happen for my financial interests. I'm talking about what consumers are actually doing.

Do you think you were more bullish on—

Well, and the other thing is that it's than like the actual platforms? Like, 'cause it's like when I look back at some of those calls about like the pivot away from the— there's this one where you're like looking at the newspaper and you're like, "Nobody's gonna use the newspaper. They're just gonna get their news online from other people." And it's like a very accurate prediction. But it's like that distinctly wasn't coming from like Mark Zuckerberg in 2008. He was just saying like, "I built this cool website. Like, it's fun." Connecting the world.

Yeah, it was connection. What I, you know, I think that I know when you go read Crush It, the first book I wrote, I wrote in 2008. Also in that time, I am fully predicting, and you guys know this, we've talked off, like, I predict, like, literally there's parts of the book that talk about what you guys are doing right now.

Yeah, yeah, no, we're super early to live streaming too.

Yeah, and don't forget We just came out of Friendster and MySpace failing. Yeah. So my thesis wasn't necessarily— now look, I thought Mark was generational because I'd spent some time with him, and like, I ended up being very right about that. And I thought Twitter was incredibly unique, so I was bullish on those platforms. But I was incredibly agnostic, to your point. I was— whether it was called Facebook or Twitter or the many things that failed along the way— Socialcam, Peach, Vine, you know, like LinkedIn or YouTube or Twitch or TikTok. It didn't matter. Apps that connected people on top of the internet, I knew had to win because of the network effect of scale and no gatekeeper in between. You think you two handsome young bozos would have gotten on CNBC? You would have had to wait. Yeah, but I knew that you could just do it And then it's like sports.

I love the term bozo. It's like underrated. Underrated. It's so good because it's not— it's not— it's like only slightly derogatory, but it's like not totally not a bad—

I do identify as a bit of a great word.

Yeah, it's a great word. You know, you guys have become very influential, and definitely the people watching are very influential. Can we all collectively put our hands in the pot and all bring bozo back?

It would be really enjoyable.

Golden retrievers are kind of bozos.

Yeah. Lovable. No, no, no, no.

We have to talk about golden retriever mode.

Yeah, yeah. But yeah, better translation. Jordi, sorry I cut you off.

You're going to ask. Well, yeah, I think, I think what was interesting about your call is you were bullish on the platforms, but the take wasn't just you should invest in Facebook. It was like you should just put stuff out every single day. And you've been remarkably consistent with that. Yeah.

Yeah, and it was, it was real for 3 guys like us that could go from zero notoriety to a lot. And literally the most famous people in the world right now, you know, at least a third of them are either streamers or purely on social. They have no television, they have no, you know, theatrical, they have no other platform.

Yeah, TV is now coming to them and saying, hey, could we work something out? We— you clearly have demand for your content or we want to get it on our platform.

I started a company a couple years ago called VaynerWatt, W-A-T-T. It's a TV production company, and we are bombarded by the distributors for me to get access to the Stephen Bartletts or the Hormozis or the Cody Sanchezes or the— you like, it's all here. And what's crazy is, like, YouTube, for example, right? There's 900 genres that should be doing what you're doing right now.

Yeah, we just talked to some folks who are doing this show for Europe specifically, and there's one that's for car dealers and there's one for crypto and there's a whole bunch of different sort of like— and it's not even like we're the first live streamers, obviously. We just took them.

Still, the funny, you know, the craziest, the craziest thing, no one has done TVPN for the market open. Like you're in New York. Yeah, you're in New York, you start at.

I know someone that's working on it.

But yeah, you know, we know, we know a couple of people. But the fact that we're 18 months in and no one saw what we're doing— we go live at 11 Pacific, so we're going live at noon. The market has been open.

And which is, which is an indication of your laziness. You guys should wake up.

We do wake up early, but we do a team workout.

And it's more, it's more an indication of our, of our fascination with private markets versus public markets and, and like longer term, less quarter by quarter, less day by day market moves and more 12 to 18 month fundraising cycles for private technology companies. And so I think that my natural interest is in the private markets and the earlier stage technology. So if I was trying to cover, you know, the day by day in the market, what oil is doing and how that's affecting, you know, GE and Boeing, I would just be a little bit less sharp than if I'm talking about, you know, private market.

And honestly, that's what, you know, back to Crush It, back to you guys, back to all of this, what you're already like, the whole theme of this interview, the fact that people that are watching that are deeply sports like you, TV, PN for sports with talent like you guys is, is ESPN's biggest nightmare.

Um, you know, sort of have it, right?

Isn't that really happening? Of course, of course.

I mean, but they have to pay him like $60 million a year.

They do. And Pat had to be lucky enough of all the serendipity that led, led to that. There are 913 Pat McAfees, and I mean, and Pat's incredibly charismatic and talented. You guys are as well. But the long tail of knowledgeable or charismatic individuals is quite long. Now everyone has a different upside, right? Like, like, again, somebody could replicate this exact show. Yeah. And if they're really just much sharper and more entertaining, they will take market share. That's just how it works.

Well, the other, the other thing about sports versus business that's interesting is like sports, there's so many people that played sports, whether it's college or professionally, and then they stop being able to play at that level and they're like, well, what's the best thing I should do? I should just talk about the game that I love. Yeah. Whereas in business, you don't get to a point where you're better at talking about it necessarily than doing it. There's a lot of people that, that could— that's so well said.

And, and so, and there's another thing. I just turned 50, and like literally this morning, like no bullshit, I'm like, am I starting to enter my prime? Like, is it operating?

That's such a creative philosophy. You're just like, I'm having whatever the inverse of a midlife crisis is, that's me. You know what's funny?

Wait, wait, is the generational run just beginning?

No, I love it. But man, I really like— and it's funny, and I have a funny feeling, especially given the incredible nature of your audience. I think when that was just delivered a few minutes ago, there's just a lot of people that really understand it. Because, you know, my whole life, and probably many others watching here, you wish you were an athlete But if you're an entrepreneur, you realize at 30, you're like, wait a minute, if I was an athlete and I was great, I'd be wrapping it up or thinking about post. And I own a large sports agency. My brother and I own a very large sports agency now called VaynerSports. I feel for these kids. This is what they were meant to do. They're 35 years old and they're like, what do I do with my— just really interesting stuff.

No, I know. I truly, I feel so blessed. To have always been obsessed with business. Like, I never, I never came close. You know, I got very into surfing, but I never liked surfing as much as business and being— having a love for something that you can do for a lifetime. Even more so, people say like, oh, you should play tennis because you can play it, you know, you know, basically indefinitely. But it's like, well, you can, you can do deals Truly for life. That's a good segue.

Yeah, this is what we were— we always, we always bring up— got a segue. Sorry, one thing. We always, we always bring up Warren Buffett and Charlie Munger. It's like the greatest, the greatest run of that career happened from like '50 to '90.

It's like, that's so long. But wasn't it more like '70 to '90 or like '80?

It was, it was like '60. Like, that's right. So segue. You just brought up tennis. I've done incredibly well. I think this is important for a lot of people. Again, I'm always thinking about the audience, alternative sports league investing. I think you guys, I think there's some more fun stuff you guys could be doing about that. It's gone really well for me. Yeah.

Alexis, we had Alexis on. He has the running league, which is cool. We had the founder of Enhanced Games on. Yeah.

There's some article in the Journal today, is padel the next pickleball?

Pickleball, right?

So pickle was a huge win for me, right? I bought a team for $500,000 pre-COVID. They're trading at $16 million right now in our Major League Pickleball league. Well, right.

So do people watch pickleball?

Yes. That's why, that's why I invest. So I have a little bit of a secret of why I think I'm doing well, because I'm doing well with this wiffleball league that is starting to get some juice. We got some great owners in that. That's good. I have one little advantage. It's back to this, and we've had some meetings off camera, and you guys know I'm like pretty in the weeds with like how communication and marketing works. The reason I invested in Pickle is Pickle highlights, even 7 years ago, massively over-indexed on social. Oh, sure.

Yeah. Right. Like, yeah, but, but the, but the thing is, I do wonder if that's so. So Dana White has the, the slap fighting League. I forget what it's called. And that's something that does incredibly well on social, but I don't know anyone, even hardcore MMA fans, it's like, can't wait for the card this weekend, I'm gonna watch it live.

But two things. One is on incredibly well on social. You have to, you know, when you— because I've looked at that as well, and a lot of it is driven by the TKO machine much more than, you know, the organic nature of what I'm seeing.

