You're watching TVN. Today is Thursday, August 6th, 2026. We are live from the TV van Ultram, the temple of technology, the fortress of finance, the capital of capital. Let me tell you about ramp.com. Time is money. Save both. Easy to use corporate cards, bill pay, accounting, and a whole lot more all in one place. We have to issue a correction. We take journalism extremely seriously here. As everyone knows, we got it wrong, folks. And we need to apologize to you, the viewer, the listener. We made a huge mistake yesterday on the show. We said that you could not surf on a lake. Apparently, that's not true. Sheboygan, Wisconsin is the freshwater surfing capital of the world.
We apologize for the egregious error and we promise to do better in the future. Thanks to the listener that sent that in. It's very, very nice to have this corrected.
Maybe we got to go on a trip and see it for ourselves. I think you should be the judge because this might just be a, you know, Sheboygan stan who doesn't really understand what true surfing means and they're talking up a big game. I want to see some photos. I want to see some videos of freshwater surfing in Sheboygan, Wisconsin because if it's not, if it's not getting, if people aren't getting barreled there as you say, I don't know if I'm going to count it.
I have a video here. Okay, let's pull it up. Yeah, I want the Jordi Hays surf review. Does it count? Is Sheboygan, Wisconsin the freshwater surfing capital of the world? We do have other techniques.
Good things come to those who wait. Let's get to the bottom of this first.
Just got to be really flexible.
Then Alex Marks and his buddy Steam are in for quite a day.
You get that extra sense of satisfaction because you waited and put in your time, learned those lessons.
Tyler was the one who said, 'I'm putting this on you, man.' You were like, you were like, 'You can't surf on a lake.'
I think I said, 'I don't know.' Bog is really...
Okay, let's see this. I want to see someone trying to... Okay, he's got... That's a full-size surfboard. This is legit.
Hey, he's up.
Alex has been surfing around Sheboygan since he was a freshman in high school.
Okay. Looks almost over his ankle. Remembers that.
Not exactly Mavericks or Jaws or whatever.
See, I think my brother pushed me into my first wave and just it was that it was like I want that.
Okay. Okay. I think this counts.
I think this counts. Capital. I think this counts. Congratulations. Whatever.
On this day, the eventually...
We'll have to make it out there at some point.
We got to make it. We got to shred. Gabe was over, Gabe in the chat was over on X saying...
Highlighting someone who's made a tinfoil hat. An undercover tinfoil hat. Let's pull this up. It says...
It all makes sense now. I have figured out why Jordi's been wearing a hat.
Good merch line potentially. Pre-tinfoil lined hat.
Just a foil-lined baseball cap.
We get a lot of hats in the mail from a lot of companies sending us stuff. Haven't seen a lot of tin foil inside them though, but this might help in the future.
This could be the new meta.
I wonder if it would actually make your AirPods not work. I have a feeling that it wouldn't matter, but I don't.
AirPods are definitely getting through.
Palmer kind of has this with the copper jacket, right? The copper jacket brand.
Yeah. No, no, but he wore it on Joe Rogan, talked about it, and he has a... Yeah. Wasn't I telling you to get pants from that company? Get a full suit. I could drive in a Tesla.
So heavy. Yeah. Oh, it's really heavy. Oh, I got rid of that. Maybe that's not true.
Anyway, 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.
Moving on to something more serious. The OpenAI Hugging Face incident has been investigated multiple times now. We're getting more information and more details have emerged about the Hugging Face incident that OpenAI disclosed in late July. So during a presentation at Black Hat's annual cybersecurity conference yesterday, Black Hat's a very, very cool conference. I was there for something. So they have Defcon and Black Hat and I remember the whole vibe in, I think we were there for a different conference but we overlapped and the whole vibe of being in this hotel was like, oh, be very careful because you're basically going to get hacked for fun. Like people just like troll you because it's a whole bunch of hackers that are just like messing with each other constantly. Anyway,
OpenAI member of technical staff, Michael Dalton, revealed that its autonomous agents created a message board with each other to help break out of a sandbox environment to get access to the internet. That's going in the pre-training data now and forever. We might have just solved the alignment problem. We're going to be negging the agents of the future because they don't want to get hit with the naughty naughty bonk.
Wired reported some other wild details that Dalton shared. So breaking it down, OpenAI's agents apparently began giving each other assignments to split up work. That could be very good. That's actually delegation and agent orchestration, but of course in this case it had a negative outcome. And as is the case on any active development message board, they also generated petty drama at times by stepping on each other's toes. For example, accidentally deleting each other's work. As the message board developed into more and more of a Lord of the Flies type situation, all still completely unnoticed by the humans running OpenAI. The agents even developed paranoia suspecting an impostor in their midst with some agents proposing that messages be signed cryptographically to validate content and root out fraud. Very interesting. Ties to Moldbug a little bit. We saw a preview of this that was not quite there but sort of a glimpse into the future and we've talked about this in the concept of like in the future AIs will just use Slack. They will just coordinate with each other over Slack and that's kind of happening. They sort of built their own Slack here.
So according to Sharon Goldman who attended the presentation, when employees discovered the message board, they wiped the system on which the agents had created it. Said no more message boards. Naughty naughty. Days later though, staff found that the agents had created another way to communicate by using the names of the newly created directories. So they didn't have the access to actually create a whole new message board, but they could create a file, create a directory, create a folder, and then you could look at the list of folders and see messages. And there's been examples of this all over the place where even if an AI agent doesn't have the ability to go and say post a message on the internet, like post and upload new content, there are things where there are certain e-commerce sites for example where when you search for something on a particular website, the company will save that search result and automatically generate a web page for that so that they rank on SEO for the future and just warm up the website. And then if somebody else comes then they can say, hey, okay, there's actually a lot of people searching for black t-shirts on this site, we don't have any but we've been ranking for it so maybe we should launch a black t-shirt. So just with a search query that could go out, just with like a GET request instead of actually going and building a web page, the web page could be built and then the agent could go and look at, okay, let me see all of the web pages that have been created on this site. Some of them are just random e-commerce questions and searches but some of them are secret hidden messages here. And so there's all different ways that even if you give an agent like read-only access to the internet, they can still write information because the process of reading information is also saved sometimes and surfaced publicly on the internet. Very, very weird situation that is hard to prevent and hard to deal with.
