OpenAI to Remain Non-Profit, OpenAI to Buy Windsurf, Jon Voight's Plan to Fix Hollywood, Kelvin Yu, James Blom, Santi Ruiz, Jacob Kimmel, Fil Aronshtein

You watching?

TVPN today is Tuesday, May 6, 2025. We are live from the temple of technology, the fortress of finance, the capital of capital. I love those cards. What would you say you're doing, Jordy?

I'm keeping my cards close to my chest, John, because if I don't, our public information will be.

It's kind of a Russian roulette scenario. Yeah.

One of these cards, one of those.

Cards is actually our card. And we are live. So we can't edit it out or blur it after the fact. But that's a great opportunity to tell you about. Ramp time is money, save use, corporate cards, bill payments, accounting and a whole lot more all in one place. Let's go. Anyway, the COVID of the Wall Street Journal. We have some breaking news from the COVID of the Wall Street Journal. OpenAI has abandoned the planned for profit conversion. We are devastated.

AGI, the real story here. AGI lost the top spot to the Met Gala.

The Met Gala did, actually. It's like really, really important. 300 billion dollar story. A couple people dressed up and had.

A party in Costa.

Which one's more important? Yeah, which one's more important? I don't know. The Met Gala is important. I believe it's. I believe it's very high margin event. I saw that it was $75,000 to go to the Met Gala. And I love that because I think all that money goes to subsidize my opera tickets.

Yeah.

And I think that otherwise the Met would not be the. They're not profitable just on the opera tickets. But that's key. That's obviously like the most important thing that they do. So the gala raises a bunch of money for opera and. Yeah. Let us know if you'd like to join us at the opera. I'm considering putting together a TVPN meetup. Maybe at the Met. Maybe at the LA Opera. Maybe at the San Francisco Opera. We'll have to get everyone together. It'd be great. Anyway, I'm super frustrated by this situation.

Looked fantastic in the tux last night.

He was at the Met Gala. His company can't convert to for profit. He's hanging out with maybe. I think the Met is a nonprofit.

Founder hanging out with another nonprofit nonprofit organization.

Yeah.

Makes sense.

Yeah. Maybe he's getting some tips. I mean, I. We've been very pro for profit conversion. I would like to see the Met convert to for profit. I want to see PETA convert to for profit. I've said this before, but PETA is such an interesting business case because they Obviously know where all the most delicious animals live. They know what, what cuts are the best of those animals. They know the cooking temperatures and the cooking times. Yeah. And so if you were to do a 4 Prop Conversion PETA, you could unlock billions in shareholder value. And obviously a similar story is unfolding at OpenAI, but looks like they're taking one step back before maybe they take two steps forward. Who knows?

Yeah.

Anyway, the news we will read through, OpenAI abandons its plan to convert into a for profit. And this was always kind of difficult to wrap your head around because they were never going to get rid of the nonprofit.

Yeah.

It was never a full conversion. The nonprofit would cease to exist and it would become a for profit. It was more like the for. The nonprofit continues to exist and fulfill its mission and just has the most banger asset on its balance sheet of all time, which is a ton of equity in, you know, a high growth, you know, multibillion dollar consumer tech company, which is the best thing you could possibly do.

The big question was, was always that the, that the investors in the for profit unit were capped at 100x.

Yes.

Which is still kind of a big question because there has to be certain investors that are already at 100x.

Yeah. And yeah, we were joking about like do you sell your shares and then it re triggers another hundred x or is the individual share like a hundred x maxed out? Like how does that work? Because yeah, if I buy from like, like an employee, many employees have probably already a hundred x, although I don't even think they're subject to that. I think they might be on a schedule.

I think the only fascinating entity that could figure out what's actually going on. And ASI. Oh3.

Oh3.

Maybe not ASI, but I think O3 could do it if you just really dumped in all of the. Every single.

And yet it seems like O3 is not, is not up to the task because OpenAI has abandoned the for profit conversion.

Yeah.

And so you would think they would have put ASI or AGI on the task, but they came up short. And I do wonder what this means. I wonder how temporary this is. This feels like a bump in the road and we are eventually. It just feels unfathomable that, you know, a billion people are going to be daily driving the ChatGPT app, subscribing for $20. They're going to have ads in that thing. It's going to be like this, like, you know, generating billions and billions of dollars of like profit and revenue and cash flow and it's still going to be in this wonky scenario. Like at some point it's just going to become a normal tech company. I would have to imagine like that just has to happen. But it's going to be a long road.

