--:-- John Coogan: You're watching TBPN. Today's Tuesday, November 10th, 2026. We are live from the TBPN Ultradome, the Temple of Technology, the Fortress of Finance, the Capital of Capital. We have a great show for you today, folks.
--:-- John Coogan: First, let me tell you about ramp.com. Time is money, say both. These are used corporate cards, bill pay, accounting, and a whole lot more all in one place.
--:-- Jordi Hays: The stock market is crashing. Where did it all go wrong? That's from the chat. Thank you, Goldrock. I like that one.
--:-- John Coogan: No, we're talking about Wong, Alice Wong, because the spy gate continues. It's getting worse. There's a whole bunch more allegations. It's very dramatic. And she's going on the naughty list. And the good thing is, if you think there's no way it could get worse, well, that feels like it's on the table.
--:-- Jordi Hays: That is on the table.
--:-- John Coogan: But yes, let's pull up the naughty list because we put Jensen on the naughty list, but I think Alice has to go higher on the naughty list, because she's the one that is the spy and she's working for the geopolitical rival, the near peer adversary. She's working for China. She started it, but it's getting crazy.
--:-- John Coogan: There are seven new accusations. Some of them we touched on. Some of them we haven't. I can kind of take everyone through this. First is that Nvidia may have interfered with a federal investigation. Commerce Department investigators were investigating Megaspeed for potentially smuggling NVIDIA chips into China. After NVIDIA General Counsel contacted Commerce, investigators faced unusual restrictions. They needed to route NVIDIA inquiries through a senior political appointee and avoided references to Megaspeed, which is very odd.
--:-- John Coogan: Two, NVIDIA intervened regarding Alice Wong's detention. General Counsel Tim Teeter called Commerce Undersecretary Jeffrey Kessler after Singapore detained Alice Wong. The question is whether Jensen was leveraging NVIDIA's influence to help her.
--:-- John Coogan: Also third, NVIDIA undercounted 50 GPUs sold to Megaspeed. NVIDIA told regulators that its inspectors had accounted for the chips, but Bloomberg identified approximately 50 additional GPUs worth billions of dollars. NVIDIA said it had inspected additional warehouses.
--:-- John Coogan: Fourth, the FBI, Congress, and NVIDIA's board are now investigating. Trump said the FBI was looking into Wong, but it was unclear about exactly what. It might be a different thing. NVIDIA's board opened—
--:-- Jordi Hays: The way the quote from yesterday made it sound like there's something else.
--:-- Jordi Hays: But there's so many things. It could be within that or separate or linked to it or something, but never want to be on that side of the phone call with the FBI.
--:-- John Coogan: And so NVIDIA's board has opened its own investigation and lawmakers raised questions about what other NVIDIA executives knew.
--:-- Jordi Hays: So I'm actually unclear on whether this will wind up with a testimony on the Hill, like a congressional inquiry, like the famous Mark Zuckerberg, "we sell ads" moment. Is that how this will play out or is it all behind the scenes?
--:-- John Coogan: Absolutely.
--:-- Jordi Hays: Absolutely. There's just other ways to deal with it.
--:-- John Coogan: So the entire country, for the most part, except that one place in North Dakota, voted against AI.
--:-- John Coogan: What do you think their representatives are going to be doing come January?
--:-- John Coogan: They want to – Jensen is one of the faces of AI. He's the most outwardly supportive of the administration.
--:-- John Coogan: They will use him to put on an absolute show, right?
--:-- Jordi Hays: Yes. A performance.
--:-- John Coogan: Yes. The exact details of that and how it goes down I don't think are obvious yet, but he will be starring. They're going to throw the book at him. It's going to be the Girardian scapegoat. Someone must fall just like Sam Bankman-Fried sort of was the power law recipient of negative attention post-crypto collapse. This could be a lot of attention on Jensen, less attention on other people in AI potentially.
--:-- Jordi Hays: But my question is if it really is like a level of national security, those might be just closed-door hearings or closed-door investigations. It might be like there's still a conversation and you're under oath, but it's not going to be televised on C-SPAN because—
--:-- John Coogan: I'm sure they'll figure out a way to televise some of it.
--:-- Jordi Hays: Yeah. No, I think you're right. I would just probably handicap it at like 70 percent. I don't know.
--:-- John Coogan: We were talking with a friend last night. He studies history, greatest entrepreneurs. He doesn't really know much about the news, but we were sort of explaining what happened to him. And I was like, is this the largest unforced error in business history? And I think that it might be.
--:-- Jordi Hays: No, no. I was going back to Rockefeller, Henry Ford. There was never a crisis of geopolitical consequence quite like this. It really is uncharted territory. There's been plenty of business scandals in history. There's been political scandals in history. But we were even going back to Clinton, and we were like, that did not reshape the geopolitical dynamic at the time. It was not like, okay, all of a sudden, Japan had more power than America because of the Lewinsky scandal. That was a separate thing.
--:-- John Coogan: I don't know. I just think I don't know enough about that to have a strong stance, although my assumption would be that the president—
--:-- Jordi Hays: Notical that he was an American.
--:-- John Coogan: Yeah, I know. I know. But like having a president be disgraced obviously has— It was a huge, huge story. But just purely from a business standpoint, a screw-up of this magnitude with these kind of implications, eviscerating hundreds of billions of dollars of market cap. I don't see how, knowing that there's just a lot there, it doesn't feel like we're hours away from a positive headline. It feels like we are months away from, we're years away from potentially a real resolution to all of this, right? It's going to be this massive overhang on the business.
--:-- John Coogan: And it has been interesting to track people's perception of this whole thing from Sunday. People are like, okay, wow, he admitted to having an affair. And then it's just got worse and worse and worse to the point where I think most people, even if you believe that Jensen is the right person to run NVIDIA from a value creation standpoint—
--:-- John Coogan: It's hard to imagine him staying in the seat.
--:-- Jordi Hays: Yeah. I mean, in terms of... It's just too... Unless there's some... What the opposite of a smoking gun? A cold gun that has not been fired? Help me out, Tyler. Opposite of a smoking gun.
--:-- John Coogan: Hmm. A cold gun that has not been fired. Help me out, Tyler. Opposite of a smoking gun. What do we got?
--:-- Jordi Hays: Insufficient evidence.
--:-- John Coogan: That's terrible. That sounds horrible.
--:-- Jordi Hays: Yeah. That sounds horrible. What do we got? Chat? Help us out. Muddy waters. I don't know.
--:-- John Coogan: Yeah. It's a question like the other. Everyone's waiting for the other shoe to drop. It feels like there might be 10 shoes here. There's certainly more. I didn't even get through the list.
--:-- Jordi Hays: An armory of untouched guns. That's the best I can come up with. Armory. Look. None of them have ever even been fired.
--:-- John Coogan: Yes, yes. They're all factory new. Never been fired. Anyway. So there's three more. Three more things.
--:-- Jordi Hays: There we go. Dan. A nothing burger. A nothing burger. If he's got somewhere. He's like, here, look. It's a nothing burger.
--:-- John Coogan: And video's general counsel was a stone cold alibi as of last week. So this raises the possibility that, oh, we got more. I get out of jail free card. If he's got might be on the table. If they have one of those, it's going to be controversial. But so there's a question ability that also sounds bad.
--:-- Jordi Hays: Yeah, that's iffy.
--:-- John Coogan: Yeah. So there's a question right now about did the senior executives knowingly misrepresent the relationship or were they misled by Jensen? Also, Jensen had quietly hired crisis PR firms a year earlier. This was from Wired. They reported that Heller had been managing his behind-the-scenes communications since at least 2025.
--:-- John Coogan: And Tyler, you hired the crisis PR firm Heller recently, right? Because you were trying to get in front of the skipping leg day allegations, is that right? Because that could destroy you if it gets out that you've been skipping leg day.
--:-- Jordi Hays: Certainly not me, I would never. But yeah, if you've been slacking at the gym and you're skipping leg day, you gotta give Heller a call. It might look bad, but better to have them in your court than go it alone, certainly.
--:-- John Coogan: Anyway, the potential cover up creates personal liability for NVIDIA's directors. This is a big deal. There's already a lawsuit filed, but if Jensen concealed the relationship from the board while Nvidia publicly characterized it as ordinary business, directors could face more shareholder litigation potentially, depending on how that breaks. So there's a lot of complications there.
--:-- John Coogan: Also, the Information has a fantastic breakdown that goes way deeper into Alice Wong, the James Bond of AI, I suppose. Megaspeed. And then there's this new company that they dug up, Century Litong. Highly recommend reading this. And they actually pieced together the puzzle a lot more. I was trying to show the connections between these firms. Remember I was saying like, okay, Megaspeed is related to Seven Road Holdings. And then there's Alice Wong's Holdco that's in the BBI's, that whole thing. Remember that confusion?
--:-- Jordi Hays: Yeah. You're like, I was tuned out for that. Most suspect.
--:-- John Coogan: It was extremely suspect. Corporate structure all the time.
--:-- Jordi Hays: Yeah, you shouldn't be able to raise a seed round if that's your structure. It's a more complicated structure.
--:-- John Coogan: So there's Shanghai Century Litang, which was the founding sole shareholder, exited June 2024. Then that's Shanghai Century Long Tang, which is indirectly owned until June 2024. Then Shanghai Yunxi Investment, which is owned 99% until 2017. It's insane.
--:-- John Coogan: Also, I didn't know Alice Wong is married. And so they're both actually married. Alice Wong is married to Kang Wei, who's the founder and major shareholder of Shanghai Hexi Capital, which was a shareholder in Zhejiang Century Watong, which is a shareholder in Seven Road Holdings Limited, which formerly owned Megaspeed. So her husband was a shareholder in Megaspeed through like five layers or something.
--:-- Jordi Hays: So I, and I don't even know, is he stoked about this? He's like, good job. You honeypotted successfully. Like that was the name of the game. Or was it like, oh, I wasn't expecting you to honeypot like that.
--:-- John Coogan: It's possible. The marriage was like part of the whole corporate structure. No one knows.
--:-- Jordi Hays: Yeah. No one knows. I mean, have you ever seen The Americans? I was just, I didn't know. And in the show, it's about two Russian sleeper agents that basically build a whole life in America. And as part of that, they are honeypotting a lot of different people. But they're in a relationship, but they're both agents, so they willingly go out and honeypot people all the time and then they come back and they're like, okay, how it going? And then there's some jealousy that creeps in. It's a great show. You should definitely watch it if you haven't. Anyway, sorry. Catch y'all.
--:-- John Coogan: Yeah, I saw a retail investor over on Instagram crashing out because they were a big NVIDIA holder. Well, probably not a big NVIDIA company, but a lot of their portfolio was NVIDIA. And they were talking about how they bought the dip Friday, had a great weekend, and woke up to discover that they weren't even referencing the CEO by name, right? Because I don't think they're that focused on it. But they were like, and now I find out that the CEO had an affair, but not just a run of the mill affair, an affair with an alleged spy.
--:-- John Coogan: Like, and anyways, it's just, yeah, they're calling it the greatest unforced error in business history. Fatality. Hopefully not. Hopefully not.
--:-- Jordi Hays: Is she the Michael Jordan of spying? You kind of got to give it to her, right? She's like one of the greatest spies of all time, potentially. A Hall of Famer.
--:-- John Coogan: Hall of Famer. But a true American tragedy. Really, really bad.
--:-- Jordi Hays: I don't think you can be a Hall of Fame spy if you get found out.
--:-- John Coogan: Oh, okay.
--:-- Jordi Hays: Like we know about her.
--:-- John Coogan: Yeah, yeah, yeah. If you're a spy, we don't know.
--:-- Jordi Hays: Sure, sure, okay. So a firm place on the naughty list, but not going in the spying Hall of Fame.
--:-- John Coogan: That's exactly what a spy would say, an active spy would say that wants elite spies to get more credit, but can't say, like, I'm the top guy.