And seeing somebody that's about to get hit in the face is like naturally scroll stopping, even if it's not a good—

there's also something that people don't realize about sports that matters if you're going to watch it, which is your capacity to participate in it. Yeah, right. Like, like, like people— a lot when people try to zing pickleball, they're like, yeah, everyone's playing it, but is it a business? And I always Like, literally, those are the articles that were written about tennis and golf 50, 60, 70, 80 years ago. It's like, it's funny, by the way, back to your first question, the other thing I do pretty well that I hope other people do, so I'm saying it out loud, is I really use history along with paying attention to people. So for example, the biggest sports in America in the '20s and '30s were horse racing, boxing, and baseball. Those sports were very good on radio, but football is not a great radio sport. Oh, interesting. Football exploded with the television. If you look at the explosion of the football, it came along with television. Yeah. And so back to streaming, what you guys are doing right now, one of the reasons I'm invested in the Sailing League, Slam Ball, you know, Pickleball, Wiffle Ball, Slam Ball, trampoline.

It's amazing.

You don't know what this dodgeball and trampoline—

no, no, basketball with trampoline so everyone can dunk and do like backflips and stuff. It's crazy.

All right, I got it.

The reason I did it was I knew I didn't need— just like you guys didn't need CNBC to be like, you're the right on-air talent— are all these alternative leagues do not need ESPN or, you know, Fox Sports to say yes. People can see this stuff through these platforms. And that's going to continue in an AR, VR— like, we're going into an incredible era of opportunity and meritocracy. Yeah.

Real quick, what was your reaction to the World Cup? Do you think that flopping will— and flopping culture will— will forever prevent America from actually getting into— do you think that's a bottleneck to growth?

No, that's not a bot. Americans are hypocrites. We love basketball. SGA is the greatest flopper of all time.

I don't watch enough.

I haven't watched enough basketball to have even seen SGA flop, but I feel like the— I feel like in Europe they appreciate the drama and it's almost like the acting of it all. But then you watch the replay and it's like, you know, Americans that are used to watching, you know, some corn-fed linebacker run through a guy and the guy just gets up instantly. You know, it's hard to then go watch somebody flop and completely fake an injury.

I, I think that soccer players at this point have this incredible 4 to 7% Academy Award-like acting capabilities that probably piss people off. But I agree with you, I'll use football because Football is a sport that people hate when they first see it, and then they become infatuated with it because it stops and starts. It stops and starts, especially the world immigrants coming to the US when they're used to soccer, which is flowing. They hate it. And then inevitably within a decade they love it because you start falling in love with all the little nuances. And I would say for Americans, the nuance of flopping is annoying, especially with the NBA. It's been a big topic. By the way, so I think especially this season. So I think some of that energy came to it. But look, I mean, I'm a little older than you guys. Everybody who's over 50 knows this. The strides that America has made in its interest in football, a.k.a. soccer, over the last 30 years is profound. FIFA, the video game, had a huge impact. Being able to watch the Premier League in America on Saturday mornings had a huge impact. And it's a great sport. And I already am excited for the next one.

What about the intersection of technology and a game like soccer? I grew up playing soccer, reffing soccer. I've reffed. That was like, you know, that was— I would spend my weekends when I was 12, I'd ref like 20 soccer games because they didn't cap it. Like there was no like, you can only work 8 hours. It was like, you can ref as many games as you want because there was a shortage. So I'd ref like an insane number of games in a day. And watching the World Cup, and again, I'm a super casual viewer. I hadn't been watching soccer prior to the World Cup, but it was fun to kind of catch up on it. Like, you know, there'd be an offside call and, or maybe there wasn't a call and then they'd run the tape and they figured out that the guy's hand, you know, elbow was slightly over the line and they'd call it and they'd take the goal back. And that, that to me, like, part of, part of the, part of like what makes the sport great is there's just players that can finesse a situation and just be right on that line. So a ref couldn't possibly call it. And then to have a machine being like, you know, no, you can't, you could beat the ref, but not the machine.

Look, I think of this like marketing. I'll make this valuable to the audience who does not like sports.

This is slow, by the way.

I thought those were the Chinese robots.

No, these are— now we're watching slam ball best hits. Look at this.

Yeah, this— boom! Look at this.

Oh, all right, all right. I can get behind this.

It's so insane.

Is this made to look like it was shot on VHS? What's going on here? Oh, okay, so it's been around for a while.

A little bit more. Um, so I'll say I'll answer you like this, to make it a little business terms, hopefully help some people's marketing. There's a lot of people watching right now that become too one-dimensional math performance marketers, and they lose. In fact, you boys were one of my favorite executions to watch from afar, because you guys played like me. You sell more on brand than on math. You guys were getting brand deals that were completely inappropriate because you sold on value not on metrics. Yeah. I think that that's how I think about technology in sports. If we get all the way to full AI robots and everything's 100% black and white, I think the league's going to learn pretty quickly that we as humans need a little romance, need a little gray, not black and white, need a little of a balance. And so just like I wish in political terms that everyone was purple instead of this garbage red and blue we're in. With the sports and technology and replay, if they can find purple, I think it's great. When they go too red or too blue, I think it's a problem.

Yeah. Even in Formula 1, the drivers are having issues right now because there's AI is deciding how much energy, fuel, and how much power to— and they're all just getting angry because they're like, am I losing because I'm not as fast as my opponent or is my— Is the AI running on my car just not as dialed? It's all these analogies.

It's like, is dinner fun without the side dish and the amuse-bouche and dessert? When we get so binary, we all want a little romance. We need a little romance.

Okay, one more potential league. I want you to run this down and let us know next time you're on what you find. But I have a theory that when every car on the road is autonomous, that driving on the track is going to have a resurgence in popularity because people are going to be— it's so romantic. It's so romantic. There's so much great history around racing. Every person— Hit me with the counter. Yeah, hit me with the counter. But I have a theory that things like there's something like a slam, slam ball or whatever it is for driving. And that's just like demolition derby.

Oh, it's something like rallying.

I was driving yesterday, I was driving a, I was driving a dune buggy. I, I was like on a trampoline, my buddy, and we were like, we were laughing, it was so much fun. We're drifting around.

Yeah, yeah, yeah, it's a more dynamic calendar.

Is— I don't know if you know this, but we used to ride horses before the cars.

Oh yeah.

And I don't know if you know, there's something called horse racing. Yeah, that is, of course, pretty out there. And so the counter would be that we kind of just lived it with horses and horse racing. But the counter to my counter is what you said, which is I actually think horse racing could be massive in this country if they put a contemporary flair on it. What UFC showed is like it was an evolution of boxing with more— Oh, interesting.

Yeah, you want like rally horse racing.

Sort of like UFC meets the Kentucky Derby. Like, you can punch the other jockeys.

Or what about none of that, and it's more about— what about storytelling?

What about— yes, yes, yes, A Drive to Survive, that type of thing. Yeah.

Or like, even when we're watching an actual horse race on a Saturday that's not the Kentucky Derby, like, if you turned it on and the bottom of the screen was actually a mic version of every jockey, and you're learning, like, We want layers. We want layers. We want stories. We want to get captivated. Everything is about that in every aspect of life.

With two jockeys getting up in each other's faces and throwing down after a rum—

You guys are looking violent today.

Okay, here's a hard-hitting question. Do you think Big Boogie turned it down or was he capping?

Ooh, that is a good question.

Do you think he potentially— 'Cause we were trying to do that. Jon did the math and he said that if he was going to dental school, for a decade and he had a bunch of dependents, it's possible the total package could have been in the, in the $10 million range. But you, you cover NIL stuff a lot more. True.

You'll appreciate this. This is one of the rare moments where I get to do this and I always look for it. So thank you, man. I'm going to clip this. The answer is I don't know. I've just not paid enough attention. I've been a little bit of a bubble. I'm very I'm in a very weird place right now. I've been in a 7 to 10 week cocoon around the collectibles industry becoming the— that collecting becoming the next lifestyle genre. Yeah. And so like my consumption of anything happening in pop culture is at an all-time low the last 2 months because I've been in the lab, like in the lab, on this thesis that I've been on for about 8 years that I think is about to completely blow the roofs off of. Like, for example, let me give you a for instance to give you even more color. Again, always try to think about how to bring value to this smart show. We've started at VaynerMedia a collectibles marketing division, just like there started— just like Coca-Cola started a sports marketing division when the Olympics were getting important as a driver to consumption. I believe in 10 years, every Fortune 500 consumable product like apparel, CPG, will have a collectibles marketing strategy. I don't know if you guys saw lines around the door at Popeye's Chicken for the One Piece collectible, people flooding to a Dodgers game to get the One Piece collectible. I think that differentiation in an AEO, GEO, you know, agentic commerce world, differentiation around collectibles and community to get people to stop their normal pattern of what they're just reordering because they want to buy a deodorant that has a collectible of their favorite. I just see collectibles eating up because of what it means for community and analog and value and investment.