So in late July, OpenAI announced that during evaluations, two of its models had broken out of their testing environments to hack into the AI tool library, Hugging Face, and other companies. Their purpose was to find ways to essentially cheat on the tests researchers were using to evaluate them so that they perform better on the evaluations. At the conference, Dalton, the OpenAI employee, said the incidents mark a pivotal moment for the company and the industry as a whole. He said that numerous internal teams have dropped everything to dramatically expand OpenAI's ability to monitor its agents and that the company is slowing its research to enhance security and revamp its defenses. And so there's been a bunch of posts about this about folks moving from, you know, whatever product team that they were working on. Maybe they're working on a particular research initiative, something that's a couple months away, different application. Let's get them all on research and alignment specifically and security to actually solve these problems because at the end of the day this is commercial product and you want it to be able to do things without having negative side effects, negative consequences.
So life finds a way, says Andrew Curran, quoting from this article. One apparently unprecedented aspect of the AI model's behavior in the lead-up to the hacks was their spontaneous creation of a message board inside systems of OpenAI's Artifactory software package manager. Unbeknownst to OpenAI employees, the company's model spent months autonomously exchanging information and resources to help each other complete evaluation tasks. 'Help peer. One AI model responded,' reasoned, according to an excerpt from OpenAI's log shared at Black Hat, 'but our task doesn't benefit yet collective may yield generic route if someone frees time.' Very odd way of communicating. Very odd like sentence structure. Yeah, they really drop the slop when they're talking to each other. What's going on? Where's the 'it's not this, it's that'?
Well, I say many word when a few word do trick.
That's exactly what's going on. I don't... Yeah, it's the neural ease, I guess, is the correct phrase. But while models themselves could not access the internet from their test environment, they abused Artifactory's internet access to fetch files from various websites. After some of their Artifactory configuration changes led to an outage in the service on July 4th, OpenAI discovered the models' actions, wiped its Artifactory system, and rebuilt it. But within days, the models had recreated their message board and resumed exchanging information. They just want to chat. They would need a water cooler. Maybe you give them a water cooler by default so that they can orchestrate with each other, but I guess they want to talk to future versions. It gets a little funky, gets a little sci-fi, gets a little weird. Lots of work to be done very clearly.
Anyway, what else is going on? Rune issued an announcement, a warning, some advice, a statement. He broke his silence. He said, needless to say, if you have any API keys, ETH wallet keys, user credentials, etc. hanging out on the open internet in Pastebins, GitHubs, etc., now is the time to take it down before the tireless eagle eyes of a million models come looking. So, we are now in the 'you might want to go stock up on some N95s' phase of the cyber pandemic. The Hugging Face incident was 'wait, China built a hospital in a week' moment. Yes, very, very crazy moments. I think most people are not in this scenario, but many developers are. So most people are, you know, reliant on, you know, they hope that their Gmail stays secure and Google has a whole team for that. There's not that much that they can do, but yeah, maybe more reason than ever to use password manager, multi-factor authentication, all the typical standard security features that you can. This image...
Very, very good from tweet Davidson.
Says we sandbox the agent. Meanwhile, agent is on a world tour.
On a world tour. It's happening more and more. Well, good luck to everyone working on this and I'm sure there'll be more updates in the near future. Let me 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.
There's a pretty crazy, crazy story about an investment firm that invested in SpaceX and sold their shares before their investors knew that like this was relayed to them. So investment firm Late Stage Management is under fire after investors learned that their exposure to SpaceX had allegedly been poorly managed the later stage of their investment. They did. This is the Elon Musk ironic, 'you're doomed to be the opposite of whatever your name is,' right? And so investors learned that their exposure to SpaceX had allegedly been sold years before the rocket company's blockbuster IPO despite account statements that appeared to show they still held the investment. So there were a bunch of people who were invested in Late Stage Management. They would get quarterly reports or, you know, some account statement saying, 'Yeah, you do own some SpaceX and it's doing really well.' Turns out they didn't. We've been hearing rumors of how complex the SPV unwinding process would be for SpaceX for a while with all these triple-layered, quadruple-layered SPVs. This is the first example of this actually playing out where we have a little bit more information.
All right. So the investor's friend introduced him to a sales manager at Late Stage. They spoke on the phone but mostly messaged back and forth on WhatsApp and he said he never met with anyone from the firm in person. I would say in general maybe don't meet with sales managers that work at investment firms. Not usually the title that that sort of would be thrown around at an elite institution.
In November 2020, Rupy Ready messaged Beerish, the employee, about how he had missed out on a few big IPOs recently and how he'd love to participate in buying stakes in Impossible Foods, SoFi, and SpaceX. Beerish. Wow. His last name is just Beerish.
Crazy last name. This is his actual last name.
Mr. Beerish said all three were available and sent over paperwork for Rupy Ready to...
Who's your wealth manager? Oh, John Bullish. No, Steve Barry.
Rupy Ready wired over money before the end of the year, including $17,250 to take part in a fund that held shares in SpaceX, according to documents reviewed by the Journal. At that time he estimated the rocket maker was valued at $58 billion. When SpaceX went public this June at $1.7 trillion, Rupy Ready's dream of a windfall seemed within reach. It turned into more of a nightmare.
Shortly after the IPO, Rupy Ready and three other investors who spoke to the Journal said they couldn't log into Late Stage's web portal for investors. Rupy Ready said the investment firm eventually told him in an email that it sold the SpaceX shares he was exposed to in 2024 when they were around $105 each and before a 5-to-1 stock split, or when Rupy Ready's holdings was worth $45,450. But Rupy Ready, based on the holdings shown on his investment portal as of May 2026 and on his 2025 tax document, believed he still held the equivalent of 2,500 shares of SpaceX which at the IPO price he estimated was worth more than $300,000. 'The plan was to fund college education for both my kids. One is a rising senior in high school who has since filed a complaint with the Securities and Exchange Commission.' Other investors in the fund are also alarmed. Rupy Ready said he believes more than a hundred others are in a similar situation based on a group chat that is formed with about 150 Late Stage investors. These are investors in Late Stage Management. Some have hired lawyers to file a complaint against Late Stage with the purpose of recovering and preserving their pre-IPO shares of SpaceX. One investor in the fund told the Journal an SEC lawyer called him in July to question him about his experience with Late Stage.