This may be less eventful than people were expecting and I would imagine it's less eventful for Sam considering he was trying to turn it into a for.

Profit, which would be an eventful, which.

Would be an event.

So yes, yes, the event not happening is less overall.

It, I mean it's a, it's an interesting situation because it seems like all of the main actual OpenAI stakeholders are on board with the conversion and it's glorious.

Well, we're going to go into that. There's an article on Microsoft in their position.

Yeah, I guess Microsoft is setting it up.

Yeah, I mean there is a question as like, you know, the nonprofit's goal is they have a fiduciary duty to humanity. Their goal is not profit maximization. And so is the nonprofit board able to make an argument that this spin.

Out is good for of humanity?

Yeah, I would argue that yes it is obviously. I think that, I think that for profit companies have financial like fantastic benefits to humanity. And I think capitalism is the greatest system and benefit to humanity in the world. And so I think every nonprofit should for profit if they want to benefit humanity. But who knows, you know, the internal politics of Open AI, the, the nonprofit, the for profit all extremely complex. So let's go through a little bit of this. The move could complicate the company's future fundraising efforts. Although they've been able to raise a ton in this weird scenario already. But I think eventually investors will potentially get tired.

SoftBank's the terms of SoftBank's original deal $30 billion round was contingent on the for profit conversion.

On the conversion.

Now I believe what we're seeing is that the investment will still go forward somehow even though the conversion isn't happening. But certainly it's in the interest. OpenAI just has so much momentum and traction. I saw Will Depew shared yesterday that ChatGPT is the fifth largest site on the Internet. So they can basically effectively raise on any terms that they want, just given how much momentum the business has. But as an investor, if you're putting in billions, ideally you have a pretty clean structure. Right.

Especially at this scale.

Once you get into that scale, yeah, it's hard to. This is not no longer becomes a flyer. It's like Mazda is leveraging the entire SoftBank balance sheet. You know to make this investment.

And yeah, it's very different than, oh, there's betting.

He's betting the company on, in a.

Way, it's going to test something out. We have a handshake deal. Yeah, I wired in the money. We have a safe. And is the safe, like, real, the most ironclad contract? Probably not, but it's fine because it's like, you know, 1 on 10. But this is 1 on 10 trillion at this point. Or 0.5 on. It's 5 billion on 500, something like that. Anyway. So OpenAI started to work on a change to its business structure after CEO Sam Altman's surprise firing and reinstatement in 2023. Its big investors, including Microsoft, watched its temporary ouster from the sidelines, unable to wield official power.

The outcome that was so wild because it seemed obvious that Satya had no real hand. And in hindsight, I think we know he wasn't involved with that process. And so. And at that point, what he had just invested, I think $10 billion or something like that.

Yep.

And so to make that sizable investment and then watch the CEO get booted without your involvement, we should do a.

Historical timeline review because you can search your ex timeline from those dates and just see, like, what was your timeline? All the people you follow, what were they posting? Because it was a crazy time and it really, like, the consensus was like, oh, it's a zero. Like, given the chaos, like, this is clearly not a functional company and like, everything will break, everything will fall apart. But gotten to a place where it was like, it was even. It was a going concern. It was a company. Yeah.

The best content, I think retrospective on that period is a podcast that Jessica Livingston did with, with Sam. Yeah, just about that kind of three, four day stretch. His reaction to it was, yeah, there's.

Two books coming out about it too.

Yeah.

And I think Sam participated in both of them. We, we read one of the excerpts that was in the journal and I really didn't like some of it in the sense that it just didn't go deep enough. And it was just, it really, like, it just was like, oh, yeah, like the nonprofit board, like, they asked these questions and. And it never went anywhere further to ask, like, okay, were those questions reasonable at all? Because the whole framing was, oh, Sam wasn't doing enough about safety and like, he let chatgpt out into the wild and it just wreaked havoc. It's like, okay, like, what, what has it done other than just like generate a bunch of revenue and help people do their homework and prep for business calls. I don't know. It's just like I have not seen the damage. I still have not seen the damage.

It's interesting.

Maybe it's Corpus. Ellen Toner.

Ellen Toner, has she commented at all on the sycophancy crisis?