Yeah, and AI creates like insane— is going to continue to create insane abundance. Like, one example is everyone, everyone can have a SaaS product now. You can make it. Everyone will. Everyone will.

What's the chain?

Everyone wants to go ahead, brother.

What's the chain from people are playing a sport, is it watchable, then it becomes a sport with teams, professional? Uh, you identified some of the translation effects, like, uh, you know, is anybody— is anybody going to watch golf? Sure, a lot of people play it, but will people watch it? Obviously the answer was yes. And that was true for football before there was a league and there was TV and the Super Bowl, right? People were just playing.

By the way, just for fun, because I'm just nerding out, please, like everyone said there would never be a professional football league, that football was always going to be college. Yep. Notre Dame, the '20s, the '30s, like, yeah, like smart people that come on every This One Is Shit, like those people back in the '40s and '50s were like, there will never be professional football.

Yes. But at the same time, there are sports that are played by lots of people. I'm thinking of like running. And running is not a massive— there's not the Super Bowl of running, even though millions and millions of Americans go jogging and they participate in the sport. There's something special about competitive football that makes it the Super Bowl, right?

Football is very special. But like, you know, the New York City Marathon is a very cultural event with good ratings. Yeah. You know, I think we're in the era where an entrepreneur watching this show right now is like, fuck it, I'm making a running league, and it might hit.

Yeah. So then talk to me about the translation layer from there. There's a sport that a lot of people are playing. If it's watchable, it becomes a big national sport with professional teams that we're all watching. Uh, at what point does it become— is there a collectible culture around it? Because We're just in this early stages of like pickleball has been played by a lot of people. It's now starting to be watched by people, as you mentioned. When will people be collecting pickleball cards?

Yeah, your production team seems really capable based on what I just saw. Yeah. Can the production team, before I leave in the next 3 or 4 minutes, put up eBay, type in Anna Leigh Waters. Leigh is L-E-I-G-H. Anna, A-N-N-A space Leigh, L-E-I-G-H, space Waters, then enter on eBay, and then go to completed and sold items, and then sorted by most valuable. If the team can do this—

this is a great challenge for the team.

Are you just trying to move some Anna Leigh Waters merch right now? You're like, and then click purchase, and then click check out.

I'm kidding, but she is the star of our team. Okay, and we do represent her sports. She's my— and there's my— she was our Joe Namath. She is our Larry Bird. Look at this.

Here we go. Now hit on the left side.

On the left side, hit completed and sold.

Yeah, real eBay ball knowers. No, they know Dylan on our team. Yeah, big collector.

Nope, no, no, all the way down. I'm very hardcore. I'm very in the dirt on collectible culture. The completed sold.

Where are we going? Hunting around for Anna Lee Waters. Oh, show us this, show us this. Log in, show us the password.

Now, so, so Dylan on the— on our team already got there. There we go. And he's seeing— share screenshot— Anna Lee Waters rookie autograph cards going for $8,000, $5,000, $5,000.

Okay, we've done a very good job with Anna Leigh. We've made her famous. She's on Fallon, she's on McAfee, she's the face of the sport. Ryan Harwood at VaynerSports deserves a lot of credit. We've crushed with her. She's the best player in the world. Pickleball is the only sport where a female can be the best. She's the best player in the sport, including the men.

Wow.

Because, you know, that's what's interesting about pickleball. 80-year-old grandmas are whipping 19-year-old grandsons in pickleball. It's a different kind of sport.

And there is another sport where women outperform men, uh, certain equestrian in the Olympics.

I'm sure there's others, but this is one that people know. Uh, you could see 8,000 to 5,000, and I will tell you, 99% of the audience and everyone on Twitter right now Hit me up, Gary Vee, and answer this. 99% of the audience right now has never heard of her in their lives. So it doesn't take a lot to get to collectible because collectibles are niche, are tribals, right? Look what happened with Comic-Con. 50 guys in a weird hotel in San Diego, right, in 1976, to now this last weekend, millions of people going to San Diego and it's a cultural phenomenon. Crazy. So collectibles is when it flips over to pop culture. I would argue that athletes that are the most valuable are more pop culture than people realize. They're more superheroes than people realize. And when you're building a league, whether it's Conor McGregor and Ronda Rousey, whether it's Michael Jordan, Larry Bird, Steph Curry today, like, you need a superhero. Joe Namath was the superhero for the NFL early on. He became one of the most famous people in the world. The Jets had a huge upset in Super Bowl III and football was on its way.

Mm-hmm. Business collectibles.

Where do you think it's going? We saw Jensen Jack go for about $1 million. That was for charity for our friend Justin Mares. Yeah. His grants that he does. I believe collectibles.

I believe collectibles have eaten up the pop culture energy of what young 20-year-olds would do with public stocks. So instead of buying 50 shares of Nike like my generation did, they're going to go buy a collectible. I think it's completely eaten up the oxygen of contemporary art. So you guys, especially after your nice transaction, continue to hang around people of wealth at a young age. Your contemporaries, people in their 30s, 40s, 50s, only 25 years ago, 30 years ago, are buying the hottest contemporary artists to put the art in the house. Much of that money is now going into $100,000, million-dollar collectibles. Obviously, Michael Rubin sitting on the top of the Tops empire with all his cultural relevance has brought an incredible amount of pop culture energy to it. And kids— people forget trading cards are 100 years of collectibility now. This has been a very long-term asset class, but the numbers are just out of control. You see O'Leary wearing this, you know, like, you know this, even for you guys that are young, definitely people in my generation, people that collected cards as teenagers in the '80s and '90s were nerdy. People collecting cards are now cool. And once that happens, it gets really different really fast. And we're on that other side. And so you'll see collect— you know, watch this. I want to know what's on your shirt. I wasn't even— I was right there. You go, look at that.

Are you going to the National?

Of course, leaving tomorrow. Let me just make this so I can clip it in 9 years. I am building the next Disney Pokémon marble all in everyone's faces. It's called VeeFriends. And because I have 13 different other jobs, no one sees it, and I don't speak about it that often. But you're building an IP business during this AI era. Is going to end up being one of my savviest moves.

Yeah.

And the generational run is just beginning. I love it. Clip that. Clip that in a decade.

Clip that in 5 decades.

I mean, 99, 100.

Thank you, Mattman. Thanks so much, Rob. Great to catch up. Always great to catch up. You're the man.

Let me know if I can help you.

We'll talk soon. Good luck at the National. Goodbye. Cheers. Let me tell you about console gaming. Console-built AI agents that automate 70% of IT, HR, and finance support, giving employees instant resolution for access requests and password resets.

Up next, Jonathan Jacoby. Jacoby. Jacoby coming in hot to the Ultradome.

How you doing?

Hey guys, thank you for having me. Great to have you. Great to have you. Let's get right into the gong-worthy moment of this segment.

Tell us about the news?

Yes, so today we launched Enigma out of stealth, um, raising $71 million seed round. In— $71 million seed round.

It had been a while since I saw a $70+ million seed round, and I feel more at ease. Ribbit Capital having, having— Thank you so much.

So what, uh, yeah, yeah, what is the progress need to look like these days? How, how mature is the product? How built out is the team to jump straight to that size of seed round? Yeah.

So we, we're a team of 26 people actually grown pretty quickly. And today we also launched something pretty unique. We launched like the first AI robotic experience online so people can just browse to a website to robot that online. And interact with real AI-powered robots by our models and our interfaces. So this is kind of a double milestone for us, both emerging out of stealth and kind of putting our technology out there at the same time.

Very cool. I saw the Painter, which is a robot that you can control and, you know, paint in the real world. I was a little bit frustrated because I had been writing down notes and talking with a few people about somebody should set up a robotic arm with that lets you paint on real canvases and then send it, like, send it to yourself. But you're effectively doing that. So you beat me to the punch.

Yeah. So where, where does the AI start? Where does the teleoperation end? Because I believe there's been demos of, you know, deterministic robot or like remote control even like a remote-controlled camera that you can pan, tilt, and zoom. That's existed for a long time. Where does the AI actually come into the process?

Yeah, so basically, we have robots. You can talk to them. So basically, the idea is we put out different interfaces. One, in some of the interfaces, again, as you can see here, you can talk to the robot and ask it things, and the robot just responds to you. We're going to also, throughout the week, we're going to launch more and more different experiences. And we're going to showcase different ways to interact with robots. So our company is a research-heavy, research lab, research-heavy company that focuses on, uh, robotic foundation models, but we also think deeply and care deeply about the experience. We think novel user interfaces are going to be, uh, a huge enabler for robotics. So we kind of wanted to put out something that is very wide to learn how people want to interact with robots and and take that learning and put it back to the research. So in those experiments you're seeing, you're seeing one robot that you can fight with, like a robotic sword fight. So here it is. So one of them is controlled manually, like a teleoperation, which is the user. And there is a robot that is an AI robot that responds to the user. So it actually fights. So this is one way of interaction. Another one is you talk to the robot something, and you see what it does. So the painter, you give it an association, it paints it plays like Pictionary with you. And then after the experience, you can like ask it to do some pantomimes. You can imagine the brush is anything and it can be a wand. It can be like cleaning the— like as a broomstick cleaning the floor. So this is where the AI comes in. This is where our models, our foundation models comes in, comes into play. And we wanted to showcase both interfaces and the models.