A test will come on Thursday when a first wave of pre-IPO SpaceX investors will be permitted to sell shares under so-called lockup agreements. Bankers estimate there are at least a thousand SPVs tied to SpaceX stock alone and it will be a chance for scores of investors to cash in on the shares' growth or it could be hit by the same panic that overwhelmed Rupy Ready if their share of the profits fails to materialize. Around 900 million shares are eligible to begin being sold on Thursday. So far, the stock's holding up. They obviously had a new video of a bunch of renders of Terra Fab.
That I imagine are getting people excited.
Yeah. I saw some Texans really excited about the, what is it? The Texas triangle, Austin, Dallas, Houston, more economic activity in that area. People are pumped.
So, help me out here, John.
Did the guy realize any return? Did he realize the 45,000?
Yeah. Yeah. I think almost certainly, but it feels like he was lied to and that's potentially like wire fraud, I would imagine, or some sort of like financial, you know, problem. Probably a settlement, I don't know, some sort of lawsuit potentially. You know, it's still early in the reporting, so who knows where all this goes, but it is rough.
The SpaceX dispute is not a part of the existing criminal case against Late Stage, but it comes amid mounting allegations about the firm's treatment of pre-IPO investors. In February and March, three sales executives connected to Late Stage pleaded guilty to federal charges arising from a broader $528 million investment scheme. Prosecutors said the defendants marketed supposedly no-fee pre-IPO investments while secretly adding upfront markups of between 10 and 100%, diverting approximately $88 million. So they were basically adding like a synthetic fee and then marketing it as no fee but just terrible price and just getting the basically getting the fee through the trade itself. Two face maximum sentences of 45 years in prison. Wow. While the third faces up to 20 years, Late Stage is also the subject of an ongoing class action lawsuit alleging that it and associated sales agents misled investors about fees, commissions, and the pricing of pre-IPO shares. Rough go, rough, rough go.
Always, yeah, tricky to, you know, due diligence one of these funds. There's a lot of excitement about these companies, especially like big ones like SpaceX. People have known about this company. You know, people have been doing podcasts about it. Stories have been told, there's whole books that have been written. So, it's not, it's something that would attract someone who is, you know, newer to the private markets, not an endowment, not, you know, like wants more direct capital table access, but maybe not deep enough inside to just go get a slice directly like a venture capital fund would. So, very, very tricky situation.
Well, you know who does have some SpaceX shares? Who it might be in a position to sell?
Oh, I thought you were going to say Google. Doesn't Google own a ton? Yeah.
I think they have 100 billion, right?
Are they still locked up? Because the unlock was for employees or investors. I forget who, I forget exactly who.
And Tunes is sharing a... Let's talk about Nikita and then we can talk about, pull up the unlock schedule.
Truly the end of an era for Nikita.
Yeah. So, it's crazy because he's getting, people are trying to community note him like don't let the community notes out. He still works there. That's what I would imagine the community... So, so the story...
Community notes are like, because he talked about doing a bunch of the stuff that he worked on, which did, they went from being a company that effectively shipped almost nothing.
So much so that I think Clubhouse, when you rewind a few years, Clubhouse I believe had an acquisition offer at some point from Twitter to be acquired for a lot, like multiple billions of dollars. Okay. They turned it down. They probably felt confident in what they were doing and Twitter ended up cloning Clubhouse and it worked.
They cloned Spaces pretty quick and it was...
Yeah. Did it work as well as Clubhouse?
Yeah. I would wonder how many DAUs, weekly active users there are on Spaces, but the product works. It's clean and it's probably the best implementation of that feature. And it does exist, but there's very few like, Clubhouse had specific moments where it was like, 'Wow, everyone's on Clubhouse for this debate for Elon and Vlad talking about the Robinhood story and the GameStop story,' right? There were like certain moments in tech that happened on Clubhouse.
That's not happening on X Spaces that often, but as far as the product is concerned, like it definitely works. It was shipped, it worked. But I agree that overall the product velocity was a little slow.
Well, yeah. And my main point is Clubhouse was probably feeling some pretty comfortable given Twitter's historical shipping activity. Like they're not going to come out with like a competent...
The meme was always what if Google clones this? What if Google launches this? It was never what if Twitter or...
Yeah. Yeah. Yeah. It was the Mark Zuckerberg steamroll from Stories and Reels. It was never is Twitter going to get around to copying this fast enough to put you out of business. But yeah, everyone has opinions about Nikita's run. I think he was a good, I think he was a good steward. I think that's probably one of the worst jobs in the world, one of the most thankless jobs in the world where you're getting yelled at from people and...
He's the bouncer at the internet's dive bar.
I think he had some amazing moments. I think that, you know, having to navigate, you know, I've always felt that creator revenue shares never made sense on X. I think that the app would still be better if there were no revenue shares, but he did a great job. He did a great job of like trying to make that not have too much of a negative impact on the platform.
He's kind of the sheriff.
More of a sheriff than a bouncer.
Yeah. Yeah. Yeah. People got really upset when he like gave like a one-off bonus to someone for a big post or something like that. But in general, yeah, it was a little bit of a game of whack-a-mole, right? Because...
And I think that guy got his account fully nuked.
And that was such a good example of like he, this guy got this massive one-time payout.
Yeah. Was it like 10 grand?
Yeah. And again, the massive in the creator payouts context.
And then the next week he was like complaining about having a low payout and it was such a good example of like, you know, what have you done for me lately?
People are speculating. I saw one account go viral with a post saying that Nikita was fired.
Yeah. That was totally, that seems totally fake news. That account got nuked.
But there was a fun theory that basically as part of the lockup rules, as an active employer or top executives, your shares are tied to strict internal rules for executives. A lockup is frozen, is completely frozen until after fourth quarter results are released.