Yeah, yeah, yeah. Like, like that would be a good angle if she's, if she, if she really has stuck with it. But who knows, she might have had a complete change of heart. Like we don't know. Like the original position was, was ridiculous. It truly was ridiculous. Anyway, the conversion would have changed open business to a public benefit corporation, which is interesting. Not just like a vanilla C corp. So there's like B corpse and good corpse and it's some, some rig and roll.

I might be totally off here.

I don't know.

To my knowledge, a public benefit corporation is mostly branding.

Yeah.

Just says it's legally required to consider the public good and its decisions alongside its profit making goals.

Which is so weird because you just naturally being a public company like you should need to consider the public good because if you have don't, you're bad pr. That's going to destroy shareholder value.

They also. So it's basically they're profit focused.

Yeah.

But they also have to have a legally defined public benefit purpose. So this is not even the same as this sort of 1% for humanity type programs where you're sort of legally, you're basically getting certified that you're going to give, you know, a certain percentage of your profits.

How about this? You. You legally have to have something that benefits the public. How about a stock that anyone in the public could buy and it goes up their portfolio? It's not that complicated.

Yeah.

What about, what about a website that you can go to to get intelligence too cheap to meter? What about that? Is that.

Yeah, well these were originally very popular with, with the whole sort of sustainability.

Yeah, I know, I get it, I get it. It just, it just, it just seems like it's like a complete like end run around just like actually thinking through the how a for profit like benches.

A public benefit corporation. So they're benefiting humanity with tasty ice cream.

Yes. Yeah. I don't understand it. It doesn't, it does not make a lot of sense to me.

Patagonia is the most high profile one.

Aren't they even more complex than that? I thought that they were like actually like co op.

Well, yeah. Could they give.

Generate meaningful like the members or. That's rei, I think. Rei, if you're, if you're an REI member, like you're entitled to like votes and dividends just as like a customer, which is kind of interesting. I don't know. I mean that's the beauty of American capitalism. Like you can do anything. You can go and start a co op and you can say, you know, we're not going to have a traditional cap table. Like we're going to, you know, we're going to dividend. Like the employees are going to own 100% of the business. Like you can do that. That's great. Yeah, you usually get smoked, but you can try it. And so you know, all the, all the communists and stuff can just be like, yeah, you can go do communism over in that corner of the of the world. Like that's fine. Like you're good.

Yeah. The shield shared yesterday the OpenAI apparently updated their structure section of their website because there's so many different SPVs and stuff at this point going into OpenAI. They they create this huge disclaimer on the site. Investing in OpenAI Global LLC is a high risk investment. Investors could lose their capital contribution and not see any return. It would be wise to view any investment in OpenAI Global in the spirit of a donation with the understanding that it may be difficult to know what role money will play in a post AGI world, which is hilarious. Marketing, combining sort of brand marketing with, you know, with your legal disclaimer. The company exists to advance OpenAI's mission of ensuring the safe artificial ensuring that safe artificial general intelligence is developed and benefits all of humanity. The company's duty to this mission and the principles advanced in the OpenAI Inc. Charter take precedent over the any obligation to generate a profit, the company may never make a profit. And then it's just other legalese.

But I mean the way I think this plays out is the consumer tech company spins out, becomes a for profit. The nonprofit remains extremely well capitalized. Probably the best capitalized nonprofit in history because it will have something like 30% of a $300 billion company and that will produce cash flow and dividends and secondary sales and all these different things. So basically this nonprofit can do whatever they want. They can hire all of the best researchers to go do AI safety research and then I think at some point they're going to discover something new and they're just spin out another for profit. I'm not even kidding about that because when I think about like. Like what? Like I'm not very doom pilled but I do think that there's a huge value in AI safety research for sure. Just like there is a huge benefit to cybersecurity research. Let's not, let's not let cyber attacks happen. Same thing with defense technology. Let's not let anyone use planes to hurt humans. Let's not let them use guns to use humans. Let's not let them use AI to hurt humans. I'm. These are, these are all in the same layer of abstraction. And so what happens when OpenAI Nonprofit becomes really, really good at preventing AI from attacking humans? Sell it to the government.

Yeah.

Become a, become a defense contractor.

Yeah.

It's almost taxpayer will pay for that.