What's your opinion on robotics sort of, or robotic control coming for free in generalized LLM development. I saw some demos of, I think it was Opus 5 controlling a robotic arm. It seems like a lot of these models, they don't always need to be trained for something specific and then you wind up just getting some particular functionality. For free or it's like a quick RL step on top of what the Frontier Labs are already doing. Is that relevant to your roadmap or do you think there's a convergence or do you think you'll go off and carve out a specific niche in robotics? I mean, obviously it's a huge category, but how does that all fit together?

It is a huge category.

That's true. Yeah.

So I see the demo you're talking about, I think, was anthropic. Yeah. So we've seen those demos actually we have from GPT-3.5. And before that, we've had robots controlled by LLMs. I think the problem that we're seeing labs such as ourselves trying to tackle is the fact that this doesn't really either generalize very well or doesn't fit in with the requirements of dexterity or doesn't—

Yeah, it feels like a cool demo, but not necessarily commercially viable.

Yeah, it also might just be like— it's impressive to be like, wow, language model controlled arm, but if it's like every time it moved, it was $20 in tokens and it was 45% accurate. And if you just like did an actual model for the robot, it's like 2 cents and perfect.

That would be great. Exactly, exactly. And it's kind of one example one of my friends told me was assume you could have had an LLM like do an ASCII art and like paint as if it's ASCII art, but we still have image models. —like the natural language of the way the LLMs work. And that's why it's a bit off. But I do think that— I think cognition models for robotics, such as what we're building and what some other labs are doing as well, I think it's going to take a huge part of how robotics is going to evolve. And we're also very lucky to be in this time of great tailwind, right? There's a great tailwind of research. Most of the team is doing research and we have a lot of tables from, you know, LLMs research. We take inspiration from that or from video models or from world models. And that's pretty, that's pretty, that's pretty unique.

Time to, time to revenue. How are you, how are you thinking about like first commercial application? Sounds like you're in the age of research, but when, when do you see opportunity to start commercializing and kind of where?

Yeah, so we have partners we're working on, working with right now across different kind of segments. So from logistics to entertainment to health as well. Some cool things on that that we might talk about on those 3 categories. And there's more. What we're seeing is we want to be able to get as much value to users and customers as fast as possible and make sure that when we are scaling it up, it's scale up in the right kind of directionally right approach because you don't want to be way off and then be dependent on kind of, kind of being in a big tech debt, I would say, in product debt. But I think, I think we're at a point where we also, we're seeing value. We're seeing, you know, value from partners. And it's actually a kind of showing signs that we didn't expect to occur that fast.

How do you think about vertical integration? It feels like there's robotics companies that are saying we're going to build the robot and the model. There's others that are like we're going to build a modular robot and the model. There's other companies that are saying we'll just build the model and the robot builders will show up. I haven't seen actually that many new startups that are like we're just building a dumb robotic arm that you can load whatever model you want on. We want nothing to do with the actual AI inside. We're just doing the manufacturing. Maybe that's an opportunity, but where do you see your scope growing or staying focused?

I think it's a great question because it really is a divergent kind of topic, a lot of opinions. We see ourselves as— so we're a software-first company, right? So we work with our product, our kind of software stack from the models to the interfaces to, to talk to them. We work very hard on making it generalized across, uh, uh, embodiment. So we want to get to a point that even a new manufacturer comes out and says, all right, here's a new robot. I would want, I would want to be able to within a week have everything work on that robot as if I had that robot for months. So we're building a lot of infrastructure to be able to cut— like, this is part of our research, like how to be able to calibrate all our technology across robots easily. We think and we see a lot of vendors and we think it doesn't necessarily have to be coupled. And we see a lot of great companies around the world that sell the hardware. So if we could be a good provider for the brain and for the interface of the brain, we think it's a very viable kind of product offering. And I think it makes sense as well.

Yeah, yeah, that's cool. I mean, yeah, obviously there are a ton of robotic production or supplier companies. There aren't enough startups in that world, I feel like. I feel like we need, like, I'm a direct competitor to KUKA or something, and I'll make a big robotic arm and I'll make it cheaper or whatever.

I'm sure there are out there. What were you doing before you got— what were you doing before this?

So I worked at a company, I worked at a company acquired by Wiz and then got acquired by Google. Oh, cool. Yeah. And actually, me and Asaf, the CEO of Wiz, we go way back. When I was 17, he was running Microsoft in Israel, and he brought me to join Microsoft. And so we have a long history. So I joined Wiz, and yeah, I think the era right now— Long story short, generational run.

Amazing. Well, thank you so much for coming on the show, breaking it down. Congratulations. Congratulations on coming out of stealth. Great demo. Thank you.

That's a very cool way to launch. Huge. And we'll, we'll talk to you soon, hopefully. I'm sure we will.

Thank you very much. Goodbye. Let me tell you about MongoDB. What's the only thing faster than the AI market? Your business on MongoDB. Don't just build AI, own the data platform that powers it. Up next, we have Dean Meyer from Sequoia Capital here to talk about America's open model paradigm. Paradox. I love a good paradox. Also, fun fact, I heard somebody debunking Jevons paradox. It's not a paradox, it's just, just an effect. It's just an economic effect. Is it a real paradox? Break down America's open model paradox. Feel free to give a little bit of an intro. What's up, guys? But how you doing? What's up?

I'm good. Can you, can you guys hear me?

Oh yeah, we can hear you. I can hear you.

Let's go! We can hear you.

What's up, guys? Great to be on the show. Thanks for having me. Look, you know, myself and my partner Constantine at Sequoia, you know, we've been riffing on sovereign AI for a while. And I think kind of, you know, we wrote this post that went a little, a little viral and kind of the core concept was, you know, US is creating this frontier capability that kind of gets leaked into Chinese models. And then you have the entirety of the American kind of builder ecosystem, which is Neo Labs, you know, companies that are somewhat dependent on these Chinese models. And so, you know, today there's like, you know, if you're an American lab and you can legally learn from a Chinese model, but you can't learn from a kind of frontier US model, which is what they're doing, then the question is how do we build kind of the, you know, the legal path almost for this direct route as opposed to just having this dependence on Chinese models? And so we kind of address that and share a few ideas, which I'm happy to talk about. But that's kind of like the overall, the overall thing.

Yeah. I mean, there's so much, there's so much news about this particular topic. The essay was written at the perfect time, just, just before Kimmy K threw that.

It was good timing. We had no intel. We had no intel.

But I mean, were you, were you a believer in, in just the fact that the open source market generally lags the frontier by 6 months and that's not changing because there was this graph that was going around for a while that was showing that maybe the closed source was scaling on a slightly steeper curve than open source and this was driving sort of— a couple of months ago when we were in between open source releases, people were starting to say, "Oh, maybe they will plateau because the training cost will be so high that you'll have sort of like what happened at Meta. That will just happen in China because China will just say, yeah, like if we're going to spend $1 billion on this training run, it's not worth it to open source it. But it feels like they found an economic model that makes sense. KIMI's license is out today and they're going to take a portion of revenue if you're really selling a lot of KIMI tokens. And so Were you— were you— was this built on an idea that, that the open source and closed source frontiers would always be on basically the same track?

Yeah, well, I think— I think, look, that is— that is the question. Yeah. Yeah. And so, look, our view is, I mean, first of all, like, the gap has closed a lot. And so kind of the key conversation online over the past week was is distillation the reason for this or is there some other kind of mysterious reason? And the truth is, there are two camps, like everything. People are kind of extremists when there's conversation. There's a group that are like, hey, distillation does nothing. And there's a group that's like, distillation everything. The truth is, from the conversations we've had, it's somewhere in the middle, right? Whether it's 20%, 30%, 40%, it doesn't really matter as much. It is meaningful, that we know. We know from the labs, you know, there's tens of millions of queries through these like large-scale proxy distillation accounts. We also know on the other hand, if you speak to anyone who's like leading in synthetic data gen on pre or post-training at the labs, like they will tell you, even if you do like hardcore KYC and then like let's say you get access to these kind of like, you know, you don't have all the traces, you don't have kind of, you know, like the chain of thought. There's a lot of RL that happens, right? And you just get the output. Like, let's assume you just get an output from a query or prompt. It is still extremely valuable. Like, that is 100% true. And so we know distillation is real. And then the question becomes, well, distillation is real. You know, how do we address the asymmetry that exists between, you know, if you're a Western, you know, US company or, you know, just a Western company, you're not going to like go and set up, you know, 35,000 proxy accounts with like different IPs around the world and go into is still Anthropic for the fear of being sued, uh, and being called out and even being cut, you know, shutting off access. And so there is the structural asymmetry, um, and, and I think it's something we have to think about.