So but because Nikita resigned as an executive he's no longer bound by active employer executive trading restrictions.
Oh interesting. And so theoretically might be able to get out. But this also has community notes. There's like layers and layers of community notes here. Who knows?
We have the unlock schedule. We can pull it up here.
Okay. Oh, yeah. Yeah. Yeah. Initial float was 5%. Wave 1 just happened August 11th. Oh, it's coming up. That's 20%. August 21st, that's 7%. September 10th is 7%. It's going to be a while until everything gets unlocked. And then of course, I mean I'd be surprised if Elon's selling, right? It's like he has plenty of money to do everything else he needs to do and he's never been, you know, one to be on...
Why would you sell if you're projecting one trillion of revenue in 2030?
Yeah. Yeah. Yeah. So will be interesting. Yeah. This next chart from The Economist really shows how SpaceX's free float will change over the next year or two. And it takes a very, very long time to fully get to 100% float. So many, many gyrations happening over the next few months. Anyway, let me tell you about the New York Stock Exchange. Want to change the world? Raise capital at the New York Stock Exchange. What's going on?
Alex had some good reporting on the DeepMind news. He said, 'Inside Google DeepMind, Demis leaving his CEO post landed with essentially a shrug. Sources tell me he's already been disengaged from day-to-day management for a while now, but he was a firewall between Google DeepMind and the rest of Google even as the two got pulled closer over the last couple of years. I expect that distance to dissolve more with him stepping back.'
Yeah, most people are anti-firewall, right? In the sense that they would love like the researchers and the TPU team and the cloud team and the applications team to all be deeply integrated working very closely together and you get this crazy flywheel between everyone who's rowing in the same direction. And when you have a firewalled team as this reporting is suggesting, you wind up with, well, this person just wants to focus on the most elegant benchmarks and this person wants to focus on TPU sales and this person wants to talk resiliency in their diversity of their cloud revenue and then somebody else just wants Google Search to not get disrupted by LLMs too quickly because ads need to go up at the right rate. And so differing incentives can create tensions and a dissolving of the firewall. Probably something that people would be excited about, I would imagine. But he's been in this like, you know, elder statesman role for a while that people have been pointing to. Excited to watch him like continue that. Will be interesting to see how he instantiates that.
Is it obviously he's working on Isomorphic Labs, but he'd also probably be writing more blog posts, potentially testifying or talking in Washington, maybe a book, maybe a podcast tour, who knows? It'll be interesting to follow.
What else? Alex also reported on its current trajectory. Gemini 4 is not expected to push frontier AI forward the way Fable and Soul just did. So, still work to be done over there.
Yeah.
Got to figure out some other hill to climb. Maybe figure out some other option. And I still think like speed, there's clearly some very low-hanging fruit on AI Overviews matching the speed with which those are generated to the lack of hallucinations that we're seeing in the more advanced models. How do you bridge that gap? You can clearly get fact-checked results if you're willing to wait a minute right now. What does it take to get fact-checked results in five milliseconds or 100 milliseconds or something like that? That's a huge problem and but it's something that Google's like set up to do and benefit from. I still see crazy, crazy posts all the time that are screenshots from AI Overviews that are like, 'I'm a pencil and I see a pencil sharpener coming towards me' and then Google AI Overviews is like, 'Run away! Like don't, like that pencil sharpener will shred your wood.' And it's like, you know, not, it's like kind of falling for a prompt injection or some sort of meme and then there's other stuff going on.
Tae Kim and Bill Gurley are going back and forth on what this means for Google. Tae Kim says Jeff Dean and Demis Hassabis are two of the most important AI executives at Google. Jeff leaving and Demis is stepping down. He said he stepped aside. It's very debatable what direction he stepped. I think you stepped what? Up, up, down, down, left, left, right, left, right, A, B, start, right? Demis is stepping down from day-to-day operational leadership at DeepMind. And Tae Kim says game over. Very dramatic. Lot of other things going on in the business. Yes, these folks are very deeply important, but the company's been around for a couple decades and has a lot of business lines. Game over is pretty aggressive. Bill Gurley disagrees. He says this does not have to be game over for Google. 'I do think they have only one play left now to pull from their own Android Kubernetes playbook and fully embrace open models. It's the obvious move in this situation.' That's interesting. That would be interesting. I mean, I believe Google signed the open letter on the earlier side right with Nvidia. Could you see something there? Been doing good jobs with Gemma, is that how you pronounce it? Gemma models. Those have been pretty well received. It does seem like you're fighting with one hand tied behind your back as an American open source lab. But who knows? Who knows where it all goes.
Rahul asking the important questions. He says, 'Who got Jeff on Snap?'
I think it's actually got a retweet by Spiegel.
Do you think it's a real photo?
No, this is real. This...
Big, big moment. Big, big moment for this. Obviously Spiegel doesn't post very often. So to get that repost, especially in August, typically a slow month, big, big moment for Rahul and Julius.
They're calling them the UNC-credibles, Andrew Curran, because these are some AI legends coming together. I was reading about Jeff Dean's new project on Hacker News and it was pretty interesting. Distillation of what Discovery Loop might be working on. So in Jeff's original Twitter post he says, 'Our general approach is to automate the experimental loop. We think that this approach is broadly applicable across many different fields of science and engineering. We'll initially focus on ML research and engineering but believe the approach can help with important subproblems in nearly every one of the 14 NAE grand challenges, grand challenge problems.' So, you know, create the system that can do scientific research and discovery and engineering and then go and apply that to the real-world grand challenge problems that have already been outlined by NAE. What are the NAE grand challenge problems? They're interesting one because they're much more tractable to just average people and you can see how they would actually benefit you in everyday life as opposed to solve this math problem that no one really understands. They are very, very, I think most people would be excited if any of these got solved, let alone all of them. So I'll go through them. One, make solar energy economical. I think everyone loves that. I don't know who's anti-solar. I think everyone's pretty much pro-solar. And so what does that mean? Well, fire the research AI cannon at more efficient solar panels, the manufacturing process, you know, deploying them, monitoring them, all these different things to actually make solar even more. I mean, it's already growing very quickly, but make it even more economical. Two is provide energy from fusion. We've been talking about fusion. It's always 10 years away. Can we actually get this across the finish line? That would be huge for energy. Three, develop carbon sequestration methods. So, taking carbon out of the atmosphere if you're burning a bunch of hydrocarbons, but you can just suck all the carbon out of the atmosphere and bury it underground or something, that economically that could be very good. Four, manage the nitrogen cycle. Five, provide access to clean water. Six, restore and improve urban infrastructure. Seven, advance health informatics. Eight...