--:-- John Coogan: Well, 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.
--:-- John Coogan: Going on the nice list, we have our first nice list addition, the United States Marines. Happy birthday, Marines. Today is the 251st birthday of the Marine Corps. And Jackson over there wrote a very nice tribute in the newsletter to his grandfather. Do you want to take us through it a little bit? Does he have a microphone or should I read it? You want to come back there?
--:-- Jordi Hays: Yeah, give us the high level. You don't need to read the whole thing, but give us a little overview of your family history. Are both of our hosts sick?
--:-- John Coogan: Yes. Yes.
--:-- Jordi Hays: How did you know?
--:-- Jordi Hays: How did you know?
--:-- John Coogan: I'm trying to hold it together so badly. We're down bad, but the show must go on.
--:-- Jordi Hays: Must go on. So, Jackson, take us through the story of Michael Fordyce and your grandfather, Tom.
--:-- Jackson: Yeah, my dad joined the Marines right out of high school. He went to the Middle East, was in Gulf War, served in Operation Desert Storm. My grandfather also went to the Marines out of high school, was sent off to Vietnam, fought in the Battle of Quezon, which was, I think, one of the most grueling and bloody battles of the Vietnam War. Got his arm shot off, unfortunately. And he has, he explained it to me today. It's called a myoelectric hand. So he thinks and it moves and picks up things.
--:-- John Coogan: Interesting.
--:-- Jackson: He actually has like the Luke Skywalker arm.
--:-- John Coogan: That's incredible.
--:-- Jackson: I asked him how it works. He says, I don't know. It just does.
--:-- John Coogan: That's amazing. I love it.
--:-- Jackson: But that's it. They're awesome guys. I love them a ton and I appreciate their service.
--:-- John Coogan: Fantastic. Yeah. Thank you for their service. We love them. And happy birthday to the Marines going on the nice list.
--:-- Jordi Hays: We got to go hunting for some more nice list additions because we're kind of overweight, naughty list here. And I feel like in general, Santa's list should be 70% nice, 30% naughty at most.
--:-- Jordi Hays: Right. Or two weighted. So we got to dig in. We got to dig up some good, maybe Jackson, maybe Jackson goes on the nice list. He's so good, but also maybe should be on the naughty list. Cause he kind of broke the family tradition of serving out of high school, podcasting at war.
--:-- John Coogan: Yeah. Men used to go to war serving this country, via now they podcast.
--:-- Jordi Hays: I don't see it now. Your whole family's on the nice list.
--:-- John Coogan: Yes. And we're grateful to have you here.
--:-- John Coogan: Well, we have Chris Miller joining at 11:45, the author of Chip War, returning to the show for the third time to take us through more of the geopolitical implications of the NVIDIA situation. Ed Bolian from VinWiki. Then we're going over to Car World, baby, talking about the cannonball. He has held the record for fastest cannonball. And then, of course, we're going into fundraising news and closing it out with Christopher Kircher.
--:-- John Coogan: Let's dig into more posts from the timeline. Amir over at the Information gives more context on Alice Wong. Says forget the role of Megaspeed, forget the word Megaspeed. Megaspeed doesn't matter. It almost a distraction. Her real story is her role at China-based Shanghai Century Litong, which I mentioned before in that very confusing chart, which became a preferred NVIDIA cloud provider or partner in mid-2023 and facilitated over $7.5 billion in NVIDIA chip shipments, including securing banned chips.
--:-- John Coogan: Remember, the CHIPS Act, that's Biden era, pre-Trump, well, after Trump won, before Trump too, the CHIPS Act restricted the A100, the H100. And so eventually NVIDIA came out with the H20, which was sort of the nerfed version of the flagship chip. But Century Leetong was still able to get the unnerfed chips, the best that NVIDIA had to offer, exactly what the government didn't want to have happen.
--:-- Jordi Hays: And now there is a world where you get duped and you accidentally sell something to someone who is exporting it and going around export controls. And that would be the opposite of the smoking gun, the nothing burger, right? If there's clear evidence that he was duped along that way and it really was like, well, I went to their facility in Shanghai. Or not Shanghai. Sorry. Singapore. I went to the facility in Singapore. I talked to the companies that were there. They were not involved in the CCP or in China at all. And so it was totally fine. And, like, I had done all the checks and we had all of our teams check that the GPUs were staying there. And they were not going to be used internationally.
--:-- John Coogan: I know. Alice was detained by the Singaporean government. We have a lot of evidence that it's worse than that, but that would be the best case scenario, given what we know. Even if they were originally tricked, there's been a year since that detention where they could have been front-running this story.
--:-- Jordi Hays: Yeah, not good.
--:-- John Coogan: So Alice Wong and her sister and her husband have built a labyrinth of companies spanning China and Singapore to acquire NVIDIA chips, according to people who know Alice Wong and corporate registration records. At the center of that corporate network is Shanghai Century Litong, a China-based company that was found in June 2023 and by October was added to NVIDIA's list of preferred cloud partners. That's fast to start a company and rocket and actually get to NVIDIA's list of preferred cloud partners.
--:-- Jordi Hays: And by October was added to NVIDIA's list of preferred cloud partners. That's fast to start a company and rocket and actually get on NVIDIA's radar. I feel like you talk to startup founders all the time who are like, I would love to get a meeting with a big company, but I'm six months in. They won't answer my emails. Not so in this case.
--:-- John Coogan: So there's other news. When a shipment of banned NVIDIA chips could not be obtained from NVIDIA's distributors, Alice Wong stepped in. One of the people who knows her said her role was to secure the chip orders when they got stuck in the supply chain. She told some colleagues and customers that she could always get the supply, citing her close relationship inside NVIDIA and with distributors. Another person who knows her said.
--:-- John Coogan: Anyway, where should we go next? To Matt Levine or over to Negligible Capital.
--:-- Jordi Hays: Imagine telling someone six months ago during the SaaSpocalypse that Microsoft will have a greater market cap than NVIDIA by November. It is remarkable that NVIDIA was actually the biggest company in the world right as this was all happening. So it's literally the target number one for a geopolitical rival. There's no other CEO that you would go after. It's truly the most important. And here we are.
--:-- John Coogan: Here we are. But I guess...
--:-- Jordi Hays: Yeah, that's why it's ultimately... No, it's... That's why it's a black swan. It's a black swan, for sure.
--:-- John Coogan: I was... Two weeks ago.
--:-- John Coogan: I was racking my brain. What could the black swan be? Trying to think about what... And of course, that's the nature of these things. But like, you know, I don't see how the question is like, can Jensen while he's under investigation from multiple groups, continue this sort of deal making spree, or does a lot of that have to just hit pause?
--:-- Jordi Hays: Right. And part of, you know, people talk about moats all the time, but like moats tend, you know, a factor if, if, you know, if people are willing to be honest and moats are just like, you know, the people running the show and the overall talent density. And so much of NVIDIA's moat was just Jensen's incredible tenacity and work ethic and basically schemes. When you look at the NeoCloud space and how he's approached those relationships and supplier relationships, the sort of Jensen discount that you have to put on NVIDIA, even the possibility of it is like extreme.
--:-- Jordi Hays: And then the impact for the private markets is like, I think a lot of VCs were basically like, you know, shoot for the moon, but even if you miss, like Jensen will give you another up round or he'll, you know, buy your company. I mean, even I was sitting here looking at CoreWeave as it's been beat up, being like, there's no world where this company fails because Jensen, I don't think will let it happen. And there's so many more questions around how he'll approach running the business.
--:-- John Coogan: I can imagine as long as he's CEO, he's probably going to stick to his guns and keep doing what he's doing. But it's just like his entire strategy of being everywhere all the time, just being again like he's not actually like Switzerland in practice. Right? Like NVIDIA is very aggressive, wants to do everything it can to win, but generally being like I'm going to fund everyone. The sort of like Switzerland, sort of like neutral approach. He just wants AI to grow and he wants people to buy NVIDIA GPUs, and if that's happening he's happy. He's going to support all the Neo labs. There's like a ton of Neo labs that have raised like a billion dollars.
--:-- Jordi Hays: Right. And a lot of them now have to be, they were hoping like, if we can just prove out that we are on the path to making something commercially viable or interesting, like we can probably count on NVIDIA for like half a billion or a billion dollars or to anchor a round. And it just like, as we've gotten into this period where it feels like capital is the primary constraint to the build out, right? A lot of people want to build data centers. A lot of people think that there's going to be exceptional demand for compute regardless of the midterms.
--:-- Jordi Hays: And if capital is the, if we're all realizing like, hey, you know, the rates that developers and everybody in the sort of seven layer cake are having to pay now to finance their own CapEx, whether they're on the bottle, the sort of AI infrastructure side, all the way to the real estate and power shell side. But one of the biggest financiers of this entire build-out is like effectively you can't lean on them anymore in the same way that you could. And that's why I just think we have a major reckoning.
--:-- John Coogan: We got to do stimulus checks. We got to get the money printer going. We got to get the crank.
--:-- John Coogan: It is such a black swan. And it's interesting because we lived through a very crazy moment with situational awareness that still didn't feel like a black swan because there were so many people from day one being like, oh, this is not going to work or this is so risky. It's too good to be true. And that puts it in like it was surprising and it was a very rough moment and it was tough to watch and it was sad. But it was not as much of a black swan as this.
--:-- Jordi Hays: Tyler, you had a comment and then I want to know where the prediction markets are.
--:-- John Coogan: Yeah, that's what I was going to say. Please. So, up almost 10% from yesterday of if Jensen is out before July of next year. It's at almost 40% right now.
--:-- Jordi Hays: 40%, yeah. And yesterday it was around 30, 31. Yeah, I've heard a lot of chatter about people saying like he can't stay, but at the same time there's not a very clear succession list and it's a crisis. It is just a crisis.
--:-- John Coogan: Anyway, there is so much more NVIDIA news, but we should probably move on. The Anthropic news, there's a ton there. They're on track for IPO. They put out their prospectus and offering documents and videos, and they have a whole host of ways to interact with their data. And also Fable 5.2 came out and is doing very well in benchmarks and people are starting to show demos of that and that's very promising.
--:-- Jordi Hays: Really rough timing because Jensen is sucking a lot of the air out of the room. At the same time, it is sort of a vindication of their positioning that the geopolitical conflict with China is extremely high stakes, extremely serious, and will be taken to the extreme.
--:-- John Coogan: And they have a new 80 billion ARR number. I think there's a lot of... There's still a million reasons to be bullish on Anthropic. They also cut the valuation. And if you were bullish...
--:-- Jordi Hays: Yeah, the valuation's cut. You have a better entry point if you do want to invest. And so I think in hindsight, I think a lot of people will be like, wow, this was like sort of a golden opportunity and that there was so much bad news. And yet Anthropic's business was still working in the way that it has been. And that creates an opportunity for everybody participating. And my assumption is that retail will be able to get massive exposure at the IPO.
--:-- John Coogan: Did you see that on their website? They have a whole – they have a bunch of case studies of companies that are using Anthropic. But then they also have a section where it's like, how can you participate? And it says like, click here to know the Charles Schwab route, click here for the Robinhood route, click here for the E-Trade route. And they all have like, you can click your brokerage and like go through a flow. And so that felt, I don't know, it felt like down to earth or something to see. I mean, maybe it's standard practice, but it seemed like, I don't know. It was kind of like, it wasn't highfalutin. It felt very like, yeah, like we're—
--:-- Jordi Hays: Well, I think a lot of people's mental models is still like a company goes out and they raise $4 billion in their IPO. And there's just like plenty of big investment banks who will fund that. And now it's like, yeah, you can carve out $10 billion for retail or more.
--:-- John Coogan: Well, if you're thinking about going public, do it on the New York Stock Exchange. Want to change the world? Raise capital at the New York Stock Exchange. And also, if you're investing in an IPO, do it on Public. Invest across all asset classes. Automate anything with AI agents. Run trading strategies, risk management, money movements, and monitor markets with AI agents for investing on Public.