It's almost more interesting to cap the non profits, you know, basically royalty or whatever. However the sort of transfer of value happens and just say, hey, you have a billion dollars a year to do AI safety research forever, and you can spin out for profit entities. But this is your sort of.

I just, I still just don't understand why we need to do safety research in the context of a nonprofit. It's like, I'm an American citizen, I'm a taxpayer, and I don't want to be turned into a paperclip, therefore take my tax money and bid out a contract to a bunch of different tech companies to prevent paperclipping. Just like you prevent terrorism, you prevent war, you try and prevent all these things, these are massive problems. So they need to be centralized through a big government project, but they can still be bid out as programs record from the dod. I think that's the future anyway.

You could be onto something, John.

Who knows? OpenAI said it scrapped the Boulder plan after discussions with civic leaders and the attorneys general of California and Delaware who would be required to sign off on it. The company will instead transform its for profit subsidiary into a public benefit corporation that is controlled by the nonprofit parent, the company said Monday. And so I wonder what control actually means in this context, because obviously there are, there's the nonprofit, there's Microsoft, there's investors, and there's employees. And so typically, if you just think about the nonprofit, the for profit stakeholders, you would think there'd be some sort of mix of those. But it seems like the nonprofit will have increased control. And then of course, there's the discussion over who's on the nonprofit board, who has control. And the nonprofit board has been changing around. I mean, I'm not actually sure. You might want to look up who, who's on the nonprofit.

So I'm pulling up a graphic right now from OpenAI's website. Michael has it basically over giving an overview of the entire kind of corporate structure, which I think is just helpful to revisit. But let us know when it's up. Michael.

So the change is a win for Musk and other OpenAI critics who believed that the company has strayed too far from its founding altruistic mission. OpenAI's ChatGPT has become lead contender in the global AI race, amassing hundreds of millions of users.

So here you go.

This is the most recent.

This is on OpenAI's website on stock.

The board of directors.

So you can see it on your, on your screen as well. But you have the board of directors at the top that controls OpenAI Inc. Which is a public charity, the OpenAI nonprofit.

That's nonprofit.

And then that charity owns a holding company for OpenAI nonprofit plus OpenAI employees and investors.

You know what? This, this, this, this reads like the nonprofit owns the employees. Like that's technically what this chart says. It says it owns a holding company for OpenAI nonprofit. It owns the employees and it also.

Owns the investors, which is maybe basically true.

Maybe it's not literal ownage, but it's like in the gaming context, like they completely own the investors. Like the investors got owned.

Get owned.

Get, get pwned.

Yeah, get pwned. And then the employees and investors are coming in and then they are the.

Majority, majority owner of the OpenAI Global.

LLC, which is a cap profit company, OpenAI GP LLC. I somehow is controlling the holding company for the Opening Eye nonprofit. And then the holding company is the majority owner of OpenAI Global LLC, which.

Is, this is what people when they say they're too smart to make money in crypto, too dumb to make money in AI because like, I don't understand this. I could not make money in this market because of how it, right now I'm literally too stupid to understand this.

No, I mean, I feel like you could write, you could write a book on OpenAI. Yes, those are happening.

They're happening.

Just writing a legal case study on OpenAI and all the thousands of different decision making decisions that got us to this point, which is fantastic. And it's kind of funny because you just have Microsoft just on the left off the org chart, basically just being like, yeah, we're just going to come into the one thing Satya is like, yeah, I just kind of want to own OpenAI Global.

Los.

Yeah, I just want to own the products.

And what does GP stand for in this context?

I think a partnership.

OpenAI GP LLC. Wow, we really did our research on this.

No, I mean there's just so much.

And then there's ownership and control. So the nonprofit owns, wholly owns the LLC which controls the. This is so complicated. How do they come up with this? It's amazing.

It's honestly Art Musk's Elon. The Journal says the move is a win for Musk, who runs a rival AI company, has sued OpenAI over various issues for more than a year. Musk, as everyone knows, has not been in favor of the for profit conversion. I think specifically takes issue with Altman getting it, you know, meaningful stake in the for profit. But Mark Toboroff, must lead counsel in the lawsuit against OpenAI says this changes nothing. The move is a quote, unquote, cosmetic restructuring that converts charitable, charitable assets into private billions. The founding mission remains betrayed. So I don't. This. I don't think this is going to sort of solve the legal battle between Musk and OpenAI, but it's certainly. I think that the result here is that it allows. It will allow new investment to kind of move forward, which feels like the kind of thing that you wouldn't want to get bottlenecked for too many months.