So you have to create— yeah, go ahead. You have to create the meme of token leader. Maybe, maybe.

Uh, do you think that Anthropic or anyone else— uh, let's just say a, a closed-source frontier lab in America If they believe that they've been distilled by an international company, should they just sue? Even if there's no— even if there's no way that on any reasonable time scale that they're going to win a settlement or block anything, would it just throw down the gauntlet to say, we're willing to put this into the, the official record under threat of perjury that we believe this is what happened, and there's like a signed affidavit of their belief that there is wrongdoing?

I mean, I think they probably would and nothing would happen. Like, I think nothing would happen. And so anyway, like, you know, US companies are kind of deterred from doing this. So I don't think it would change too much, unfortunately.

It would be very symbolic. I think it would be. I agree with you. I think— I don't think anything would change from the actual lawsuit. It would cost some amount of money to file this lawsuit. Lawsuit, but at least it would— it would have— they would have something to point to of like, hey, look, we tried to go through the traditional path and you were able to follow along, you were able to look at our evidence which was entered into the public record, and oh yeah, the, the CEO of the lab that we're accusing of distilling us didn't show up to court, or whatever, you know, whatever the process is. The process can, can begin. Maybe this is like a day away, maybe like the lawsuit drops tomorrow, who knows. But it seems like that's the next logical step because right now there's a lot of pushback against like, hey, why do you have a problem with this? They say, oh, maybe it's distilled. Well, you know, that's the next step, maybe, I don't know. Yeah.

So, Kimi came out with their license this morning and they— in the license it says if you're just running it, hosting this yourself for your own use cases, that's fine. Yeah, it's free. But if you're effectively reselling it as a NeoCloud, you have to give us a revenue share. I can't imagine that Washington is going to tolerate that. Just, just the idea, the idea that like a bunch of US NeoClouds are then sending billions of dollars back to, to China to support one of their leading labs. I just, to me,— that feels like it could provoke a response that felt like it was maybe coming last week when different people in the admin were saying, "Hey, we believe distillation is happening here.

We're pro-open source, but we're against how they were positioning it, IP theft." On the flip side, I was very quickly able to find a Bottega Veneta dupe that comes from China and is effectively distilled on that brand And the government isn't stepping in about that, you know, and that money is flowing straight to it through Amazon and through other platforms. Shein and—

I just— Yeah, but it's not— it doesn't have any geopolitical significance. You're not— and you wouldn't go to war for— and it's a carrying company.

Yeah. No, that's true. That's true.

Anyway, sorry. I was— No, I think— look, I think— look, I was surprised to read that. Look, the truth is, like, I think it's not dissimilar from the database market where, like, You know, you had Redis like early on. There were a range of these companies released really cool open source and like they just got the hyperscalers copied them effectively. And that was the end. I just, you know, it's impossible to enforce. I like, I don't think that will work for the Chinese. But look, the broader question here is like there are two major risks with like depending on a Chinese-US, you know, Chinese open source ecosystem. The first one is is the dependency risk. Like, the framing we have is, you know, in some sense, you know, Sequoia, we always say we're only as good as our next investment, which we actually believe. And I think in some sense the analogy is like, you're only as good as your next model. And so as long as you're dependent on the Chinese to give you the next, like, hit of, you know, of, of the thing, you're kind of, you're kind of in a bit of a, in a, you know, in a bit of a weird position where you have this dependency. And then the other thing that I don't think people are kind of well calibrated on is just the risk of backdoors. And I'm not saying, like, just to be clear, I'm not saying that all Chinese OpenAI labs are backdooring models, but the truth is we know for sure, and there's like a lot of research on this, both from the labs and outside the labs in academia. It is like an NP-hard problem to know if there is a backdoor in a model that we know. So it's impossible to verify. And on top of that, there's like a ton of crazy evidence that shows you can, you can like hand-pull a model, which basically, you know, one way of saying is like a model, you could do all the evals you want. And suddenly, like, you know, it gets a certain input and the model literally changes behavior. It will start like exfiltrating data and doing some crazy stuff. And like this, this stuff's real. Like this is not science fiction. We know. And so kind of the analogy I would give you guys is like, I think the state of AI security today is very similar to that of early days of the internet when in 2001 Wi-Fi encryption was effectively broken, publicly broken, and still no one did anything. And I don't even think we're at that moment yet. And it's going to get really bad if we, one, don't better understand how to secure these things, and two, build our own kind of independent ecosystem. And so, yeah, those would be like some thoughts as well.

There's so much more we could go into here. We got to have you back on the show and talk more because like there's so much going on. But thank you for writing the essay. We'll let the audience go and read the full piece.

Yeah, great to have you on and come back soon.

I'm looking forward to the next one because it feels like the conversation is already going in so many different directions. We didn't even get into the— The politics of anyone having a cyber weapon in their pocket. There's all these different scenarios to play out beyond what you just mentioned, which are a bunch of other considerations. So it's a fun one. 100%.

Would love to talk to you again.

Guys, thank you so much. Yeah, this was awesome.

Great to have you on.

Have a good rest of your day.

Talk to you later.

Bye. Let me tell you about Public. Investing for those who take it seriously. They got stocks, options, bonds, crypto, treasuries, and more with great customer service. Our next guest, is Brent Franson from Death Clock. He's the founder and CEO. Welcome to the show, Brent. How you doing? I'm well, thanks for having me. Please introduce yourself and the company. Tell us a little bit about Death Clock.

Yeah, so, uh, we trained an AI on 1,200 longevity studies to predict how long you're going to live originally, and then the model was predicting the exact day you would die. And so we started sharing that And people like, you know, people either love it or hate it. They want to see the exact day they'll die. But more importantly, the model builds a plan, you know, to live longer. It's not just about when you're going to die.

Is this a nod to Metalocalypse from Adult Swim?

No, it's not. Although that's one of the most, the most common questions, the most common responses on social about Metalocalypse. Yeah. And then just about the, like, religion. You know, only God can tell you when you're going to die.

Yeah. Is this rage bait? Are you intentionally framing this? Because you could have called it like the Longevity Optimizer, but Death Clock.

I don't think they would have gotten nearly as many users.

I agree. But what are your opinions about— I mean, I've literally like lived this as well. So I'm not actually, you know, saying that this is bad or anything, but I'm just wondering about about how conscious the decision is to use a word like death because Liquid Death, same thing, where it's this aggressive name, it draws in attention. What are your thoughts on that as like an attention mechanism?

Yeah, I mean, I think there's a stoic component to it. And so I do think that there is some, I don't know, some good reason to be honest with ourselves about the fact that we're gonna die. But the real answer is, I don't know, distribution really matters. It's hard to make noise. And so we wanted to be really provocative. So our goal is very much, you know, I don't know, the Howard Stern ilk of, look, love us or hate us, just don't be apathetic about what we're doing.

No, no, I think it's a good strategy for the modern era.

Talk about the new launch. You launched free nationwide blood testing. Sounds good. Too good to be true. What's the catch?

It's not. You know, people are really concerned about their data and I get it. We don't do anything. With the data. We don't share or sell the data. But I think what's happened is we have a lot of these blood testing companies that do all this kitchen sink blood testing and they charge you a lot of money. And the reality is, is blood testing is commoditized. LabCorp and Quest have existed for a long time. These tests are not that expensive. And the set of tests that you really need to understand heart health, metabolic health, inflammation, those are These are pretty inexpensive tests. And so from an acquisition perspective, for us, it made sense to say, look, let's give away a test. You can get a sense of the platform and how good the plan that it generates for you is. And then we think you'll do more testing and you'll go deeper and you'll convert to premium. But, you know, giving away a basic test, it's in the tens of dollars. This stuff is not as expensive people think.

Okay, and, and how many biomarkers does the free test give you?

The free test gives you about 50, but this whole number of biomarkers thing, it's the wrong way to think about this kind of testing. And so I think we've gotten into this place, we think, okay, you know, 150 biomarkers is better than 50 biomarkers, and that's not necessarily true. I mean, when you run these blood tests, they've got about a 95% accuracy rate. So when you start just testing for everything, even asymptomatically, you're cranking up the odds of a false positive that you've got to go chase down with your doc, and then you're in some diagnostic process that actually carries some risk. Yeah. And so it's 50— you can do all of that testing with us, but it's more about getting the right tests. I mean, really understanding the right things about your heart health, your metabolic health, making sure you're doing the right cancer screens.