--:-- Jordi Hays: Bill Bishop is just showing the... The Bishopinator?
--:-- Jordi Hays: The Bishopinator is doing a bunch of pungentry. He says, NVIDIA execs ran a smear campaign against proponents of chip export controls, bullied people on the Hill. They made so many enemies, and now no one is afraid of them.
--:-- John Coogan: I think we are still in the tip of the iceberg phase of what is going to come out. He thinks that there's going to be many more shoes, a whole closet full of shoes dropping over the next couple months maybe.
--:-- Jordi Hays: Imagine they open the closet and like 300 nothing burgers just spill out.
--:-- John Coogan: Wow. Yeah.
--:-- Jordi Hays: Yeah, fire up the grill. We're cooking up a nothing burger.
--:-- John Coogan: I haven't seen anyone post the nothing ever happens guy yet. You got to post that.
--:-- Jordi Hays: It could be, yeah. You have one of the all-time worst nothing ever happens posts right before the start of the Iran war, I believe.
--:-- John Coogan: Yes, correct.
--:-- Jordi Hays: Wow. True L for you. Fade Tyler.
--:-- John Coogan: True L. But that's what you come to TBPN for, hot takes. Travis Kalanick chimed in with a vague post, sort of. I mean, everyone knows what it's about. He's talking about NVIDIA. But he says, had a little time on my hands this morning to do some writing. Send along to someone you know who may benefit. Practical principles to counter a modern-day coup, PSYOP, struggle session. Own what you did, but fiercely resist accommodating the mob. The op is not about your past actions. It's about your enemies, and this event is their opportunity. No matter how bad it gets, nothing is as bad as it seems. He's saying it's a mindset issue.
--:-- Jordi Hays: The interesting thing is going from where every single day that's passed, well, it's only Tuesday, the last couple of days, every single day it seems it's actually worse than I originally thought.
--:-- John Coogan: True. Yeah. It does seem to be getting worse before it gets better. But hey, nothing burger could be coming at any minute.
--:-- Jordi Hays: Well, it does seem like Anthropic is on track to IPO next Thursday, which will be very exciting. I believe SpaceX went out on Friday and didn't really start trading or moving around much later in the day. So we'll be covering it. It's potentially landing on GTA 6 launch day. So we're going to figure out the schedule, but it should be a fun day tracking how the market reacts to that. I'm pretty optimistic, but we'll see what happens.
--:-- John Coogan: Also, we're going to Tomorrow X next week and we're going to be talking to a bunch of the most qualified people, many Anthropic investors, many of the folks who are underwriting the IPO. So be sure to tune in next Tuesday and Wednesday because we have an all-star lineup to cover this IPO and understand what's happening both with the AI trade and Anthropic specifically. It should be a really great week of coverage. But of course, then we have to lock in for GTA because it is a historic moment.
--:-- Jordi Hays: First question for Strauss Zelnick. When is GTA 7 dropping? That's my first question. What have you done for me lately? We're not even giving you one hour to savor the moment.
--:-- Jordi Hays: What have you done for me lately? Give me GTA 7. I want a timeline. Lock it in. Commit. Commit to a launch date. Make some commitments.
--:-- John Coogan: Well, here we go. Amazon unveils a flying home security drone. Ring Robotics has two autonomous security devices, a $250 flying camera for homes, a $1,000 roaming robot for businesses. The Ring drone cam flies preset routes indoors, automatically investigates triggered alarms, and records 2K video. It's AI pilot maps rooms and navigates around obstacles. The Rover cam autonomously patrols businesses, investigates disturbances, and records up to 4K video. I think this is quite cool.
--:-- Jordi Hays: We have a security system in the office. The cameras are at very specific angles. I would love for a drone to launch when there's an intrusion. Does it ship with kamikaze mode? Is that an add-on that you have to subscribe to, or do you have to use a zero day and a model to root it and then side load your own operating system?
--:-- John Coogan: They should just come as a swarm. You should be able to get thousands of these drones, all coordinated. Because once you unlock kamikaze mode, you don't want to be limited to just one shot.
--:-- Jordi Hays: Yes. So, very cool. I think this will be maybe successful. I don't understand why it's indoors. Don't you want outdoor patrols, outdoor security? Don't you want the drone on your roof docked and then flying a little patrol every once in a while based on cameras? It feels like, why are you waiting until there's an intruder inside to launch this little drone that it can just swat away?
--:-- John Coogan: No, I mean, if you get an alert that there's been some sort of intrusion, I think your default assumption should be that it's not actually an intrusion. If you have a home security system over the course of a decade, you're going to get a bunch of notifications. Probably, hopefully 99% of them are like, somebody came home that lives there and didn't know it was armed. So there's going to be a lot of false positives. The drone is how you quickly determine, is this a false positive or not? Without having like, I'd rather have a drone. Indoor cameras. I'd rather have a drone than 20 indoor cameras.
--:-- Jordi Hays: Yeah, you don't want indoor cameras. When Nat Friedman said that he had a series of indoor cameras and he was able to get his open claw to kind of bully him into drinking water. You remember that at the Stripe sessions? I was like, that's very cool that he's such an early adopter of this, but I don't know if I want a bunch of indoor cameras. So having the ability to snap your fingers, put a camera anywhere in your house without needing to install always-on cameras that immediately become worse spyware than your smart TV potentially, that is very cool. But I want to see what these things can do outside. Can these things run around outside? Can you trick them into thinking they're inside when they're actually outside? Because the outdoor patrol seems like the thing to get. But maybe an opportunity for a startup.
--:-- John Coogan: It is funny the sort of contrast between these two things. The drone seems kind of sinister. It's the Matrix. We're living in a sort of a dark sci-fi future. And then the little home robot is very WALL-E coded.
--:-- John Coogan: And it's like maybe you could have made the drone a little bit more friendly.
--:-- Jordi Hays: Oh, I would have gone the other way. I'd have made the rolling robot a little bit more intimidating. Threw some barbed wire on that thing.
--:-- John Coogan: The crazy thing is, this product launching right now is tough timing because of all the new cybersecurity risks.
--:-- Jordi Hays: Oh, yes. And I wouldn't have even thought twice about it six, 12 months ago. Whereas now I'm like, do I want something that can take off in my home and fly around?
--:-- John Coogan: Yeah, yeah, yeah. Or at the very least, you've got to really, really trust Ring.
--:-- Jordi Hays: Yeah, totally. And Amazon, which is in the top five most trusted companies. They can spend more on cybersecurity for this product than any other home security company. But it is still like, there's always a human. But I...
--:-- John Coogan: You accidentally gave a password away or something. With the ADT acquisition, I expect them to introduce a bunch of products like this. And that's what I want out of these security companies. I talked a lot about the pitch for humanoids as a home security deterrence product. I think it's going to be far easier to have a humanoid be disruptive to a home intrusion than it is for a humanoid to fold laundry well or do the dishes. So bullish on this category.
--:-- Jordi Hays: Yeah, no, it's cool to see that Ring is still innovating even inside of Amazon. It seems like they're making good progress.
--:-- John Coogan: Let me tell you about Shopify. Shopify is a commerce platform that grows with your business and lets you sell in seconds online, in-store, on mobile, on social, on marketplace, and now with AI agents.
--:-- Jordi Hays: Fireworks has been hacked. Inference platform hacks. Obviously, Railway has been going through it. Fireworks has been hacked. Open Code was another that was dealing with a bunch of cyber issues. Basically, a month ago, various groups decided, like, we are going after the inference providers. One, you can potentially get access to compute. You can get access to compute, but you can also exfiltrate weights. And this Fireworks incident is the first example where I think they were not able to exfiltrate all the weights. Again, Fireworks is hosting open models that have been post-trained by companies. There's a lot of data in those models too. There's a lot of reasons why people would want those. But fortunately, it seems like they only got some small percentage of the weights and they caught it the same day. So I kind of expect it to be like roving efforts where we're going after startup, scale-up inference providers, and then it'll probably go to another category of company.
--:-- John Coogan: It seems like Ilya Sutskever sort of predicted this back in September. He said, Neoclouds have limited cybersecurity. Next time agents successfully go rogue, they'll try taking over a Neocloud to run more copies. This is bad. Thus, Neoclouds should greatly strengthen their cybersecurity and every company with strong cyber models should help with that. Raoul said, never wrong, just early. And this wasn't quite that because I don't think they got access to all the compute, but they stole a lot of data, stole a lot of weights. And they were not able to fully take control of the Neocloud.
--:-- Jordi Hays: Also, I think Ilya here is talking about rogue agents. And this was not. I mean, I don't think we believe that this is. So this is a proof point in the chain that an inference provider was able to be hacked, which is a novel and new problem.
--:-- John Coogan: Anyway, what do you think of hobbies?
--:-- Jordi Hays: Wait, actually, we have Chris Miller. Let's bring Chris in. We can have our hobby debate later. That's much less important than Chris Miller. Chris. How are you doing? Great to have you back on. It's been too long.
--:-- Chris Miller: Thanks for having me.
--:-- John Coogan: Thank you so much for coming on. We were just about to debate whether or not you should have hobbies because that was taking over the timeline. But there are much more important things in the news.
--:-- Jordi Hays: The best hobby this week would have been spending the last decade obsessing over global export controls and semiconductor supply chains.
--:-- Chris Miller: Yeah, I think you nailed it.
--:-- John Coogan: So thanks for getting a head start.
--:-- Chris Miller: Yeah.
--:-- Jordi Hays: How have you been processing the news? Is it as vague what actually happened to you as it is to us? Because it feels like every day it's getting worse, but there's also more tentacles. There's like seven different points of contention right now. How have you been processing the overall story as it's been developing over the last few days?
--:-- Chris Miller: Yeah, I mean, I think with scandals like this, we should expect they have a long way to run. You find out new things as the investigation deepens. I think already, though, for NVIDIA, the damage is clear. I think the other damage is in the Trump administration, which let this go on for a year without notifying anyone that there was a Chinese spy involved with the CEO of one of the most valuable companies in the United States. That is a puzzle as to why that happened.
--:-- John Coogan: Yeah, that's very odd. One thing that I've been really interested in is, obviously it's an American tragedy. It is a disaster. It is a crisis for NVIDIA. At the same time, it feels like it's a victory for the CCP. It's a victory for the spying that they wanted to do. They were successful in many ways. We don't know how successful. We don't know if they had access to devices and networks in NVIDIA, if they were able to steal designs. It's so much more a mixed bag. In some ways you lost your greatest ally, somebody that wants to sell as many GPUs as he possibly can to China. He doesn't want to lose the Chinese market. He knows they're going to still build their own independent supply chain anyway. And you lost somebody who was the most important person in terms of financing the global AI build-out, who had clearly aligned interest with you. By default, do nothing, win. All the same aligned, not total alignment, but enough alignment that he was going to do everything in his power. My big question is, why did NVIDIA not start this cleanup a year ago? It seems like they hired the PR firm Heller. So clearly they assumed that this would go public at some point. But it feels like they could have reduced the damage by just coming out and being the ones that led with this. And they would have controlled a lot more of the narrative. Whereas now I'm sitting here being like, I expect to get more bad news about NVIDIA every single day for the next six months. I just don't know where the bottom is.
--:-- Chris Miller: Yeah, I think that all sounds right. And I think if you look at the response in Washington, it's going to be to put tougher controls on NVIDIA and tougher controls on the Trump administration to make them enforce the rules that are actually on the books. So it's an interesting question for China. Are the second order effects negative rather than positive? If you can no longer smuggle as many chips in through Southeast Asia or get access to as many data centers that are in turn training and deploying Chinese models, it'll be an important question to see what, if anything, China did learn as a result of this. Right now, too soon to tell. Maybe we'll never be certain. But if the Chinese firms can't get access to Southeast Asia, that's a big deal for them.