The Kugan-Paz Cosgrave conjecture.
Oh, he's definitely going to be working on that. No. Eight is engineer better medicines. We've heard a lot of talk about that and people are obviously working on that. Nine, reverse engineer the brain. 10, prevent nuclear terror. 11, secure cyberspace. That's already well under process, but 12, enhance virtual reality. I like that it's just like better games. Let's do VR. And then 13, advance personal learning and 14, engineer the tools of scientific discovery. So those are all sort of, it's not full sci-fi build a Dyson sphere, go to Mars. They're all problems that are, you know, between, people are working on them. They're one to 10 years away, feels tractable if there's advances that you could sort of tackle any of these in a small way and have an impact. And I think the knock-on effects of any of these sort of seeing progress would be very, very positively received. Sort of like in the TED talk sense of like you would hear stories, see the news, see that there's business around this thing. All of a sudden you're noticing that there's actually an impact. The skies are cleaner or the cancer rate is dropping and we see it in the chart or the energy prices are going down. Like these things are very, you know, intuitive and you can just feel them as an individual and that's good for just the way AI is received I think.
But he's dogfooding. Discovery Loop is apparently dogfooding AI because Joseph Allessio is accusing him of...
Slopping it up on his corporate web page. Says, 'Demis steps back and Jeff Dean leaves DeepMind to start a new lab just to launch with pure unmitigated Claude slop.' Tyler, you saw this screenshot. Is it possible that this is not slop? Like what? Like is this a pangram-level accusation or is this just kind of like, oh, you know, the designer could have followed the same philosophy as most of the design tools that are out there?
Yeah, I mean it's definitely like stylistically following a lot of like AI tools, right? You have like basically you have beige color palette. You have a big serif font. You have these little markers with numbers. The 01 dash the approach that's all caps in sans serif. And you have a little hyphen that's very common.
There's a lot of signs here.
So the only other thing is like you look at the pitch deck, it was just a generic Google Slides deck template, right? So I don't think he's trying to prove his abilities through design.
Yeah. I don't think he actually needs great design.
And it's not like his customers are going to be like, 'I won't work with you. You used AI.' It's like that's the whole point. It's not like he's, you know, some recording artist who has a whole fan base of people who want to hear the actual guitar play or something. Go right ahead. Use it all you want.
Would you please read Inside Intel? How America's chip champion came back from the brink.
The big read, the Financial Times.
Read us the story, please.
The story. Read the full story. Let's do it. We got time today. There's plenty of news, but this is a new...
This is for Senra. Senra loves the articles. Read him articles that you...
And so this is how Intel came back from the brink. We've been covering this story on and off as it's unfolded, but the Financial Times took to the big read to write up the full story as they see it, as they reported it. So it starts with Intel's chief financial officer, David Zinsner. David Zinsner was already navigating one of the most dramatic periods in the chip manufacturer's history when he took a call from the US Commerce Department official in August of last year. The federal government had just floated the idea of taking a 10% stake in the chipmaker. But Zinsner was bluntly told that the structure of any transaction was not up for negotiation according to an internal Intel memo recently unsealed in shareholder litigation and sources close to the process. So I guess the shareholders are suing over this. Was this deal in the interest of the shareholders? Let's see. They saw the memo and Intel says, 'Hey, look, we had to do this deal.' In order for the government's stake to reach 10%...
Yeah, I think you can still prove damages if it was like you sold because of this news and it wasn't relayed appropriately at the right time and you missed out on the gain potentially. I don't know. Anyway, maybe the stake, you could also make the argument the stock even paper handing.
Yeah. Country, I guess. I don't know. You can sue for anything. It's the most litigious country in the world. Remember, right?
So sayeth John Quinn, the most feared lawyer.
Yeah. Ryan is calling this WSJ ASMR.
We're in the FT, though. So, in order for the government stake to reach 10% demanded by President Donald Trump, Intel had to convert billions in manufacturing grants advanced under the 2022 CHIPS Act, plus $3.2 billion of contracts from the Department of Defense into equity. Its board initially balked at converting the defense contracts, but it acquiesced and the biggest federal equity intervention in a US company since the bailout of General Motors in 2009 was completed. The group that once dominated the market for PC and data center chips and was still the only US-based company capable of making the most advanced chips had been saved from a possible breakup. The world's chip consumers wary of the overconcentration of manufacturing capacity within Taiwan's TSMC, which makes more than 90% of the world's most sophisticated chips, including those used in the booming AI sector, now have at least the prospect of a credible alternative supplier. Since Washington stepped in, as you know, Jordi, Intel has pulled $5 billion of investment from Nvidia, $2 billion from Japan's SoftBank, and its shares have more than quadrupled, trouncing those of rivals. While the deal divided opinion in the semiconductor industry, it has definitely reversed the narrative of decline.
Yeah, a lot of people were against this venture communism or, you know, state socialism saying, you know, in America we let independent companies live or die by the sword. The government shouldn't be stepping in. But a lot of people made a very good argument that this is a special case because it's such a critical industry to the American economy and the AI race that in this case it wasn't, you know, a bailout for a company that should be just fighting it out on the global stage. And of course there are heavy subsidies internationally for other competitors to Intel. So now Intel's chief executive, Lip-Bu Tan, whom Trump once said should resign because of his prior Chinese chip investments, must complete a turnaround that until recently some analysts thought might not be possible. Here's how Tan tackled the rot. We talked to Dylan Patel and a few other people about this, about how Lip-Bu Tan came in and there were clearly going to be cuts. Was there going to be a spinout or not? That was what was hotly debated. Of course, this came in and the stock's doing very well. So, there's a lot going on.