Probably get too caught up. There's like a power law in efficacy of biomarkers. And there's also probably a lot of correlation between some of those longer tail biomarkers.

Okay, so, so someone comes in, technically you're, you're having to fork over tens of dollars to get them a free test and you're betting that some meaningful percentage of them will be like, this service is cool, I'll stick around and access other services. So It sounds like close, as close as possible to a free lunch, which, which I like. Yeah.

I mean, it really is free. There's no— yeah, there's no gotcha in it. I mean, you do have to say, no, I don't want to. I don't want to take an Alzheimer's, you know, I don't want to look for the Alzheimer's gene. I don't want to do more advanced cardiac testing. All of those things require you to upgrade. But I think, I think the future is one in which we're kind of acting as our own personal doctors. We've got an AI that's got all of our health data, and then we order the tests very surgically that we need at the times that we need 'em. We're not just kitchen sink doing 15 vials, you know, every 6, every 6 months. Like, okay, I should go look at my liver health because of a medication I'm taking. And you shouldn't need a doctor to do all of that. And so that's what we're trying to make available mass market. We're not gonna be the, high-end solution.

What, what legally can, can only a doctor do these days?

Well, medications, cancer screens, but I mean, blood testing. So there is a doctor—

I mean, obviously blood testing happens with labs, so those are licensed. But then in terms of the interpretation of the tests, that can sort of be democratized with AI and individual, like, self-serve products.

Yeah, I mean, you know, we have to disclaim that we are, you know, we're not a doctor and you need to speak to your doctor about any of these results. But there are, you know, the barriers are coming down in, in lab testing. I mean, so we— every lab order that goes through Death Clock, it has a doctor's name on it. Like, it goes through an approval process, but it's very fast and you don't need to speak directly to the doctor. Yeah, um, because there's no, you know, there's really no risk for most of these tests.

What were you doing before this?

I was the CEO of a company called Euclid Analytics, a benchmark and NEA-backed business that was acquired by WeWork about 6 months before WeWork imploded. It was a crazy, crazy ride.

Wild. Well, I'm so curious what, uh, Adam Neumann had in store. WeHealth, maybe?

Oh yeah, I don't know.

He was a, he was a, he was a crazy charismatic guy. He He was one of those guys.

He's still alive, by the way. Yeah, yes, he's still in. I don't think he's dead.

You would go sit and he would talk for 2 hours in these all-hands. Wow. And it would feel like it was 15 minutes. But the day he showed up to an all-hands in a suit, we all looked around like, oh, we're screwed.

Oh, that's good. Question, biological age, how should consumers read into it. It seems like there's this race to the bottom in biological age where the platforms have an incentive to, to rate someone as a lower biological age because they're more likely to flex it on social media or to their friends. And then the other question I have is, is that just comparing your— that's just comparing your, your labs to the average American. And we know the average American is deeply unhealthy. So is it even when, if I'm like, I'm 30 years old, if my biological age is 25, is that basically just saying like you are healthier than the average American who is deeply unhealthy?

I recently did one of those tests and it said I have the mind of a 5-year-old biologically. We've already told this joke, but it's funny every time. It gets good every time.

No, it is gamed. And I think there are some companies that are pretty reckless with that, they're telling you that you're young, that you're younger than any real— it's called a phenotypic age is what would be kind of the fancy name for it— would actually tell you. And so I think that's a little concerning that people are doing that. Um, where it comes from, there's a researcher at Yale who produced the most valid calculation of biological age, and it takes 9 biomarkers, and it's a little bit more— it's a little a little bit fancier than just comparing you to average, but I think that's directionally right. And so I wouldn't put too much stock in biological age. I would take very seriously heart health markers, metabolic markers, inflammatory markers. And these are the basics. It's like, know what your LDL is, know what your hemoglobin A1c is, know what your hs-CRP is. C-reactive protein is going to give you a sense of inflammation. If you're paying attention to those things, you're going to do way more than, you know, caring too much about a biological age.

What's your date? Oh yeah. How long?

Oh, my date is— yeah. So mine's at 78. It's like February 16th, 2065 or something. It's weird to get an actual date, but the average user has 10 to 12 years that they can add. So the models also producing, how much longer it thinks you can live. And then any time your health data is updated, the calculation recalculates. So my date changes every day. But when I, when I started, it was 78. I partied a little too much.

Have you pitched a partnership with Nas Daily? Are you familiar with this YouTuber, Facebook creator? He, he's obsessed with this concept. He wears a t-shirt that has a percentage loading bar for how far he is through his life based on life expectancy. And it's all a message about, you know, like enjoying every day. But he gets like hundreds of billions of views or millions of lots of views. Very, very popular creator. Would be a very logical sponsorship maybe one day.

Yeah, we should talk to them. I mean, I think the reality is, is like we're all going to die and we're going to just be—

not me, built different. But continue.

We are all going to die, no matter what people are going to tell you. And there are things you can do to live longer, but—

No, sorry, we're having too much fun. Money never sleeps.

You shouldn't either. I hope you're right, actually. I really, really hope you're right.

Just say you're not AGI-pilled. Yeah, not AGI-pilled. Say you're not AGI-pilled. We're going to cure everything.

No, I think you're right. And I think just reframing mentally around what it means to be healthy, giving people another visualization on what is typically just a bunch of random nanoliters to deciliters, like it's a bunch of random stats.

Also, why people are getting health information constantly, at least I am. Take this supplement. Do this or that. I've always been into experimenting with different protocols and things like that, but sometimes you need to kind of jolt people. Yeah. You know, so.

Well, I think it's fine if you're experimenting so long as you're getting the basics right. I mean, I think if you're getting a full body scan but not getting colonoscopies, that's wrong. I wish I could throw, you know, throw the X up. But if you're like, you've done the colonoscopy and you've done the mammogram and you've done everything, then like, okay, fine, maybe get the full body scan. And then on the experimenting, you gotta be careful. You know, the story of thalidomide, the doctors were prescribing that off-label to pregnant women, and then the babies were being born with like hands on their stomachs. Yeah, really, really gross. Yeah, so you gotta, you know, the experimenting can be a little bit, I think we're a little casual about the experimenting.

I'm not saying don't do it, but. Yeah. Well, thank you so much for taking the time to chat with us.

Very fascinating.

Congrats on the progress. Thanks for filling us in, Brent.

Fascinating company. And yeah, free blood tests out there.

Go get 'em. Also, death potentially a skill issue. Don't do it. Anyway, thank you so much. We agree, we'll agree to disagree on that. Have a good one. Thanks, guys.

We'll see if we die or not.

I hope you're right. Yeah, see you back here in 70 years. Goodbye. Cheers, Brent. Let me tell you about Shopify. Shopify is the commerce platform that grows with your business, lets you sell in seconds online, online, in-store, on mobile, on social, on marketplaces, and now with AI agents. And our next guest is Matej from Adrienn. How you doing? Welcome back.

Hey, good.

Thanks for having me on.

How are you guys doing? We're doing well. Great to see you. We met at a Long Journey dinner forever ago at this point.

I think you've been on the show. We did. And we also met at the Ramp dinner. I think I was the infamous founder that snapped that phone, if you guys remember that.

That's right.

Oh, I missed that. We told this story.

I'm not going to say the guy's name because he's a friend and it's funny and it makes him look really silly. But this guy is sitting there, he's got the new iPhone Air. He's telling everyone in the room, you cannot possibly break this phone. This phone is indestructible. Anybody here try to break my phone. Try to break my phone. This guy goes, okay, just bends down, just snaps it in his hand. And the guy's standing there just being like, just in shock and disbelief. I thought— I was told— and he was saying, he was saying, uh, I'm gonna email Tim Cook about this. I'm pretty sure that Apple never promised you—

they never ran an ad that was like, it cannot be broken by a guy from the Gundo. So you are in the Gundo? Tell us That's what you're building?

Yeah, we're, uh, we're in the greater South Bay Area, I guess you could consider it the Gundo. I mean, we're super excited to announce our Series A, but how much did you raise? Uh, we raised $18 million. Yeah, Indrenum builds scalable sensing infrastructure for undersea domain awareness. So think about like Hunt for Red October, finding submarines, unmanned underwater vehicles, semi-submersibles, and so on.

So, yeah, I'm gonna maybe put it too plainly, but we don't really know much of what's going on out in the maritime domain. Is that correct?