--:-- Jordi Hays: The other thing is, NVIDIA, as they navigate through communicating through this crisis, they can in my view take a stance on either side without looking bad. If they say, all right, we're going to change our ways, it looks like there were people that were effectively compromised. And if they just continue the status quo, it's like, okay, also, you know, there's no winning from here either direction you take. For me, when you look at Jensen, he seemingly was doing two to three public speaking events a day, 365 days a year. I don't know when the guy takes a day off.
--:-- Chris Miller: Yeah, I mean, I think it is sad. It's kind of tragic to see what happened. And I think if you're NVIDIA, though, you've got to change your relationship with Washington. And it's been super adversarial, positive with the Trump administration, adversarial with Congress over the last 18 months. And now Congress is in the driver's seat on this issue. They're going to investigate both NVIDIA and the Trump administration as to how this happened. And I think NVIDIA needs a big reset with Congress. Otherwise, investigations are going to keep rolling.
--:-- John Coogan: Talk about the difference in NVIDIA's perception on the West Coast versus in Washington. On the West Coast, NVIDIA is just going around being like, you get $100 billion, you get half a billion, you get $2 billion. This Neolab gets $5 billion. This Neolab gets $3 billion. This inference provider gets a billion. They're just giving money to everyone and through that are very popular. Even you saw this the last couple of days, a lot of people defending Jensen just because he is a friend to many people on the West Coast, but also because he's funding all of their companies and portfolio companies. Whereas everyone in Washington that I know is like, this company made so many enemies over the last couple of years that the balance of power has shifted very much, like you said, to Washington and nobody seemingly wants to give them the benefit of the doubt on anything.
--:-- Chris Miller: Yeah, I just look at who's sponsoring legislation in Congress that's going to restrict NVIDIA. You've got one piece of legislation that has Senator Jim Banks, one of the most conservative Republicans from Indiana, and Senator Elizabeth Warren. They agree on nothing other than regulating NVIDIA. And that just shows the cross-agreement on putting more restrictions on AI chip sales and pressuring the Trump administration to enforce the current rules. And I think for a long time, NVIDIA thought it could fight that. And now it's going to be very hard to fight that. Because the argument that NVIDIA would make was that these customers in Southeast Asia are just normal commercial customers. You can't make that argument. One of them is run by a spy.
--:-- Jordi Hays: Do you have a sense of why diversion needs to go, like the GPUs physically need to get to China? Why can't Baidu just use a data center that's in Malaysia or Singapore that's set up with NVIDIA GPUs and just do all the training and inference over the wire?
--:-- Chris Miller: Well, they are doing that. I think you have to assume that they have a bunch of the most advanced chips in China, and they're also buying as much compute as they possibly can hosted in Southeast Asia. Some of it's, I'm sure, hosted in America. It seems like there's going to be a lot of new regulation around not being able to just buy cloud services. So much of this stuff was like the market's been moving so quickly that there's a bunch of these little edge cases that are allowing their labs to develop advanced models, even on the data side. I would expect now there to be enough momentum to sort of regulate whether American data companies can turn around and sell the same type of data that they're selling to the labs to Chinese companies. The whole landscape, I think, is going to shift.
--:-- John Coogan: No, I think that's right. And I think one of the ironies, though, is that this is actually a small share of NVIDIA revenue. If you look at the best estimates of chips smuggled into China physically, plus chips in Southeast Asian data centers, you're talking a couple percentage points of NVIDIA revenue for both.
--:-- Chris Miller: Yeah.
--:-- John Coogan: It's truly one of the worst own goals or bag fumbles, unforced errors. Jensen has built a fantastic business, but it does feel like the last three years you could do nothing and win. You could even go out there and say, look, yeah, the TPU is great. Tranium is great. Intel is great. You can buy from them. Ours is good. Do whatever you want. Don't even go on the offensive domestically. Admit that it's an oligopoly. And you'd still be a trillion-dollar company. You would be extremely successful. But it was like, you've got to do more. You've got to win. There's some other motivation there to not just... Apple has not chased the AI, we've got to train a foundation model. They've stuck to their core, and it's been fine. And they're the most valuable company in the world now. And it feels like you could say, oh, yeah, I'm NVIDIA. I'm going to make incredible cash flow over the next couple of years. I'm actually not going to travel at all. I'm not going to leave the United States ever. I'm not going to take a call with somebody on the other side of the Pacific because who cares? I'm making so much money. It's great. I don't know.
--:-- Chris Miller: Well, I think that's going to be the new policy. I think it's got to be something like that. Refocus on the U.S. market. That's where the money is anyway. And Southeast Asia and China are going to be written off.
--:-- Jordi Hays: Yeah. I know a lot of the not just the AI doomers, but also safety-focused people, EAs, all sorts of people in the AI landscape, even more optimistic people, were always upset about any GPUs making it across the Pacific because they were saying, there's enough demand here. We'll buy them. There are so many American companies that would love to buy those at the full price. And so you're chasing this revenue diversification across the Pacific for not even that much of a financial gain, not just on the revenue scale, but also what it would actually do to the margins because you could still keep the prices high. One interesting question is how much of that philosophy is a legacy of the global capitalist mindset of the 90s, the 2000s, throughout the birth of NVIDIA? I was trying to steel man that philosophy because it's very out of vogue right now when we are in the era of tariffs and protectionism and geopolitical nationalism, America first. But also there's a whole bunch of other political groups that are in favor of restricting AI to China. If you go back to the nineties, it feels like selling internationally, expanding your business globally is something you would get a pat on the back for as a business leader in America. Does that track with your understanding of how the global political business landscape has evolved?
--:-- Chris Miller: Yeah, I think the way I would frame it is that most chip companies came to terms with the fact that over the last five or 10 years, governments were going to have a big say in the future of the industry. NVIDIA tried to push against that. NVIDIA wanted to stop that from happening. It thought it could tell Washington that actually its chips weren't strategically important. And now we've got not only evidence that Washington is going to make its voice heard, but that China believes they're so important that they're worth sending a spy to gather information on. And I think this is going to be the final data point that forces NVIDIA to realize that everyone else around them thinks that their chips are strategically significant, even if they don't think that governments are going to force that upon them.
--:-- John Coogan: It was always such a hard argument to make from their side, which is that, yes, AI is going to change everything. And we need to rally trillions of dollars of capital as a country and as the West to set up all this infrastructure because we are going to solve all the major challenges of humanity. But also this magical technology is not relevant to national security and we should just sell it to everyone. It was always an impossible argument to make. And I don't think they're forced into not being able to make that argument anymore. Can you give us a quick tour of the other American semiconductor companies and their stances on China? Because Google tried to go to China. They tried to bring Google search to China. Unsuccessful. They were asked to censor it. They never really developed a business in China. I don't imagine that they're selling TPUs to China. I'd imagine that they don't even take meetings with anyone that looks like they could be diverting TPUs to China, which would be another big issue. Amazon, probably similar. Even though Amazon sources a lot of products from China, they're probably not selling stuff. Intel, is that a more complicated situation? Because ultimately, Intel chips go into IoT devices, cars, all sorts of different things. But take me through some of the other landscape and the other postures that we've seen from the semiconductor industry.
--:-- Chris Miller: Yeah, I mean, I think Intel is certainly a more complicated case. I would say most chip companies want to sell their products around the world. Fair enough. That's their job. Their job is to sell. I think the difference is that most chip companies would have gone to the U.S. government and said, hey, we have this economic interest in selling to China. Or, hey, our chips aren't actually that good. Let them sell to China. Whereas NVIDIA said, actually, no, these aren't strategically important. You're misunderstanding the U.S. relationship. It's actually not about AI competition. Going to the core of what both Washington and Beijing actually think the competition is about. So it was this clash of worldviews that made NVIDIA really stand out. And I think that's what's going to have to change.
--:-- Jordi Hays: Last question for me. In America, we are overwhelmingly focused on AI. We are fully awake. ASI, RSI, these discussions are all over the place from the New York Times to the Dorkesh Patel podcast. It's everywhere. What is the status quo in China? Have they woken up? Are they taking this really seriously? Is this evidence that they take it so seriously that they would deploy a spy potentially? Or is there still a little bit of they haven't fully shifted into gear?
--:-- Chris Miller: I think that's a key question. And the Chinese system is not monolithic. On the one hand, you've got spies deployed to track NVIDIA. That's an important point. On the other hand, recent reporting this week that you've got NVIDIA H200s stuck in customs in Hong Kong, purchased by Chinese firms, but the Chinese won't actually let them into data centers. Which you wouldn't do if you were super AGI-pilled. So I think there's still a debate in the Chinese system about just what approach to take on this topic.
--:-- John Coogan: It's always felt like a longer time. Last question. I know we're out of time. What story do you think is being overlooked now that this will dominate all of the headlines for months to come?
--:-- Chris Miller: The second order impact on data centers outside of China serving Chinese customers. It's both Southeast Asian firms, but also you look at Oracle, which has ByteDance as a huge customer. I think that's going to be in focus as well as this investigation rolls on.
--:-- John Coogan: Yeah. No, that makes a ton of sense. Well, thank you so much for taking the time to come chat with us.
--:-- Chris Miller: Yeah, I really appreciate it. Have a great rest of your day.
--:-- Jordi Hays: Busier than ever. Thanks so much, Chris.
--:-- Chris Miller: Great to see you. Goodbye.
--:-- John Coogan: Let me tell you about CrowdStrike. Your business is AI. Their business is securing it. CrowdStrike secures AI and stops breaches. And up next, we have Ed Bolian from VinWiki, the founder and CEO. His YouTube channel is also VinWiki. Ed, how are you doing? Welcome to the show.
--:-- Jordi Hays: Palette cleanser.
--:-- Ed Bolian: Palette cleanser. Very well. How are you all?
--:-- John Coogan: We're great. We talk about a lot of AI. I also talk about a lot of cars on this show. Very excited to have you on here for the first time. Thank you so much for joining. Would love to start with just a little bit of your backstory for those who are watching who might not be familiar. Can you give us the high summary of the VinWiki business but also the fantastic Cannonball records that you set? Just kind of take us through some of the career.
--:-- Ed Bolian: By all means, yes. We started VinWiki about 10 years ago, but prior to that I was director of sales at Lamborghini Atlanta, and before that I had started an exotic car rental company out of my dorm room at Georgia Tech with some stated income loans back before the financial crisis. So I've had a long journey in the world of interesting cars and interesting things to do with them. Back in 2013, a co-driver and I set the Cannonball Run record. So there was a real Cannonball Run of the 1970s that inspired the 1981 Burt Reynolds movie. But there are still crazy people like me that occasionally go out and see how fast it can be done. So we did it in 28 hours, 50 minutes back in 2013. And then through COVID, that was able to be massively advanced. So now the record is 25 hours, 39 minutes. It seems like it's quite an advantage if the cops aren't working and nobody else is commuting on the road. But now at VinWiki, we tell car stories and have tried to change the way that the world looks at vehicle history reporting and automotive history to try to empower people to immortalize the best stories that they can about their cars.
--:-- Jordi Hays: Are we at the end of the cannonball era? Will we never see a new record?
--:-- Jordi Hays: Because COVID was so unique in that there were no trucks, no cops on the roads. But at the same time, technology is advanced and cars are getting faster. It presents a new risk, which is that car enthusiasts might try to start a global pandemic to create the environment that they could beat the record.
--:-- Ed Bolian: Yes, yes. That's a good conspiracy theory. There you go. Whatever means necessary. You know, it's interesting because it may have sufficiently flipped the sandbox over that no one decides to go out and try it again. Certainly under the current technological paradigm, it is problematic to really set out a plan that gives you a really high likelihood of success against that mark to beat.