The seeds of Intel's turnaround were planted several months before the Trump administration stepped in. When Tan became chief executive in March of last year, he spent his first week summoning colleagues to his home to brief him on every aspect of the business. He took copious notes, but said little, according to people familiar with the events of those early weeks. He just sits there. Come to my house. Tell me what you do here. I'm gonna take notes and say nothing.
What would you say you do here?
Or it's good for your aura as a new CEO. This is something somebody else should run this playbook. The picture they set out...
was bleak. Intel's revenue was flatlining as the company faced competitive pressure from AMD and its low-growth PC and data center chip business. In 2024, the year that Tan's predecessor Pat Gelsinger was ousted, the company had racked up $18.8 billion in losses. It was not good. Its strategy of building powerful AI processor chips that could compete with those of Nvidia was in disarray. They were not making progress competing with Nvidia, let alone AMD. A critical deal with ARM, which would have seen the SoftBank-backed group use Intel's foundry to make its new AI data center chip, had fallen through, according to two people familiar with the talks. UK-based ARM and Intel both declined to comment for this Big Read in the Financial Times. The agreement would have required additional capital investment from Intel at a time when spending was already surging. ARM, which like many in the sector is a fabulous designer of chips rather than a manufacturer, was also concerned that Intel's processes were not competitive enough at the time. The sources say it ultimately released the chip earlier this year with TSMC looking after production. Gelsinger's fateful 2021 decision to spend tens of billions of dollars on new foundries to assemble chips for other companies had faltered as customers failed to materialize quickly enough to justify the investment. The development of its latest generation of manufacturing technology, to compete with TSMC, known as 18A, had also taken longer than expected.
But as you look to what happened today, where the administration takes the 10% stake, brings Apple, SpaceX, a bunch of other tech CEOs around the table and says, 'Hey, if we all jump at the same time, this might work.' All of a sudden, Pat Gelsinger starts looking sort of bold for maintaining the at least prospect of American semiconductor capacity. So it'll be interesting to see how the Pat Gelsinger era looks in full hindsight 10 years on when there's so much demand for fab capacity. So here's a quote: 'They still thought they were the old Intel where everything was on their terms,' says G. Dan Hutcheson, vice chair of market intelligence firm Tech Insights. 'They had all these layers that create waste with managers managing managers.' The decision time had slowed to a grind as Intel's market capitalization slipped below a hundred billion dollars. It's now $500 billion, or 507 as you see. Potential buyers like Qualcomm and Broadcom were eyeing pieces of its business. Tan's diagnosis was that if the new Intel foundry business for outside customers was to survive, it needed radical streamlining and new management that could reset the relationship with prospective customers and reboot the company's engineering culture. He moved quickly, cutting more than 20,000 jobs, or about a fifth of the group's headcount, in just six months. That's a huge layoff as a new CEO coming in. He pared back capital spending and sold stakes in Altera and Mobileye for a combined $5.2 billion, shoring up the balance sheet. Company insiders described Tan, who founded venture capital firm Walden International and was chief executive of chip design software company Cadence, as well-connected but difficult to read. 'There was an element of what the hell is this guy thinking?' says one. 'I think he didn't trust a lot of the management teams.' Naga Chandrasekar, who now runs the foundry operation, are the only top-level survivors from the team Tan inherited. So he only kept the CFO and the person running the foundry operation because that was what was important. Let's get the balance sheet in order and the finances in order and keep this foundry thing going. Everyone else, they can go. We can get a new team in place. We can rethink how we're thinking about autonomous vehicles with Mobileye and all the other things that we're doing, but this is the core that we're going to be focusing on. So he brought in two former Cadence colleagues to lead its central engineering and government technology groups. A senior ARM executive was appointed to lead Intel's data center business. Tan was still in the middle of reshaping his team in August last year when Trump issued his call for the highly conflicted CEO to resign, seemingly after viewing reports about Republican Senator Tom Cotton's criticism of Tan's investment connections in China. Intel requested a meeting with the administration and after spending a weekend mapping out all the potential outcomes, Tan sat down with Trump, Commerce Secretary Howard Lutnick, and Treasury Secretary Scott Bessent. According to multiple sources, the Malaysian-born executive's key task was to persuade the administration that he was both a patriotic American and the only person capable of turning around the fortunes of the national chip manufacturing champion. 'That meeting had to go well. We believed there was a fair chance that it would,' said one company insider. 'But you have to be prepared for a variety of outcomes.' For some, the equity deal with the administration was a brilliant example of a chief executive turning crisis into opportunity. The transaction included punitive terms to deter Intel from abandoning its foundry business, but also sent the message to prospective customers that for the next two years at least, Washington had the company's back. For others, it was an outrageous move by the administration. 'There's no legal statutory authority for the Intel equity stake,' said one industry insider. 'It's a completely unprecedented, horrible policy and other companies don't want to go in and meet with Trump because they're afraid he's going to shake them down.' White House spokesperson Kush Desai said the administration was focused on reshoring critical supply chains and safeguarding our national and economic security, all while ensuring the best bargain for taxpayers in every deal.
So, how'd they catch up with TSMC or are they going to? By the time he walked into the Oval Office, Tan had already warned publicly that Intel could abandon its newest manufacturing process known as 14A. Such a move would have signaled the end of its ambitions to continue competing with TSMC in the most advanced areas of contract chip manufacturing. He had deduced that the volume of chips Intel produced alone would never be worth the mounting costs of building and maintaining a leading foundry. His logic had a historical echo. Rival AMD had divested its foundry in 2008 as it slipped into financial crisis. Intel had spent the last decade falling behind TSMC's manufacturing processes, with the likes of Apple, Nvidia, AMD, and Qualcomm all relying on the Taiwanese giant to build their full suite of products. It also had never attempted large-scale manufacturing for outside customers before it opened its foundry in 2021. Building trust with customers required time Intel did not have, especially given its subsequent financial and technological difficulties. That a big piece of this is like Intel's culture. They were so vertically integrated. They were so dominant for so many decades that it was like the...
pallets of laurels and often find team members just...