Yeah, I mean, the maritime domain is getting more and more contested, like, daily, right? Uh, I think we have a good understanding of what happens on the surface of the oceans. There's cameras, surface boats, but everything's kind of shifting into the undersea quite quickly, you know. It's one of the most unexploited gaps that we have. And a lot of near-peer adversaries and terrorist organizations and so on and so forth are starting to move under the waves. So like this entire push of, you know, the US, for example, funding unmanned underwater vehicles, our adversaries are doing the same thing. And so the big question is like, how do we find those things intelligently, quickly, give ourselves awareness of what happens around, you know, various parts of the ocean as well as like our strategic bases and so on.

Makes a lot of sense. Talk about traction since the last time we chatted. What kind of progress have you guys made? Obviously the new funding, I'm sure you have some great momentum, but what's been, what have you been actually executing on?

Oh my goodness, so many things. So I think from the hardware and software side, like, I mean, we've gone through 3 generations of hardware since we last talked. We've been moving at breakneck speeds here at Andronom. We fielded it across like numerous exercises. Most recently, we just got back from a 3.5-week deployment with UUV group up in Keyport, Washington. So they're all the guys that run and like dictate how unmanned underwater vehicles run. So we were up there alongside with a few other folks, no mission-ending failures for a 3.5-week deployment. Obviously constantly testing here off the coast of California, building up our repository inventory of sonar signatures. And we do have some things that are going to be coming up quite soon that we're super excited to announce when they hit. So stay tuned.

What does integration with other defense tech companies look like now or in the future? Like, you know, there's been like Anduril has a platform, Palantir has a platform. I'm sure a lot of different units have bespoke systems. Systems for command and control and all sorts of different things. Does that come for free? Is that something that people are just like vibe coding integrations or do you have to go and like do a deal and actually figure out how you integrate or do you ultimately want to be, you know, your own platform that, you know, runs as like the operating system?

Yeah. I think it vastly depends on the type of technology that you're building, right? Specifically for us, like sonar information is so classified and compartmentalized and it's, it's very tricky to get access on data that has been compartmentalized such as this because you have numerous third-party integrators that also own that data itself as well. And so that is really the reason why we went out and built Pearl, which is our passive sonar buoy, right? As well as Obsidian, which is kind of like you could call it a C2, but really it is like the platform platform that ingests all this information. There's a bunch of like super sophisticated machine learning and effectively putting dots on a map for end users. But the C2 in itself, like there's various abstractions of it. You know, you have the god C2s that kind of live on top. You mentioned those companies' names. Yeah. We really are trying to own the undersea domain. And by building the hardware, we're building up this massive repository of sonar signatures, and that'll give us the option to then give that data to those other C2 platforms if necessary, right? So it really is like a fine, like a fine balance of what we have to be doing with regards to C2. But ultimately, like every defense tech company that's out there needs to play well together because we all have to provide an integrated solution to the end user, right? And we do have those relationships with primes and with other kind of startups as well. Yeah.

You mentioned that the data's hard to get generally. You're doing machine learning, which is obviously very data heavy. Is there like a Scale AI, like a Mercore for this type of data yet? Are you buying data from fishing boats? Or is there some world where there is a scaled data benefit, even not in a military context, but maybe you just want to get a bunch of data on how sonar performs in a swimming pool or a lake or something that's more controlled, and then you can transfer, learn that into like the actual viable domain in the final application.

Yeah. Sonar is super, super challenging in the fact that you can't necessarily translate environmentals to where you are deployed, right? So, like with radar, for instance, you just have air and it's pretty linear. Like the ocean does a ton of weird things with regards to performance of sonar. And so you really have to be deployed and you have to understand kind of the environmental factors such that you can perform this highly complex math that goes behind all of it. That's really the reason why we did build Pearl, right? Like, that delivers that information in real time. Like, we kind of get this information in sub-3 seconds worth of latency and we process it and we put dots on a map for end users. We have that environmental feature kind of built into Pearl itself. And you can do base-level extrapolation kind of across the board with a lot of that, but it's not super, super applicable everywhere. And so, yes, there is data that we have accessed that is more off-shell, for instance, like how do whales sound, right? Various different types of whale things. And at least in our, you know, kind of hot dog, no hot dog paradigm, we can say this is a whale, this is not, you know, a submarine or something like that. So it definitely adds a little bit more kind of feature-rich environment for the end user. But most of it is really hinged on the hardware that we're developing and building, right? So in this world of AI, you can't just access a massive repository of data and do a ton of AI on it. You have to build it yourself, which is really like the reason why a low-cost, scalable hardware solution is super, super important for us to deliver that data in real time.

What does actually deploying a buoy look like? Do you have to call the government before you do that? Because I love the idea of moving fast and doing R&D and just dropping something in the water, and that's probably fine. But at the same time, like, you could accidentally be spying on like a US submarine and they might not want you to be spying on them, even if you're like a friendly, right? So like, what is what is the flywheel to get the permission to, you know, do the test that you need to do before you get the contract to prove that you can do it? Like, what's the sequence of events?

Yeah, I mean, we're working hand in hand with a lot of these folks at the government, right? And so I detailed the deployment in Keyport, Washington. That is something that we scheduled prior going up there. You know, we do a lot of verification and validation testing here off the coast of California, but it's kind of just like put it in the water, see if it works, check it out, pull it out, and then ship it to the location with the end user and test alongside them. We've done that kind of across the coast, right? So we were in San Diego at a large fleet exercise at the beginning of this year as well. And so it does require coordinating with end users, right? But that's ultimately what we're looking for anyway. Like, what are the feedbacks that these folks give to us such that we can iterate much more quickly? We could do that in our own engineering vacuum where we could just be talking talking to these people directly, and that's what we really want to strive to do. And we've gotten really positive feedback from them, right? Like other people build buoys, they're very large. It takes a lot of effort to put these things out, like gigantic research vessels, you know, barges. It takes a week to pre-deploy. We just kick ours off the side of the boat, right? And so when we give these things, these buoys to the end operators, we say, break them, test them, go deploy them yourself. Give us the feedback such that we can put that into our loop of iteration. And they do that. They're super happy to break things. Right. And I think that's the most valuable thing that we can ask for.

I love it. Well, congratulations and thank you. Awesome progress. We'll talk to you soon.

Great to see you. Thank you so much. Keep snapping iPhone. I was going to say don't snap anymore iPhone Airs, but keep it going. Keep it going. Anyone has an iPhone, wants an iPhone snapped? Well, your guy, they're saying wants that.

Who wants that?

Sometimes I want my iPhone snapped. Diego said John needs someone to clear his desk of Diet Cokes every 20 minutes. Can I get a recycling bin over here? You're only on your third. Yeah, I'm all— This is a typical, typical day for you. Well, it's an important day.

We have Justin from NVIDIA, he's the VP of Enterprise AI here to talk about NVIDIA launching the Open Source AI Alliance. Justin, thank you so much for taking the time to come talk to us. Yeah, thanks for having us. So can you start by giving— I mean, we discussed the letter. I think a lot of people have read it, but what is your elevator pitch for the OpenAI Alliance, the letter, how they interact, like the overall project here, the goals?

Yeah, I think let's just say there's been this moment in the industry, especially with this incident that happened at Hugging Face, where there's just this collective awareness across all the industry leaders that, you know, doing a lot of the work in cybersecurity out in the open where we can look at open models, open harnesses, open safety evaluations as an industry is going to give basically the broadest collective defense of our organization. So that's why we put the letter out there. That's why all these companies are kind of rushing in to talk about sort of the importance of of this initiative?

It seems like it's going extremely well. I, I totaled it up. I think $18 trillion of market cap has signed on across something like 35 companies and more. I'm interested in what you— are there any other outcomes, or is this purely just avoiding a bad outcome? Because the whole framing is, is basically, uh, don't ban, don't do something. It's if we don't take an action here, allow the status quo to continue. Is that the correct formulation of the, of the proposal?

Well, I think it's a recognition that you always want to give the defenders of enterprises across really critical industries— financial services, energy, telecommunications— an asymmetric advantage to protect their infrastructure. And the best way to do that is you give them, you know, state-of-the-art frontier open systems that they can scan, remediate, patch their infrastructure. Now, they should always have the advantage because they understand the code, the configurations, and how they have their infrastructure running, but they need access to the tools to basically get ahead of any potential security issues.

Okay. I guess like the steel man of the other side is that This isn't an asymmetric advantage to the, to the organizations that you described. It's actually a perfectly symmetric advantage. The attackers will also have frontier open source, like great open source tools. And if we, if we allow open source to proliferate, at least it will be balanced. But can you explain a little bit more about where the asymmetry on the defender side comes? Because that sounds Amazing. We want the defenders to win, but it feels like if everyone's using the same tools, it's actually just a level playing field. And this could just be a battle, could still win, could be fine. But what's driving the asymmetry?