--:-- One thing that's been amazing in the years since 2013 when we set the record is that the enthusiasm for Cannonball has been made manifest in a lot of different ways. So we've had electric car records and self-driving records and truck records and motorcycle records and three-wheeled records and coast-to-coast-to-coast records and Southern route records. And so it's become more interesting for people to just find their own way to sort of have their cannonball record that doesn't have to be averaging 110 miles an hour across the country, but it can be their own husband and wife record or whatever they want to do. And so I've enjoyed that as well.
--:-- Jackson: Yeah. I mean, there's forever going to be opportunities to have the most niche cannonball. Oh, well, isn't there a cannonball that was... I think we're sub-100 horsepower.
--:-- Ed Bolian: No, I know what we do. I know what we do, John. Yeah, we do the tour bus cannonball record and we do the show the whole time.
--:-- John Coogan: Oh, I like the crew live stream the whole thing. Yeah, we did do a sub-100 horsepower race where everybody had to show up and we would dyno the cars and assess penalties for those that went beyond it. I ended up just renting a Corolla for it because that was the easiest way to get a 100 horsepower car up in White Plains. But it was a lot of fun to be honest. I've done it about 10 times now, most of them not in earnest trying to set a record, just going out as crazy as we can. A couple years ago we did it in a 1976 Rolls Royce dressed as sheiks like Jamie Farr and had a blast. It's a very different pace but a whole lot of fun.
--:-- Jordi Hays: I love that you're like, you rented a Corolla. They're like, yeah, just bring it back with a full tank. Tell everyone what the punishment was for cars that dynoed over 100 horsepower.
--:-- Ed Bolian: There were several different ones. One was that you had to assemble like a thousand piece puzzle and glue it to the hood of your car, which essentially meant that you didn't get to sleep the night before. One was either a ballast, so you either had to carry a lot of weight, like bags of cement, or keep a lobster alive since we were starting by the ocean. No one chose the lobster, but there was definitely some cement transported. Sandbags.
--:-- Jackson: Yeah. That's great. And when you did the cannonball, did you have a spotter plane or spotter cars? Like what was the actual team? Cause you said you had a co-driver, but was it just two people in the car or three? And then were there other people involved? I know that there's like sometimes whole teams that go after the records.
--:-- Ed Bolian: Correct. And the most recent record involved 35 spotter cars driving at pace with the organizers of the run, the drivers agreeing to pay tickets for them if they get stopped. I started planning this in high school, at least ideologically, about trying to find the right recipe to do it. And the record at that point, this being 2004, had not been advanced since 1983. And so while I was sort of thinking through this, two efforts were made, one by Alex Roy and Dave Maher in 2006 in a BMW M5 that did involve a spotter plane. I did not have any friends with airplanes. And so none of them were volunteering to fly ahead and let me know what was up ahead. And then Richard Rawlings and Dennis Collins from Gas Monkey Garage did it in an '07 and a 550 Maranello. So I had a couple of guys that agreed to drive like hours ahead and let me know if there was any new construction or new hazards that I might run into. But no, nothing of the sort in terms of planes or people ready to go to jail for me.
--:-- Jackson: Switching gears to car collecting. I'm interested in the story of the LP640, but also VinWiki and how those two things fit together. How much of your choice in collection is driven by what you discover when you're doing VinWiki work, which you can kind of just describe and explain some of the realizations that you have that actually impact the market versus how much of it is just you like what you like and sometimes you get lucky?
--:-- Ed Bolian: You know, that really is it. You have to start with something that you really, really like whenever you're thinking about something that might be defined as like an alternative investment class. It's not something that you set out like Dave Ramsey style, like you should go get some gigantic loans for rare exotic cars and just see what happens. One thing that I learned early on while I was selling cars is that the ones that the guys that seem to be sort of beating the game were the ones that were buying cars further along a depreciation curve where stability was a little bit more likely. And the way I used to phrase it is that I tried to look like I won the lottery 10 years ago.
--:-- Because with rare and interesting cars, generally there will be some initial depreciation as people are trying to buy the newest, latest, greatest, biggest, baddest, whatever. But I wanted to try to avoid depreciation. I call that my greatest allergy is automotive depreciation. And so I was looking for cars that seemed rare, but also seemed like they were at a point where they were a lot more stable and potentially starting to go back up, which was still relatively unheard of in the early 2009 to 2015 when I was selling cars every day.
--:-- And so I loved the Lamborghini Murcielago since it came out. I just thought it was the coolest car. I love the heritage and lineage of the V12 Lamborghinis, especially the manual transmission cars. And I learned that they were a lot more rare than really anybody seemed to know. As manufacturers started to bring out these sequential manual paddle shift cars, the thought was that it was about a 90% take rate, but it turned out now we've now learned that it was about 99%. So for the Ferraris and Lamborghinis where there was a manual versus a paddle shift option, almost all of them were paddles. And so that made them much more rare than we expected. And usually rare just means nobody bought it when they were new. And that's true because back then the only way that they could make cars faster was to make the transmissions work better. And so now I realize that they were rare. And that's a lot of the ideology around the VinWiki app is how do we document rarity and significance in cars that might eventually have something to do with the value.
--:-- Jackson: Is manufacturer data getting more reliable over time? Because it feels like that 90 to 99% discovery, that was unique. That's a gold mine. But is that also going to happen in a 2026 car run where you know some manufacturer they're vague about it or they give incorrect information and then 10 years down the line we discover through chronicling the VINs that there's some oddity with some spec that's extremely limited?
--:-- Ed Bolian: There is some chance that there will just be something that's overlooked as being important. Lamborghini didn't think it mattered for anybody to know how many of specific cars they were doing. In the same way that Porsche wouldn't necessarily think about how much it matters of how many they build of a certain paint to sample color or something like that. Those are the types of granular data points that might be uncovered later down the road because there is a lot more documentation now. The companies have grown up through all of the different ways that they have combined corporately. And so it's a little bit less likely that that sort of thing will happen. And the cars are a little bit more homogeneous. With some exceptions being like 911s range from $150,000 up to $400,000. And there's different permutations of how they're configured. And obviously, the profitability changes alongside that.
--:-- Jackson: How are you thinking about Ferrari's depreciation curve and their changing strategy? Because there's sort of two lenses. We've seen the run of the 430 and many others from that era, the sort of mid-2000s cars skyrocket in value, at the same time when the SF90 has collapsed. But I can't help feeling like in 10, 20 years, there might be a new class of 30, 40 year old guys who are like, oh, the SF90 is when it was the best. And how much of this is just a generational thing versus a structural change in the way Ferrari is producing and selling cars?
--:-- Ed Bolian: So you're exactly right that the generational aspect is huge. Right now, anything built pre-1980 is either going down or at least flat as a best case scenario, even the most significant of pre-war cars and stuff that has always been kind of the headline of auctions. And that's because most of the people that loved those cars growing up have already made the decision about whether or not they're going to engage there as collectors. And so today we're seeing the people that grew up with the posters on their walls of the iconic 80s, 90s, and 2000s cars really coming into a position where they can engage in that marketplace. And that is what's driving it. So that's why the 430s, as you said, are starting to go up and why the SF90 just isn't there yet. It isn't old enough yet. And we just brought out the 849. So what we're seeing with Ferrari is always going to be kind of on the cutting edge of what the collector hobby is going to do next.
--:-- Jackson: Do you think there could be sort of a lost generation? I'm thinking like, the 2015 to 2025, where no one ends up really wanting to collect these cars because I have felt like in the last year, the manufacturers have all got the message that they went through this era of making heavier and heavier cars with hybrids and basically every it was like the most infuriating era even as a casual car enthusiast where you're like, no one is listening to my requests.
--:-- Jackson: And then now they're getting points because at least I'm giving them points. Cause it's like, okay, well you didn't make like a true gated manual, but you made something that feels like it. But to ground the 2015 era, what are you talking about? Like the SF90 is like using that as the sort of...
--:-- Ed Bolian: Oh, okay. So yeah, more 2020s.
--:-- Jackson: Yeah, yeah. The 2020s feels like it's the sort of peak of this of not listening and obviously dealing with the pressure from various regulators.
--:-- Ed Bolian: Regulators, yeah.
--:-- Jackson: But now with Mate being like, all right, we're in our, effectively saying we're in our quartz crisis era. We need to play our own game. It feels like we're headed back to the future looks more like the past, which I think is kind of good and bad. The good is that it feels like there's no more innovation, right? It's just like figure out how to make cars like we used to make cars. Look at the Mission S. I mean, it's like a nod to that after a lot of, hey, maybe it's going to be all electric.
--:-- Ed Bolian: Yeah, that day, that Thursday felt like monumental because of the fork in the road. You had this kind of performance from Tesla and a very different direction plus the Mission S, which is like, hey, we're just going to make the thing that you've always wanted.
--:-- Jackson: Yeah, right.
--:-- Ed Bolian: Yeah, we're listening. It took us a while, but we are. And I think you're right that we may have this sort of alienated generation of cars, mostly based on hybrid long term reliability concerns that we're just going to run into cars that are a lot less functional at some point than we'd hoped for. And there may or may not be engineering workarounds to make them into more drivable or usable cars at that point. And sometimes that's a very healthy thing for the market at large because when the goalpost doesn't move aspirationally, when there's no new cars that you have to dream about or chase, it allows you to focus on the cars that are more genuinely or earnestly sought after. And the competition there does drive up prices. And so sometimes having that lost generation kind of emboldens the enthusiasm that is made manifest in the collector's attitude towards these cars.
--:-- And so now some of these things like an electronic gated manual transmission, it may work and it may succeed or it may come off as a little bit patronizing that, oh, here that manual thing you asked for, it really just sits in the cup holder and you get to shake it around a little bit in order to make vroom noises the same way that speakers making exhaust noises.
--:-- Jackson: Yeah, the end state is just you're in a driving simulator, and the car is driving itself wherever it wants to go. You make a good point about hybrid systems maybe being unreliable, but it's kind of fallen on deaf ears from the guy who daily drives a Bugatti Veyron. So how have you squared those two? Because I thought that car was the king of unreliability, but maybe I have it wrong. How has it actually been daily driving?
--:-- Ed Bolian: Well, the interesting thing is it's so daunting to think about personally maintaining like a Ferrari hybrid system. Whereas these other vehicles, they're fundamentally just a mechanical machine with more electronics. There's not really a monster in the closet with a Bugatti, something that's going to bite you one day inevitably. There's plenty of things that are very proprietary and very expensive if they go wrong. But fortunately with all cars, the more you drive them, the more reliable they are, regardless of how high strung they are, potentially even more so if they're super high strung. So my 2012 Bugatti Veyron that had not been well cared for prior to my ownership by a variety of rappers and entertainers and things like that who owned it has honestly been great for about 15,000 miles now.
--:-- Jackson: You've put 15,000 miles on it already? I had no idea. About two years? Wow. And is most of that rallies and actually driving it, or are you genuinely daily driving it?
--:-- Ed Bolian: It is more big trips. It's a lot to take it somewhere just because you don't get anywhere fast. Everybody wants to stop and talk about it. And I enjoy that too. But you have to really plan when you're going to get somewhere based on the likely obstacles in the road on that journey. So I drive it every chance I get. I give anybody a ride that really wants one. To me that's what makes it all fun. That's what perpetuates the hobby anyway.
--:-- Generally my attitude with car buying is that I'll buy the worst examples of the coolest cars on earth. And in those cases, because the cloud of their backstory becomes more difficult or problematic in terms of the storyline, you sort of redeem the value of the car by using it without calamity. So if the car is not expected to be in good shape but then you put 5,000 miles on it in a year, generally the audience is going to say, okay, well this thing probably okay. We don't need to worry that much. We can trust the car rather than worry about the backstory. And so it's easy to get scared off by that, but it's also pretty easy to overcome it.
--:-- Jackson: Yeah. I love buying the worst example of a great car, looking like you won the lottery 10 years ago. These are hilarious phrases and ways to think about it. But how much of that goes into the spec? Do you have particular opinions about color and interior? Red over red is a particular choice.