Yeah. I mean, the public benefit thing seems like a reasonable end state for this. Right? It's like, you can build a normal business. You can have investors, you can have dividends and stocks, options and stuff. It is. The dynamic with Musk is a little interesting. Like, I feel like Elon should either just make his ask very clear. Like, hey, I donated millions of dollars. I think, like, hundreds of millions to the nonprofit. I don't own a single share in the new for profit. Why? Like, give me. Get me on the cap table, get me equity. Like, that seems fair. I think a lot of people would be like, oh, yeah, like, then you're not just like, you know, going around suing, like, kluding up, like, something that's working. You're just like, yeah, like I was earlier here. I've been early. I'm like, you know, the word of Savarin of this thing, like, I was like, around, like, give me, you know, count me, like, pay me, you know, it's totally reasonable. Ask on that front. And the other side is like, Elon hasn't started another nonprofit because if he's really worried, if he really does believe that, like, the stated preference is, hey, like, like there needs to be an AI nonprofit. Like, this is essential. And my problem is, is that the nonprofit is becoming for profit. And that's my beliefs are in how the world works and how we're going to prevent AI doom. It's like, well, you're the richest man in the world. Go start a nonprofit. Right. Like and so instead with the XAI xa, I totally think it's fair to start XAI and compete in the for profit race and just go do that.

That's great.

And there's tons of benefits to X and to Tesla and tons of stuff we've talked about. But where's the next nonprofit? We should be competing on both sides, right?

Yeah.

What are you reading now?

I put the chart into O3.

Okay.

I said can you explain this like I'm five. It took it very literally because it's funny. So the board of directors is the quote unquote the grown ups in charge.

Okay.

They're like teachers who make the playground rules so everyone plays nicely.

Okay.

The OpenAI nonprofit is the caring parent club. This is a special club that cares more about doing good than making money. The teachers tell this club what to do.

Okay?

Okay. The big toy box is the holding company. The caring club owns a big toy box inside. It keeps everyone's toys and lets workers, employees and friends put their toys in too. So they all share the helper grown up. OpenAI GP this is a small helper that the caring club owns. Karen Club owns. Its job is to help look after the toy box and the money making shop. Next item. The money making shop with a lid so complicated.

It's already more complicated.

The shop can earn money but it has a lid that stops it from getting too greedy. It's profits are cap. The big toy box is the shop's main owner. So most of the shops pennies still.

Serving the box owns the shop.

So most. So most of the shops pennies still serve the caring club's mission.

Okay.

Microsoft's tiny slice slice. Microsoft is like a friendly neighbor who gets a small slice of the shop's penny. But it doesn't run the shop.

Okay.

Anyways, that's pretty straightforward. Still confused.

Still a little confused.

But I can imagine a five year old would be even more confused.

But thanks for the metaphor.

Good effort though. Good effort.

When it became clear that it was going to need more money than its founders thought, OpenAI created a limited liability company subsidiary that would allow it to take investment from big tech companies such as Microsoft. Those investors returns would be capped at 100 times what they put in and could fall to zero if the nonprofit board deemed it was necessary to support its mission. High risk, high reward. Yeah. I always found it kind of beautiful that AGI like the fate of AI is like if you want to unleash the, you know, the, the ASI future. Like you must be hyper capitalist. You must raise $100 billion.

Yeah.

You can't just, you can't just be.

The techno capitalist revolution.

Yep. It really is like the march of techno capital is what gets you to AGI. It's fascinating.

Hyper commercial.

Yeah, exactly. A lot of people were thinking like no, maybe it's just some, some brilliant algorithm thought in like the independent and you know, some one person thinks of it. E equals MC squared could come up with it and like it unleashes this AI revolution. But no, you actually need to coordinate all of humanity. You need to make money. Yeah. It has to be profit. It has to be profit driven. It's fascinating. In the new structure proposed Monday, the capped profit LLC would be replaced by a public benefit corporation. I like to hear that we're moving from LLC to corporation. When would serve which would serve both a public mission and investors. In a note to employees Monday, Altman.

Hopefully that eliminates the weirdness around the capped the whole cap profit dynamic. I never like I don't know how.