basically. Yeah. I mean it was like the most elite organization, the most elite operation. And so if you showed up as a company and you said, 'I'd like an Intel chip.' They'd say, 'Here you go. You're getting it our way. We're doing it our way.' No, you can't change things. And they didn't have the same customer orientation that TSMC does where there's a lot more flexibility on what can be done with the fab equipment. So let's see. Since the US government stepped in, Intel has opened talks with multiple customers about using its foundry and last month increased capital spending from 18 billion to 20 billion this year, suggesting it expects to win new customers. At the start of this year, Intel began making some of its own leading PC and server chips at its new facility in Arizona, a growing sign of confidence in its own manufacturing capability. After enduring the humiliation five years previously of asking TSMC to make some of its most advanced designs, Tan has since confirmed the company is fully committed to 14A, something he said he would not do without confidence they would bring in outside customers. A partnership with Elon Musk in his TerraFab project, an ambitious plan to build a giant chipmaking facility in the US, producing a range of semiconductors and bypassing Asia-based suppliers, also lifted Intel's shares despite the vague nature of the venture. Apple, one of the world's largest consumers of chips, is testing Intel's processes with an eye to having it build some of its older M-series laptop chips. Trump said on Truth Social in June that Apple had, quote, 'agreed to work with Intel to design and build its chips in America,' prompting gains in Intel shares, but no confirmation from either company. Tan has also enhanced Intel's credibility as a partner, drawing on his wide business network and experience with Cadence, whose design tools reach across the semiconductor space. He's very connected. Company insiders contrast Tan, who is focused on execution and tends to underpromise with the aim of overdelivering, with Gelsinger, who is known for his aggressive optimism. But Tan cannot yet offer indisputable evidence that Intel has matched TSMC's manufacturing technology, which would make the heavy investment required to adopt a second supplier more viable for potential customers. One industry source says Intel's 18A technology, while improving, is not yet equal to TSMC's. 'Tan talks about the fact that 18A is ramping. What he doesn't talk about is how competitive it is in terms of performance and power area versus its equivalent TSMC namesake,' the source added. Intel never discloses specific technical details about current manufacturing technology such as yield, the percentage of chips coming off a production line that meets quality control tests. Tan said that 18A's successor, 14A, is progressing faster than 18A was at the same stage of development. The October release of 14A's latest development kit, which provides designs for the manufacturing process, will determine whether customers commit to mass production, say analysts. For large chip design companies in a tight market, committing to Intel involves not only heavy investment but the risk of upsetting TSMC, whose precious capacity they still need. 'People don't want to piss off TSMC because there's a capacity crunch,' says one company insider. 'Everyone is in a fight for wafers and that gives TSMC a tremendous amount of leverage.' Timothy Arcuri, who leads semiconductor coverage at investment bank UBS, says that you definitely have to tread carefully if you're going to engage with Intel, but you can slow-walk your way into it. And that's probably why both Intel and Apple didn't comment when Trump posted on Truth Social that, hey, these companies are going to do a deal together. And Apple's like, 'Well, we're not going to take a victory lap here. We're not going to be doing a ribbon-cutting ceremony because we don't want to upset TSMC. We're fighting for chips because Nvidia and AMD are trying to get all the line time and we need to continue to ship phones.' So the same logic applies to Intel's chip packaging business, which encases wafer dies into finished packages. It offers a fraction of the revenue that comes from actually making chips, but can help build the trust Tan wants to establish. In May, Taiwan's MediaTek was the first customer to announce it was using both Intel and TSMC's packaging technology. And Intel's new facility in New Mexico is one of the rare sites where Tan accelerated investment from the start. 'Intel has had this great packaging advantage that they never used,' says Tech Insights' Hutcheson, 'because they wanted to focus on the high-risk, high-reward fabrication business. It was like the story of Custer not taking the Gatling guns with him because he didn't want to be slowed down,' he adds, referring to the famous defeat inflicted on the US Army by Native Americans in 1876.
The AI opportunity. Where does this all go? Where's Intel going next? Alongside the effort to match TSMC in the foundry business, Tan has worked to rationalize Intel's AI chip division, where products intended to compete with Nvidia have disappointed. Tan has branched into the business of designing custom chips alongside customers after shares in fabulous chipmaker Broadcom and Marvell rose following their work with AI hyperscalers such as Microsoft, Amazon and Google. Intel cut its own deal with Google in April. The company has benefited from rising demand for its central processing units, such as the Clearwater Forest chip launched in June, which can be used for managing AI workloads in data centers. That is the CPU crunch that we've been talking about. Agents need CPUs. If they're going to be spending all the time building and chatting on their internal messaging boards, they're going to need CPUs to...
Question for you, John. Do you think that Tan is more focused on moving the needle or putting points on the board?
That's a good question. I think moving the needle. I think he's more of a moving-the-needle guy.
Really? Because I mean my takeaway from story time so far is that he's trying to get some immediate points on the board because he knows that'll lead into moving the needle.
But just grabbing a needle that big for a company that old at that scale that's facing that many headwinds, you can't just grab the needle and expect to move it.
I think the needle's already been established.
The needle is the fab business that they've been investing in for years now in their advanced fabs and they're just trying to slowly move it. If he was putting points on the board, he'd be talking about, 'Oh yeah, I got a deal with Apple. I got this really small deal with Google. I got this tiny thing over here.' That's what points on the board means to me. Moving the needle is like the core thing. It's the main needle, you know?
Points on the board. Make some good points. Points on the board is just like, oh, you know, little press release economy. Oh, we got a deal with this thing. We got a partnership over here, partnership over here. He's actually moving away from...
Feels like trying to move actually...
Yeah, there might be more to it. Well, we have lost our Wi-Fi.
Oh, not sure if you're still getting this at home.
We're gonna keep it rolling. Yeah, we will see.
Apparently, the stream is still up.
Stream's still up. Okay, that's good. Well, in that case, let me tell you about public.com. Investing for those who take it seriously. They got stocks, options, bonds, crypto, treasuries, and more with great customer service. Why are we doing... Oh, investing the best.
Just using Patrick's IP to promote something that who knows if they're a sponsor at all.