--:-- Ed Bolian: It's a choice. It was not my choice.
--:-- Jackson: Yeah, but does that just not factor into it? Or does it actually just add to the lore and it's just even wilder so you're going for it? How do you think about the aesthetic piece of it? Because a Bugatti can be relatively understated when it's specced a certain way. Not yours. But it's almost more fun that way. And that's what the point is of these vehicles.
--:-- Ed Bolian: Yeah, you don't get to go to the used car factory and spec out exactly your dream every time. And so I do wind up with some pretty wildly specced cars. This car certainly being a key example of that. It was actually specced by Bugatti. It was the one they showed off at the Frankfurt Motor Show back in 2011. So red over red with red wheels was a thing. And so, yeah, I wouldn't expect one that way, but it does work really well for YouTube. The thumbnails work great. The contrast is nice, but the cops do not miss it. I'll tell you that much.
--:-- Jackson: Well, the funny thing is I think that just color matching interior to the body, to the wheels, is starting to be really hot. Even in resto mods, people just seek this out. So you kind of know it's going back in vogue. Maybe you're part of the early part of driving it. It's like the skinny jeans. What do you think about the various manufacturer programs where you basically give the car back, they'll reset the mileage to zero and then make the spec whatever you want? It feels great for the actual owner of the car, but potentially bad for, or somewhat equalizing. Like, it basically levels the playing field.
--:-- I have a friend who bought a 918, put a million miles on it, drove it like absolute crazy, was actually dailying it all over Europe. And then he knows that he can just send it back and they reset the mileage to zero and they're going to change the color and change the interior. And honestly, I think they're charging too little because the ultimate value of the car will be after that is like, you have a basically 918 with zero miles on the dash and the craziest spec, all refreshed, and you get the privilege of seeing all the other specs before you spec yours again, which you didn't have if you ordered the car new.
--:-- Ed Bolian: I think you're right that that is certainly now in vogue and most of the manufacturers are offering some sort of a program to do so. Lamborghini is probably lagging the most in that regard, although they will in theory do whatever you want. Bugatti now will take any of their own back and it's about a million dollars for engine overhaul, paint, and interior. But again, you can respec it to the moon and back if you so choose. And I think that is exciting. The Porsche Carrera GT is the first one we saw like a real ROI on that. Certainly some other 911s are following suit where you'll spend $500,000 or a million dollars and somehow you've doubled the value of your three million dollar car. So that to me is interesting and I think it's driven a lot by the fact that they're not taking too many of them. I think they've said that they've sold out the next three or five years of slots for those levels of recommissioning.
--:-- And so it's relatively unheard of that you modify a car and make it more valuable. It doesn't matter if it's turbochargers or underglow or a factory repaint. Generally, it is seen as a detraction. But this seems, at least at the moment, as potentially not that case. Now, if they all get redone, then certainly the ones that remain original will be a thing. But your friend's attitude is exactly the right one to take. It is to say that I get to use my car completely and then I can go get my get out of jail free card.
--:-- Jackson: I actually think it's great because it just means more of these cars will just be on the road. Even in Southern California, there's an incredible density of some of the coolest cars ever made, and yet I'll see thousands of base Carreras before I see like one of these special cars on an average day on PCH, even on a Sunday when they should be all out and about.
--:-- Ed Bolian: And if that cures the anxiety of utilization, then it's ideal, right? Because it gives you the chance. Even without a reset, if it's just known that they're totally redone, that's one thing. I'm not crazy about Porsche's decision is the odometer reset because I'm always the guy that loves having a high mile odometer. I actually had driven one of my Murcielagos on a long, long trip, choosing it against other cars because I knew that those miles would put it as the highest mileage one in the world. And so I wouldn't want them to reset all my hard work if they changed the color.
--:-- Jackson: I'm sure that's an option. We might charge you back 10K for it. We won't reset it for 10K. Just roll it up. Give me an extra 100. I earned bonus miles this week.
--:-- How do you feel about this new sort of wedgie, angular aesthetic, the stainless steel? I'm thinking of like the Audi Nuvolari, the Jaguar Type 01, the Cybertruck. Or the ultra-limited Cybertruck.
--:-- Ed Bolian: The ultra-limited Cybertruck.
--:-- Jackson: You got to get that on VinWiki. There might be less than 79,000. It might be going up in value. But put aside the internals, and I just want to hear you talk about what that design language. Does it speak to you as someone who grew up and has owned many Lamborghinis and engaged with more aggressive designs? Is this a welcome step forward for design language, or is it a detour along something that you are not interested in? How have you been dealing with that?
--:-- Ed Bolian: Well, the fortunate thing is that it really doesn't matter that much what we think of them, despite how much we all like to comment on them. Because we're not usually the ones that are going to sustain the values of these cars over time. The goal is to appeal to the inner 13 to 17-year-old kid that is dreaming and reading every bit of media, watching every YouTube video. Those are the guys that you really want to make love the cars. And as long as that happens, then the hobby and the enthusiasm sort of sustains.
--:-- I love simplistic designs. Lamborghini used to always have silhouettes that could be drawn with one line. And that means that the cars are much more easily recognizable in a sea of other cars. When one passes you on the highway, it makes a greater and more immediate impact. I like that a lot. And I think we are seeing more of that in these iconic silhouettes and a little bit more restrained design language. You kind of think the opposite of the Pagani retooling right now. There's still a demand for that sort of thing that's got eight different colors and four shades of carbon fiber and all of the customization that you could ever dream of. But sometimes, taking one extra piece of Johnny Depp jewelry off and leaving it maybe just slightly less done makes a lot of sense. And so when I find cars immediately good looking, it's usually because they're not so over styled and overthought that it is something that you can immediately visually at least digest.
--:-- Jackson: When the car market inevitably sees a correction, how do you think it'll be different than the sort of post-2021 era? Because in that case, you had people spending a million dollars on an SF90, taking it, and then as it started correcting, having negative equity where they just owed, all those viral videos in maybe 2022 and 2023 where people are like, I got to pay you 400 grand to give back my SF90 because they had some super low interest rate and they'd finance the whole thing. This time it's been interesting. This run up that we've seen in so many cars is not really debt financed, or some of it could be, but it's people that are paying six, seven percent. Which means that I think it will be maybe less violent of a correction. I don't really know. It might give us a softer landing if that's the case.
--:-- Ed Bolian: The bigger reason that it would be a softer landing is that most of the debt available in cars that are 10 or more years old and are sort of classified as collector cars is around 60% to 70% loan to value. So even in the event that there was a pretty meaningful contraction of value, most of that equity or the lien is going to be able to be satisfied regardless of those conditions.
--:-- What I believe we are going to see is a stronger fork in the road between cars that are legitimately rare, collected for the right reasons, versus just everything for sale, everything hot the way that we saw made most clear in the 2021 zone where we also had inventory shortages and ship shortages and people not able to spend money on travel and other things. It was all just going into hobbies and cars and stuff like that. So at that point, most of the debt was heavily over leveraged because it was more so built on how somebody buys a new truck or a Toyota because they had to be able to loan 120% of MSRP for anyone to buy a new car. And so that sort of trickled up into the exotic car world and was more problematic. Now that really doesn't seem to be happening.
--:-- We are seeing sort of cars get king made more or less now where collectors will come in that may have almost no experience in the car hobby and are treating it much more as an asset class or an investment option. And they'll come into dealers with just a gone in 60 seconds looking shopping list and say, I want a nice one of everything. And there's just not nearly enough supply for that. So you combine that, whether it's their own money or out of a family office, sometimes they're representing private equity. But that's not super common in the space just yet. It would massively change it if we started to see the way that they've engaged in art and stuff like that. But better debt and more debt available under proper terms is helping. And then, of course, there's just a crazy transfer of generational wealth right now that is fueling the hobby.
--:-- Jackson: How do you think track-only cars are going to evolve? It feels like Willow Springs is going through a renovation right now. There's more track clubs, and motorsports can be very expensive on the tracks, but that's part of that K shape that you described. And so there's a world where manufacturers lean into that more. Do you have any view on where non-street legal vehicles from major manufacturers might go?
--:-- Ed Bolian: My hope is that it always looks like a Bugatti Bolide that is so obviously not road eligible that it allows them to do crazy stuff like Aston's car built for Ken Griffin with a money no object. Like, what can you do? And I like it when on a very low volume manufacturers do that as kind of a thought exercise to say, all right, if we didn't have to worry about you not dying in a crash, we could do this. And if we didn't have to comply with any government's rules, we could do that. And to me, when it's insane, it's at its best. When it's like a track version of a car that just sort of has crazier aerodynamics and no ground clearance, nobody cares. But when it's nuts, that's awesome.
--:-- Jackson: I'm just wondering if the rise of self-driving cars and speed cameras would take some of the energy out of the fun that you can have with a road legal street car, canyon carving on the weekend, because you're 100% going to get a ticket or are you going to get stuck behind someone that's in self-driving mode or something?
--:-- Jackson: And so it pushes people to the tracks more, but I don't know.
--:-- Ed Bolian: Well, yeah, that's top of mind for us. My kids are all under five, and I assume that they will never drive themselves on the road, but I will want to give them the experience of the joy of driving sports cars. And so, yeah. The track.
--:-- Jordi Hays: Yeah, and the beauty of it, if you're just a pure enthusiast, is the car that looks like it costs $5 million is like $600K because you can only use it at the track.
--:-- Ed Bolian: Oh, sure.
--:-- Jordi Hays: And so it feels like the value you get for a lot of these things ends up being like, if you don't care about flexing it at Cars and Coffee or driving with a bunch of aero to the grocery store, you're in a pretty good spot.
--:-- Ed Bolian: Exactly. Now, we are seeing an issue, especially in the United States with these cars, that the United States does not have a good program to have an independent adjudicator decide that a car could be street legal the same way the UK always did. So like, you know, especially with rally cars and things like that, you could put proper turn signals and proper lights and proper seatbelts in them and go get the car MOT checked and run it on the road because it was sort of approved for road use. In the U.S., we don't have a system like that in any state. However, various states have much more relaxed rules about what constitutes street legal or what constitutes title eligibility. And so we're seeing a lot of these cars popping up that were built as race cars that have got titles. And that opens up a tremendous amount of legal liability and things like that. And so once again, that's why I prefer that they be nothing close to a road-looking car, because it decreases the temptation.
--:-- Jackson: Yeah, yeah. I've heard crazy stories about track-only cars getting license plates and driving around because of just like an error at the DMV. Almost like crazy stuff, right?
--:-- Ed Bolian: Yeah, they just took in the MSO and they gave them one even though they weren't supposed to. And yeah, when it was convenient to make it easier to transact them because you're not doing bill of sale only. But we got to use our powers for good, not for evil.
--:-- Jackson: The funny thing is sitting in a Valkyrie, you know, with the engine so loud you got to wear ear protection. I'm like, this thing can go on the road? That feels so wrong.
--:-- Ed Bolian: Which a Valkyrie is a show and display, so it's not really supposed to go on the road. It can do 2,500 miles a year, but it still found another way thanks to Bill Gates around the rules.
--:-- Jackson: That's right. That's right. Oh, that's good context.
--:-- John Coogan: Well, this has been an absolute pleasure. Thank you so much for coming on the show. Congrats with all the progress. You know, my final parting thought is now it's getting a lot faster to make video games. Somebody needs to make a really pure enthusiast version of Cannonball Simulator, but it takes the full amount of time so you're driving at normal pace on the exact Microsoft Flight Simulator, but for driving.
--:-- Jordi Hays: Yes. Yes. iRacing, but for the Cannonball. Somebody build it.
--:-- Ed Bolian: Yeah. You got at least three people that will play it.
--:-- Jordi Hays: That's it. We're in.
--:-- John Coogan: Three customers for you.