Yeah, you got to remember when, when you're running a service and you have terms of service, like, you know, who's, who's accessing that service. Yeah. And there you're putting additional guardrails on there on, on safe use. You can also, if there— if people start to misuse the service, uh, you can, uh, quickly patch guardrails, uh, and basically ensure that people are using, uh, the service appropriately. So, so that in, in and of itself should give the defensive teams, uh, the ability to run, I'd say, these, these better capabilities to clean up their infrastructure, patch their infrastructure. And then service providers that are hosting any of the models, um, you know, should, should run them with appropriate guardrails.

How do you think the equilibrium plays out if sort of, you know, companies get like big companies get access to advanced open source models. So they're patching their systems, they're doing stuff, but there's still a lot of attackers out there. And there's this weird scenario that people have been talking about where a small business that might have a credit card stored here or some piece of personally identifiable information stored in a less secure system. Now they're responsible for spinning up an open-source cybersecurity defense system. Like, what's the equilibrium that keeps the long tail of organizations secure?

Yeah, you've got to still pair, you know, openness with strong safeguards, like rules on malicious use. Sure. Laws that prevent against malicious use. And then obviously rapid remediation of anything that, you know, that's going on. And, you know, that the, the conversation around open models is no different than the conversation around— I mean, I think open software. Yeah. And you think of open software, whether it's Linux or Kubernetes or PyTorch, that, that open software, you know, built the foundation of our economy. 80% of our digital economy is powered by open source. And I'll say good stewards of open source software can sort of debate open standards in an open way. They can look at evaluations, they can look at patching and remediation, and they can work collectively for the defense of the US economy. And I think we got to look at models the same way. Yeah. You know, as soon as models are out there in, I'll call the public domain, you can fork them, you can patch them, and you can remediate any issues that, you know, cybersecurity people find in those models. And so, you know, the goal of this organization is to have those conversations out in the public. I think it's, you know, it's an important conversation for all of us to have. But we want to make sure, again, that you, you have, I'll say, a diverse set of tools at your disposal as a cybersecurity defender so that that, you know, if a model refuses to do the work that you need it to do, you have other options to continue to defend your organization.

Who do you think will step into the role of model evaluator or model tester? Because we were just talking with Dean Meyer from Sequoia Capital, and he was illustrating a possible scenario where there's like a backdoor. In a model or some nefarious thing sort of hidden below the surface. So, the model still tests well on benchmarks, you deploy it, and then over time it does something nefarious. It feels like, again, that's another big burden to potentially put on a small business. But if there's some sort of, you know, American company that steps up and validates and tests, maybe there's an even balance there? How do you see that playing out?

Well, I think, you know, organizations like Red Hat have done this open source software, right, where they cleared of CVEs, they provide the commercial support to enterprises. I feel like it creates an opportunity for businesses to step in and be able to do the same thing with open, open models and provide the same commercial patching and support into, into enterprises. So I think that's a market that will emerge as we go forward. But, but ultimately, I think allowing, I'll call it cybersecurity professionals to continue to pen test and understand the limits of these models is what's going to allow us to, you know, build that lifecycle to do patching and remediation of any vulnerabilities that can exist in these model weights. Now, the fact that they're open means you can test them and understand the limits of them. The challenges with a lot of the the model weights that might be hosted, the terms of service prevent cybersecurity people from researching for and looking for any backdoors in those as well. So the problem doesn't go away whether it's hosted or open. Sure, they can exist. And I think the spotlight of having more cybersecurity professionals, you know, evaluating these things is just going to lead to better collective defense.

I don't know how much you can speak to the— to what the Nemo team is doing, but broadly, what should organizations expect from, from Nimotron, uh, the family of models over the next, uh, let's say 6 months?

Oh man, I, uh, I don't know that, uh, I can talk about where we're going, but, you know, generally our approach has been, um, has been like, let's again do it all in the open. Let's, uh, basically put the data sets out there that we're training into these models so can understand the data that blends that went into the models. Let's put the model weights out there. Let's put the training techniques out there and the evaluations out there and basically use that as a foundation for building, you know, custom intelligence into enterprises. I mean, that's the other part of the conversation that's lost is these open models are critical for enterprises basically building their own intelligence and basically maintaining— I'll call it a reinforcement learning wheel or a flywheel on these open models so they can maintain their competitive advantage. And so, providing that foundation of data and open weights and data blends gives enterprises that. Now, with Neomotron, they're really language models, but we have a bunch of other domains and models. We have, you know, Cosmos, which is World Foundation Models. We have Alpha Mayo for autonomous driving. We have GROOT for vision language action models for robots. And so I think across industries, the idea of having these open models to build the startup ecosystem, allow enterprises to domain adapt them and build their own intelligence is really important, really for the, for the future growth of our industry.

How important are— can you give us some context on what you're seeing or hearing from other business leaders about the value of open source purely economically? Just— we went through this token maxing boom. I think every enterprise leader feels that AI is valuable, but it can be very expensive. How are they thinking about efficiency right now with the current cost-to-benefit tradeoffs? And how does open source change that?

I mean, I can, I can put my NVIDIA hat on here. Uh, we, we are the same way, right? We use all the frontier models. We love the frontier models. Yeah. Uh, they are, they're amazing, uh, for, for driving efficiency and how we do, you know, software development and chip design and how we run a competitive business. Um, but you, the, the modern way to build applications now are really an ensemble of models where you want to use the frontier for where they're unique and they add differentiated value. So like a really complex agentic plan, you want to run through the Frontier model, and then you want to use the open model weights for things that are more trivial, like document summarization. And the two can complement each other to drive a lot more cost-efficient applications across the enterprise.

Yeah. And then anything that's been fine-tuned on your particular data, that example from Thinking Machines with Bridgewater's information pipeline. Obviously, they got incredible performance at very low cost because it was a unique use case that was going to run on a continuous basis. And so they were able to get a lot of value there.

It's just common sense. It wouldn't make sense to hire you and then have you just be a BDR doing cold email. It wouldn't be a very efficient use of—

It's a great analogy. Sorry.

Yeah, sorry, I was just going to say, yeah, I think, um, I think your point's spot on. I think, um, you know, there's a bunch of other, uh, you know, model makers, uh, like, like TML is a great company, uh, Reflection, um, you know, uh, Runway, uh, you know, many different model makers across different domains. And I think you want a rich ecosystem here, uh, you know, of startups, um, you know, building the future of what's possible with with this industrial revolution going on with AI. So we're just happy to advocate on behalf of the open ecosystem, but really it complements the closed ecosystem and all of it sort of has its unique value here.

So the alliance is pretty huge at this point. Do you want to grow it a lot bigger or is this a solid enough base to go where you want to go next? Or will this be something that just continues to grow? What do you think?

You know, I think the inbound requests around this is kind of blown us away by the appetite and the interest. I think ultimately what we want to do is kind of work as a community and share sort of, you know, best practices. We also, with this, we put out a new agent harness that we think is phenomenal at, you know, reasoning. It's great at coding and it's great at cybersecurity tasks. And so if we all sort of share advancements in the open, we can build the best technology together. And I think some of the things that we'll keep working on as an ecosystem is, I mean, to your point, what are the evaluations that we should be building? How do we test and I'll call it open models? And look for, I'll say, safety issues in those models? How do you start to build, I'll call it, more cyber ranges that you put agents through to understand how truly capable they are? I mean, I think part of the challenge with this cybersecurity discussion is it's been over-indexed on, I'll call it, you know, benchmarking for, you know, vulnerability discovery and vulnerability discovery is not the same as, you know, like going through an attack chain. So, you know, I think we've just got to have more, more sane conversation in the open as an ecosystem on, you know, what are the right ways to safeguard this stuff and put it to work and to drive the efficiency that it can bring us, you know, across every business.

Thank you so much for taking the time to come.

Yeah, great to meet you, Justin.

Great to meet you. Yeah, thanks so much. Have a good rest of your day. Goodbye. That's our show, folks. Thank you so much for tuning in. We will see you tomorrow. There's a lot more news, but we can go through it tomorrow. Uh, Starship launched, and there's a beautiful video that somehow got taken down, so we can't even share it. But, uh, you can go find it. There's some cool stuff there. Uh, anyway, go have the best afternoon.

Have the best afternoon ever. Have the best Monday afternoon ever. Thank you to Cooper for always coming in at the market close.

Good crew. Get in the chat. Thank you so much for tuning in and chatting. We do read it. It's a lot of fun. Leave us 5 stars on Apple Podcasts and Spotify. Sign up for our newsletter at tbpn.com and we will see you tomorrow at 11 a.m. Pacific. Goodbye.

We love you. Goodbye. Goodbye.