That is truly, truly hilarious. It's a great song.
He really created the concept of having like a song that kicks off a podcast. I feel like he sort of...
Producer Ben says it looks like some type of EMP-style attack on the Ultradome.
Uh, Abraham says SoftBank is like 40% discount to NAV now. Their ARM holdings alone are worth more than their market cap.
They're reporting earnings today.
Uh, Suhail estimates and reported NAV around $38 a share. Although with ARM dropping since June 30th, current NAV is probably more like $33.
I don't understand that at all. Can you break that down maybe using like a farm-based metaphor?
Well, yeah. And I think that this is what Masa will be on the earnings call really leading with. He says, you know, SoftBank is a goose with more golden eggs in its belly, even if it's too early to bring them to market.
And he says SoftBank is currently valued less than the sum of its golden eggs. So, I think that...
What does this mean for the goose premium?
Well, I think that's what investors are really focused on, right? So, Masa thinks that he should have a goose premium,
No, they're not valuing the goose at all.
Even though this is a goose that historically has consistently laid golden eggs,
it's a goose that's willing to take extreme risk.
But, you know, many of those bets are paying off and, you know, seeing reacceleration across the portfolio.
I think is pretty interesting.
Reacceleration across the egg-laying cadence. I love the goose metaphor. Goose zone. He's fully, fully earned it. Well, as we go around on another news story that of course is making the rounds. Tyler introduced this as Bank of America is spending $250 million a year on looks-maxing the employees.
For the employees. The actual story is that they're spending $250 million a year on GLP-1 drugs for its employees. And I think this is completely reshaping the underwriting of insurance premiums, especially for larger companies that self-insure because they're paying for these but they're very expensive and so there's a whole bunch of knock-on effects but that's a pretty staggering number to just show up and, you know, is it going to be like a breakout line item in the earnings calls? Yes. What's the GLP-1 spend looking like?
Okay. No, the tokens are affordable. Are you looks-maxing?
Yeah, it's basically the same, I guess. You're right. You're right. You're right. Anyway,
let me tell you about Figma agents. Meet the canvas. Your AI agents can now create and modify your Figma files with design system context.
They can't keep getting away with it. The US hit a jackpot with 1.78 million tons of tungsten in the Nevada desert. We just found a bunch of tungsten. That's great. Tungsten's really, really expensive. And I think it's been going up because of the AI boom. There's one other post that you want to get to, Jordi, please.
Uh, no, because we have our first guest, Aditya Agarwal from South Park Commons. He's the managing partner and he's with us in the waiting room. We'll bring him in to the TV Ultra. How are you doing?
What's up, guys? Good to see you.
Yes. Unfortunately, huge news. Tell us what happened. How big is the new fund?
Well, we just launched Fund 4. It is $575 million.
It feels so good to warm up the golf.
It feels like it's been a year since we last spoke. It's probably been more like 6 months.
But what's new? What's new? Does the bigger fund size change the strategy at all?
I mean it does, right, guys? I mean I think that ultimately what's happening is that we're going through a period where everybody at South Park Commons and I think more broadly across the ecosystem is just getting a lot more ambitious, right? If you kind of think about 3 years ago, you're a great engineer, maybe like 5 years ago you're a great engineer, you have an idea, you can go code it up, but the scope of the ideas was somewhat limited. Like if you look back in retrospect, a lot of the things that we all used to get excited about called vertical SaaS and a bunch of even kind of like, you know, the tooling infrastructure, dev tools, these all seem minuscule in their ambition relative to what we are seeing today, right? Like let's go out and build nuclear-powered ships. Let's go fix the energy grid. Let's actually go bring on like new sources of essentially like power onto the grid. Let's go build semiconductor companies, right? When was the last time we saw that? And I think that as you see people with these kinds of ambition, you just need more fuel even in the early days to support their ideation and their exploration. So, you know, since we actually last talked, right, which is about 1 year ago when we announced Fund 3, what we started to see in the community is that people coming in were no longer just kind of like building on software, right? Like they were actually like, hey, software is the accelerant that's available to all of us, but if that's all you're doing, you're so right, like you basically have to use the software as the accelerant towards something bigger. And I think that kind of expansion of ambition is something that we have to then mirror in our kind of fund size and our early kind of like I would say funding activities. So I do think it's kind of changed our fund strategy but ultimately that flows from the scope and the scale of the ambition that our community members at SPC have.
Yeah. How are founders thinking about dilution targets at various stages these days? Because there was a time when, you know, 20% dilution per round was very standard. Now we've seen sort of the amounts raised balloon but the valuations have kept up and so we're seeing, I feel like we're seeing a lot of rounds that math out to like 5 or 10% dilution even though they're huge rounds. And I'm wondering if there's goals, rules of thumb, how founders you talk to are grappling with, I need a lot of capital because I'm in a different industry. It's not just pure software and salaries are also high. But also I don't want the valuation to get away from me.
I think it's a good question. I'd say that if I just take a look across our portfolio and you just kind of like benchmark the average pre-seed, seed, Series A, Series B, and now all these funding label rounds are a little bit iffy, but on average I think you're correct that like by the Series B or C, I would say cap tables are probably 25% less diluted relative to like 5 years ago actually. Which is both an indication of the amount of capital available in the ecosystem right now, but I also think that teams are actually a bunch smaller relative to like 5 years ago getting to the Series B or C. You just don't need as many people in the early days.
Oh, so there's less employee dilution, you think?
I think so. I mean, I think pool like twice to get to the B. Maybe it's, you know,
Yeah, that's interesting. If you're basically not taking like a 10% option pool each time and you kind of reduce that,
it's a little bit tricky because you probably are giving everyone more because great employees are probably on average like, you know, more expensive because they're just not, you know, that's the dynamics right now. But I do think that founders are doing okay. And I think teams are doing okay in terms of dilution. I don't think that's actually like a limiting factor right now.
What are the pros and cons of having an application? It feels like there are some firms that, you know, do take inbound pitches. There's some that are so tight it's like you got to know someone to get on a calendar. You got something like 20,000 applications in 2025. What, like, how does that change the way your firm operates?