--:-- Ed Bolian: That'd be fantastic. That's perfect. Sounds like a blast. Thank you so much.
--:-- John Coogan: Thank you guys so much for having me. Love your show.
--:-- Ed Bolian: Yeah. Great to meet you. Have a good one. Goodbye.
--:-- [Segment transition]
--:-- John Coogan: Let me tell you about Codex. Codex is a powerful workspace for getting work done with AI agents. Whether you're writing code, analyzing data, creating content, or automating business workflows, Codex helps you move projects forward from start to finish.
--:-- Jordi Hays: And we have our next guest, Justin Sealy from EBS, co-founder and CEO, taking AI-powered robots to defense factories. Finally. Welcome to the show, Justin. How are you doing?
--:-- Justin Sealy: I'm good. How are you guys doing?
--:-- Jordi Hays: We're doing great. Sorry, we're running behind. Give us the latest news. What's up, guys?
--:-- Justin Sealy: Just announced today a $30 million seed round with Kleiner Perkins and Cousin Noah. So coming out of stealth.
--:-- John Coogan: When did you actually start the company?
--:-- Justin Sealy: Okay, pretty quick.
--:-- John Coogan: How do you actually lay a wire harness with a robot? Is that like a robot arm or like gantry or humanoid? What kind of robot are we talking about?
--:-- Justin Sealy: Yeah, so quick background. Basically, we're a bunch of defense tech guys, and one of the biggest problems is getting technology into defense tech. It's like 80% IT and security requirements, and then you've got to solve the problem as well. The big problem is, how do you use as much off-the-shelf stuff as possible? So we just take off-the-shelf robots as is. You get all that security, you get all that safety from it versus building all the hardware from scratch. Then you get to go deploy much faster. And to answer the question on the wire harness thing, we try to take, if a wire harness takes 50 hours to build, you basically walk through and you find all the pieces that you can automate so the artisan can build more faster.
--:-- John Coogan: Okay, so you're building the wire harness, not necessarily installing it. Because when I think wire harness, the only diagram that comes to mind is the Cybertruck's 48-volt architecture. And you can kind of see it's pretty neatly arranged, but it still looks like a nervous system or like a bunch of arteries throughout the vehicle. And that feels like a lot of very fine-tuned pieces. But you're laying out the wire harness kind of on the floor and then delivering it, and then they go and install it into the drone or the vehicle or whatever. Is that right?
--:-- Justin Sealy: You got it. So basically, you build these things on large tables. They're called layup boards. And across the U.S. and the world, there's hundreds of thousands of humans who sit there all day and manually build these things. It's incredibly manual. There's no automation that really works for this. So as the first starting point, one of the places we can kind of speed up the defense industrial base is making these robots run when people don't want to work on nights and weekends. So then you get more capacity out of these existing artisans.
--:-- Jordi Hays: So how general do the AI robots need to be? How much of this can just be automation in the sense that you have maybe used AI to create a workflow that produces a wire harness at 99% reliability? And then you can just run it as a manufacturing line consistently. Like, you know, the original Ford assembly line is AI in some sense. It automated. There are machines, but the machine always puts the windshield in the right place. You don't just say, "Hey, make a car," and then it just figures it out. How general do you want to be in the short term, medium term, long term?
--:-- Justin Sealy: Yeah, so the difference between deterministic and probabilistic robots, right? If you're going to build, like I was a factory engineer 10 plus years ago before I went into defense tech, and if you're going to build more than 10,000 to 20,000 of something, you should just call ABB and build a factory line for it.
--:-- Jordi Hays: Got it.
--:-- Justin Sealy: Then you get deterministic every time. You get that perfect every time. But the really nice thing is actually in the defense industrial base, we don't build hundreds of thousands of anything. We built like 100 F-35s last year and stuff like that. So actually, VLAs and probabilistic is really good for the defense industrial base because that same robot needs to build 10 and 20 different SKUs.
--:-- Justin Sealy: And so that's where we're actually very bullish on this segment of the market because humans work at a workstation and they might build part X today and then part Y tomorrow. And VLAs and kind of like general purpose robot arms are really good at switching. There's no tool change.
--:-- John Coogan: Yeah. So I agree that makes sense in terms of meeting the market where it is. But how do you respond to the proposal that maybe the future actually does look like just one million DJI style drones that are extremely commoditized and we are moving into a regime of high volume manufacturing? And yeah, maybe we'll keep making a couple F-35s every month. But realistically, wars will be won and lost with high-volume manufacturing capability around extremely standardized pieces of the supply chain.
--:-- Justin Sealy: I think it's kind of, you know, $20,000 and $50,000 or something is like a good amount. It's still not hitting mass production like you see in other countries. And so for me, it's kind of like let's go hit that number first. I think we built like 100,000 and then we kind of work up there. But also, I think what you're seeing in Ukraine and others is designs change pretty rapidly. And so you need to pivot between designs really fast. So that's also kind of changes like probably really nice. It's like, oh, you had a tool change design or retool the entire line to build that thing. Now you can segment out the pockets that VLAs can just like an operator can just go like retrain the robot for maybe 10 hours and then do a new training run. I think you can have built the next thing quickly.
--:-- Jordi Hays: Got it. Last question. $30 million raised. How much of that is like general operational expenses versus actually acquiring robotic infrastructure, like effectively CapEx? I imagine that you're able to finance it if you do that, but what is the mix of use of funds for the next 12 to 18 months look like?
--:-- Justin Sealy: Yeah. Biggest thing is we probably can pay for all of our CapEx through revenue. So we did seven figures of revenue within the first six months. So the idea is like, let's go do eight figures in the next six months and let's just keep crushing. But the biggest thing is it's all acquiring AI talent and then scaling data. That's really good. And so most of that's for scaling data and we need hyper specific data where you're actually doing technical tasks. And so you have to train operators to do that, be very proficient. It's not just general data collection. These are kind of artisans in their own way. And so scaling that while also getting all the software and AI and robotics talent you need in one place is what you need cash for.
--:-- John Coogan: That's awesome. Well, congrats on the fundraise. Congrats on coming out of stealth. Great perspective and great progress.
--:-- Jordi Hays: Got to lead with the next time. We'll be leading with the revenue. We'll hit the gong for that.
--:-- John Coogan: Great stuff, Justin. Have a good one. We'll talk to you soon.
--:-- Justin Sealy: Goodbye. Talk soon.
--:-- [Segment transition]
--:-- John Coogan: 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. And let me also tell you about Console. Console builds AI agents that automate 70% of IT, HR, and finance support, giving employees instant resolution for access requests and password resets.
--:-- Jordi Hays: And our next guest is here live with us in person. Good to have you. Thank you so much for coming on the show. How are you guys doing? Please introduce yourself for everyone.
--:-- Chris Kircher: Absolutely. So Chris Kircher, the founder of Kircher Law. Before that, I spent 17 years at Quinn Emanuel.
--:-- John Coogan: That's your law firm. Most feared law firm.
--:-- Chris Kircher: Had a great run there. Represented Elon Musk with your previous guest, Alex Spiro, in the Twitter case. Also represented him on his comp case. I represented Citadel in the GameStop meme stock saga. I did a lot for Lehman Brothers and got very into AI a few years ago and saw, okay, this is pretty obvious how this can be used in the work we do. I started the AI and data analytics group at Quinn. And then in July, I thought, okay, it's time to take this to the next level, take my talents to Greenwich, Connecticut and build my own firm. But it's been a blast. I've got five lawyers down. I actually have gifts for you guys. Let me see. So I made, one of the things I made for my lawyers when they joined were baseball cards with their stats and their pictures. If you look at the back, you've got your, this is the Friend of the Firm series. The Lawyer series is slightly different, but I wanted to give you those.
--:-- John Coogan: I love the stats on the back. This is very, very detailed here. Friend of the Firm. Let's go.
--:-- Chris Kircher: It even has my height on here.
--:-- John Coogan: Correct. Nailed it.
--:-- Chris Kircher: Nice. Friends of the Firm. I love it.
--:-- Jordi Hays: Awesome. Great. Thank you. Yes. So do you feel like you just kind of locked into the perfect timing with this where you built the beginning of your career, landing all these super high profile cases at one of the most high profile firms, and then AI comes along and now you can take the entrepreneurial route, but not feel like you need to scale to thousands of people. For me, I love this business because we have 11 full-time people. It's nice that we can all be under one small roof and I enjoy that. But how did you think about the right time to take the entrepreneurial step?
--:-- Chris Kircher: Yeah, great question. I mean, I was having a blast at Quinn and working on the AI stuff there, but actually ended up teaching a two-week class at Stanford Law School last spring. And wandering around the Bay Area where it's even more AI-pilled than New York where I'm from, I thought, okay, there's a real moment in time here. I'm 47. So when I was in college, the internet was rolling out, broadband. And I kind of missed that pitch. The dot-com bust was my junior year or so. Went to law school, loved my legal career. But I thought when AI blew up, I'm like, I've seen this movie before. And I think I can take my now experience and expertise and use what I've learned about AI to build a new kind of firm. We don't need as many people. We can have much tighter teams, much less coordination tax, especially in something like litigation where coordination is so critical. And so find the people who are also really into AI who are in litigation who just want to figure out how do we do this. It's been a blast.
--:-- John Coogan: What does the future of self-representation look like? There are examples of people going to small claims court and standing up in front of the judge and representing themselves. It's not recommended usually, but it does happen. I believe it's legal to do that. Is that right?
--:-- Chris Kircher: It is a constitutional right. Less than 1% of cases, less than 0.1% of cases or something.
--:-- John Coogan: But take me out into the future. Is it going to be 10%? Because people are going to be like, I can just use AI. Or do you think that's not going to change anything?
--:-- Chris Kircher: I guess the way to think about it is the marginal cost of actually initiating litigation has been dropped to near zero. Before, you'd have to write a 30-page complaint or whatever, and that was hard to do. Now the AI can do a serviceable job. There will be problems with the courts, with everybody flooding in with what they call pro se complaints for people who are self-representing. And when the new models come out, I use them, and I think for a second, like, oh my God, why does anyone need me? But then I remember, oh, because, first of all, going to court, advocating in front of a judge, that's not going away. That's a unique human skill. You can't grow that in an AI pipeline. Counseling clients, right? When a client has a big problem, when I have a medical issue, I can look up why am I coughing like this, and ChatGPT or Claude can give me an answer. But if I am really concerned, I'm going to go see my doctor. And so when I was teaching the Stanford class, actually I taught it again in the fall, and one of the students came up to me and said, "So you're just like talking to it? Like you're just asking questions? I can do that." And I thought, can you? Who is going to use Claude better on a litigation task? Me, another litigator, or someone who has never litigated before and has no idea.
--:-- Jordi Hays: The pitfalls.
--:-- Chris Kircher: Right, the pitfalls, the strategy, the opportunities. It's the same thing. You ask an LLM about something you don't know nothing about. You're like, this is the most incredible genius. And then you ask it about something that's even trivial. For me, if I ask it about anything surfing related, I'm like, this is wrong in a bunch of different ways.
--:-- John Coogan: The exception would be design. People with no design skills, they go to the LLMs and they just get these beautiful designs. It's perfect. It's always like the worst.
--:-- Chris Kircher: Yeah, I was joking that people make fun of them, it's the next word predictor or whatever. When you give it all the words you have, the next word is the one you need, the next concept, the next idea. So you can extend your knowledge, but to make huge leaps is really hard.
--:-- Jordi Hays: The legal industry right now feels like all the lawyers that I talk to, at least in my network, tend to be pretty smart, pretty pragmatic, logical. We're in the technology industry, so these lawyers that are working in the tech industry, they believe in technology. I don't meet a lot of lawyers that are like, "Gosh, I don't think this AI thing is going to be that big." But it feels like there's structural reasons why certain groups can't do what is sort of necessary. And so I would say like you feel like this is an amazing opportunity for you because you can go find them. I feel like most lawyers, the smart ones, should already be AI-pilled. And then you have to assess if they're going to show agency themselves to figure out new systems and ways to use these tools. But then you also have the structural advantage where you're building from the ground up with this technology in mind versus thinking like, man, we have 3,000 people on payroll and we've got to figure out how to navigate through this. You're just starting fresh.
--:-- Chris Kircher: I mean, all those things, starting from the systems you're building on, is it better to retrofit AI into these legacy systems or should we just start from scratch? The lawyer training, people are coming at it from all different perspectives. Some people look at it like, "Oh, AI, that's interesting. I'll maybe get some time this weekend to dabble," versus "Let's use it on everything as much as we can. It's literally superintelligence. Where can we use superintelligence in our practice?" And that's hard at any big enterprise because you have committees, you have a lot of stakeholders who have reasonable concerns, but the technology is moving so fast that to have to sort of slow it down to do it by consensus was hard, is hard. And so doing it myself, not as hard.
--:-- John Coogan: What's your pitch? Assuming you want your most high-profile cases to be in front of you, what's your pitch to clients to go with you versus, when you're dealing with a really big problem and you have to litigate or you're being sued, your first thought, at least for the biggest companies in the world, is I want to save some money and go with the AI native firm. So what's your pitch?
--:-- Chris Kircher: So that's it. I think clients initially hear AI, okay, that's cheaper. I like super intelligence. We're giving a super intelligent technology on your problem that is inherently a question of intelligence. It's how do we argue this point? How do we find other arguments? How do we shape the narrative? What really happened here and how do we lay all that out? So a lot of engineers look at litigation problems and they say, looks vaguely like a math logic problem. I think we can engineer away the lawyers. But when you get to the end of a lawsuit, there's just a bunch of arguments about what words mean and what the significance of an event is, who's more credible. And so John Quinn always would say, the side that wins is usually the side that figures out what matters first. Superintelligence is incredible at helping you figure out what matters quickly and to develop on top of that.
--:-- Jordi Hays: Or I imagine even in cybersecurity, like every company is facing this right now, where somewhere in your software supply chain, there was this one tiny vulnerability. And it feels like the same thing could apply to law, where it's like finding that one detail that you can sink your teeth into and ideally make the whole issue about that ends up being, I imagine, like a way to win. And it feels like AI. But the funny thing is that both sides now have that same access to intelligence. So it becomes who can use it the best.
--:-- Chris Kircher: Absolutely. I mean, I had a case in the pandemic, a six billion dollar hotel deal where a buyer had agreed or our client had agreed to spend six billion on hotels in the fourth quarter of 2019 to close in the first quarter of 2020. That deal became impossible to finance in the pandemic. And usually those deals are hard to get around. But the issue was they closed the hotels. The seller had closed the hotels in the pandemic. That was our defense. "What did you want us to do?" But the contract said you had to run the business in the ordinary course unless you have consent of the buyer. And we hung on that and we won on that. And it was a huge deal. And in every case, there is a way. Superintelligence helps you find it faster and come up with more creative solutions for clients. Because no client actually wants to go all the way. No one comes into my office and says, "I'm here because I want to win a trial." They have a problem, and it usually goes through courts or other arbitration or some other process, but they don't really want to win the lawsuit. They'd rather just solve their problem.
--:-- John Coogan: Spiro was saying that he knows some people that do just want to win.
--:-- Chris Kircher: He does. He's the exception.
--:-- John Coogan: He knows some people that are just like, "I love winning and I want to sue everyone." And then he said he knows some people that don't want to sue anyone.
--:-- Jordi Hays: Can you help me understand what's going on? What happened with Weil, Gotshal & Manges? A 95-year-old law firm went dissolved a couple weeks ago. The Wall Street Journal here says it was a talent war, a record year for partner turnover, helped bring down Weil and is forcing similar firms to make changes to compete. Are you familiar with this story? Can you give me more context?
--:-- Chris Kircher: I followed it closely. I'm fascinated by it coming out of big law. Big law firms have many of the features of banks, but without the FDIC, without bank holidays, where you get a run dynamic where the revenue is typically concentrated in a small number of so-called rainmakers and you have a ton of infrastructure underneath that. And so when they leave, they take the revenue with them as opposed to, you would assume if you're sharing revenue with just a couple of people and somebody steps out, that's more for you. But in this case it's less. And especially in that run dynamic where you think this person's leaving, Jordy's leaving. If I'm the last one here responsible for thousands of employees, I don't have that much business. I don't want to deal with that. I'd just rather go down the street to your office and see the perfect costs.
--:-- John Coogan: Why have lawyers seemingly given so much power to individual partners? Like you would think that over the years the power would shift from independent partners to the firm. But it seems to be if you can just take your clients and go, that's a crazy dynamic. And I would expect this is the first story of many as people realize, hey, I can jump ship, spin up my own practice, maintain even 80% of my book, but I have way less costs, I'm instantly making more money. And it feels like you could have this massive reshuffling of big law over the next 10 years.
--:-- Chris Kircher: Yeah, I think historically, big law firms paid in what was called lockstep. So if we were the same year and we were partners and we were contributing, we'd get paid about the same. You've seen an explosion of free agency in law. So people going from one firm to another with big guaranteed contracts or deals. So there's less loyalty, less "we grew up in this together so we're going to save the firm" and more "well, these guys all got here three weeks ago, so anyone can turn over any time." So there's not that same level of investment. They'll bring their team, they'll bring their associates, there may be their other partners who help them service those clients, but do they need the whole firm? Probably not. I think that's what we're seeing.
--:-- Jordi Hays: What is the base case for a lawyer leaving a firm taking their business with them? Is it they're going to take 50%, 80%? Because there's plenty of clients that they might like their particular lawyer, but they also might just say, "Oh, we already have this relationship. I haven't talked to him in a while because we're not doing active litigation or something. So we'll just give whoever the firm gives us."
--:-- Chris Kircher: I think litigation is very different. Litigation is episodic. It's spiky.
--:-- Chris Kircher: You don't know when someone's going to have the big one. Usually most people have one "bet the company" moment in a lifetime. In transactional work, if you're representing private equity companies doing deals, I think that tends to be much more stable. It's sticking to the partner. And that's who they're looking to.
--:-- Jordi Hays: Okay. So if they leave, are they taking 80%, 90%, 100% sometimes?
--:-- Chris Kircher: Probably a pretty good chunk, especially if the client looks at it and sees, well, I don't know if that firm is going to still be there or my other options are still going to be there because everybody's like a bank run looking for safer ground. So you may end up in a situation where like in '08, '09, everybody's looking for the JP Morgan or Bank of America of law.
--:-- John Coogan: And is that emerging already? Is it too soon to call it? Because I understand there might be like a barbell effect where you have a whole bunch of smaller, more independent, more agile firms and then one bigger, big winner-take-all power. But also you might just wind up with a bunch of diffuse, smaller law firms.
--:-- Chris Kircher: I like the diffuse model, especially in litigation where conflicts are a problem.
--:-- John Coogan: That's right.
--:-- Chris Kircher: Partners have conflicts with one another. Historically, you would trade it off because you wanted access to hundreds of associates. But now if you don't necessarily need hundreds of associates, that trade looks different. I think the thing I'll watch is in law firm economics, most firms pay out a majority of their comp in January. And so people may be at firms kind of holding on until the end of the year to see how that shakes out. But there tends to be a lot of free agency in the first quarter. And that might be the next sort of destabilizing event.
--:-- Jordi Hays: When you're working on a "bet the company" moment, how are you puzzling together all the different strategies and how they fit together? There's companies that actively have to have earnings. They have to do PR. This founder might need to go on podcasts. They might also need to testify. There's discovery. There's so many different things going on. How 360 is your approach?
--:-- Chris Kircher: It's 360, all 360. And that's actually one of the great uses of AI is to build stable case knowledge where you can lay out, here's the client's objective. Here are the constraints that we're living within. Here are the risks that we're concerned about so that everything you do on the case is informed by those things and the AI can remind you, don't forget this person's testifying in this other case next week or they have an earn out coming up and so maybe they don't want to rock the boat now, all that kind of stuff.
--:-- John Coogan: That's fine. Yeah. Well, thank you. Last question. What I feel like is one of the reasons you have an interesting job is it feels like you're probably busy every day. But there's like what, 10 days a year that really matter where you are in the courtroom? And even in those 10 days, there's probably a much smaller set of minutes that actually matter. Talk about that kind of preparation for that and what you go through. Because it's super spiky where there's all the prep, but then it's like an athlete where you're going in and it's like a boxer, an MMA fighter that goes in. And in that moment, everything is decided. And then your track record gets decided, which is a lot.
--:-- Chris Kircher: I think athletics is the right comparison. I always would talk about football, right? Football is like eight minutes when you compress it down. When I was at my old firm and recruits would say, "Don't you guys work more hours than other firms?"
--:-- Chris Kircher: Like those are trial hours. That's like saying Tom Brady plays more football games. He's happy to play those extra playoff and Super Bowl games. So when you're in trial, it's a month of trial. You're not doing anything else. You're probably not seeing your family a whole lot. You're not catching up on your favorite podcasts. But you are locked in.
--:-- The whole game in litigation is to get stuff in your context window, in your head, so you can really outsource the whole thing to AI. AI has to be used to get in your brain so that when you're standing at the podium arguing with the judge or cross-examining a witness, you can call what you have to.
--:-- Because you can't — it's not like in a perfect world you have an AI running in real time that is ingesting every single thing that's said in the courtroom and then giving you something to look down at. That's not possible. You're there. It doesn't matter if you have 100 associates or 100 agents. It's all got to be up here, on game time. So that's what you're prepping for.
--:-- John Coogan: Do you think there will be a cheating scandal like what happens in poker, where someone in their shoe has something that's telling them to bet or not? Like Claude Shannon? Like the original wearables?
--:-- Chris Kircher: There was a moral salacious one in the chess community. I won't go into that. But yes, having answers secretly fed to you effectively — maybe not an earpiece, but something that's shocking you in your right leg if you need to press harder, shocking you in your left leg if you need to back off a point.
--:-- Right now, the method is someone sitting at the table writes a Post-it note and hands it to you. And you have to on the fly try to figure out what the hell they're saying and how do I integrate this. There's got to be a better way with technology.
--:-- John Coogan: Shut up. Shut up. Wow. Okay, so there is some interaction, so a little less room for a cheating scam.
--:-- Chris Kircher: Yeah, but maybe it's not cheating. Maybe there's a technological way to enhance collaboration in the courtroom.
--:-- John Coogan: Yeah, well, still. No, the lesson from the last week for me is the smartest, most successful people in the world will at times do the stupidest things imaginable. So I wouldn't be surprised if you see a lawyer with a hidden microphone and some kind of — the smartest, most high profile, most in the spotlight people sometimes do the stupidest things you can possibly imagine. So anything's possible.
--:-- Chris Kircher: That is what happens. That's why lawyers are important.
--:-- John Coogan: Where are you setting? You're setting up in Greenwich?
--:-- Chris Kircher: Greenwich and New York. But we're doing a California road show. Had some meetings in L.A. yesterday. Going up to the Bay Area tomorrow for some stuff at Stanford and some other meetings. Having a blast.
--:-- John Coogan: Amazing. Well, thank you so much for coming on.
--:-- Chris Kircher: Appreciate you guys having me on. Thank you so much.
--:-- John Coogan: Come back on. We love getting the opinion of the legal community at different points. I've been watching since the first episode, so this is an honor.
--:-- Chris Kircher: Amazing. Thanks for coming on.
--:-- John Coogan: Close out the show with us.
--:-- Chris Kircher: Thank you very much. I'd love to. Leave us five stars.
--:-- John Coogan: Flashbang.
--:-- John Coogan: Well, there's one more thing. Cisco.