Capping profits Not a fan.

It makes it much more difficult to underwrite if you can only get 100x. Yeah, right.

Yeah.

But 100x's don't get masa.

Is it possible in the morning. Has anyone done like some sort of synthetic levered SPV that's like 100x long.

Open AI something there. Especially if you put it on chain.

You can get, you get 10,000x return if you. If you're 100x long.

The more complicated the structure, the more complicated the financial products. You can build on top.

I feel like, I feel like Wall street and the hedge fund guys really aren't aware of the opportunity here. We got to, we got to bring finance to San Francisco. Yeah, save San Francisco. Save it anyway. Altman said public benefit corporations have become typical for AI companies such as its rivals Anthropic and Musk' oh, interesting. I didn't know XAI was a public benefit corporation. That's cool. OpenAI board chairman Brett Taylor said during Monday's press conference that the new structure would be simpler and allow employees, investors and the nonprofit to own parts of the public benefit Corporation. That's all the different stakeholders that are already looking for a slice. He declined to say how large of a stake the nonprofit would have in the public benefit Corporation. The potential size is being studied by independent independent financial advisors. Altman says he expects the moves to result in the nonprofit being one of the largest and best capitalized Charities in the world. Very cool. And it means we can start deploying capital from the nonprofit very soon, which we're thrilled to do. We've got a long list of things that we think will be very impactful and I do think that's like an undertold story right now just what the nonprofit will wind up doing. There's a ton of interesting things that they can do in AI research. I still actually think that they're going to discover something and then want to commercialize it again because that's just the nature of putting a bunch of, I mean, I mean the question, the question there researchers like go do whatever, of course discover valuable things.

The question there is just. Wouldn't OpenAI, just the Public Benefit Corporation just have kind of effectively a rofer on novel IP and product opportunity?

No, I don't think the, I don't think the.

I'm not saying actually legally. Yeah, yeah, but would they not.

So, so, so I think this is like the, the weird dynamic where the. Now there, there are two open AIs that you can go work for if you're in AI, if you want to work for the product company, the high growth, you know, studying churn and product and how do you actually commercialize this? How do you make this an amazing company? The next Facebook, the next Google. You go to work for the Public Benefit Corporation, you're working on ChatGPT, agents, operator, deep research, you're having a great time, you're you know, the employee basically. But if you really don't care about that and you want to work at the nonprofit, you want to work on AI safety, you want to work on foundational research, you head over to the nonprofit and all of a sudden your job on a day to day basis is much less pressure driven. You don't have the same KPIs, you're just doing research. But what happens when you put a bunch of insanely intelligent people together and give them no constraints to just do whatever they want? They're going to come up with something creative like they did before.

Okay, but I just don't know a single person that works at the nonprofit. Technically.

What really do you know? No, but, but, but, but, but once they're capitalized like they will start scaling that staff with but their mission won't be like the, the mandate for those folks won't be focused on go build products, go implement, you know, GPT4,5 in some tool. Let's figure out coding, let's figure out agents, let's figure out booking. It's Going to be something much more abstract and foundational. But you put those people with a ton of resources and you let them turn them loose. Like they're going to discover cool things that will probably be commercializable anyway. We'll see how it unfolds. Maybe that's not what happens, but I don't know. Anyway.

Yeah. Did you already share this? But Altman says it means we can start deploying capital.

What does that mean? Yeah, yeah. Deploying capital into research that the nonprofit spits out. What will that look like?

Thinking studies?

Yeah, exactly. Yeah. There's going to be stuff that's just like policy papers. Maybe those aren't commercializable. There's going to be stuff that's just like a critique of how LLMs are being implemented across the world. Right. There's going to be stuff that's not commercializable. But if they start touching research like they're going to, they're going to create value.

I would still go, I would still expect Altman to be looking over and be like, hey, you know, we could get that to a billion of arrows if you want and we'll give you a cut.

It's going to be complex. Like let's say in the nonprofit they're like, hey, yeah, we actually found a new algorithm. It's like amazing at booking flights. Like never makes a mistake.

Just like we did the meme. We did the meme.

No, this was supposed to be at the for profit. Why did this happen? The nonprofit and there's just some brilliant researcher over in the nonprofit being like, yeah, I was just looking for the most elegant algorithm possible. And then I scaled it up and now it just does everything.

He's like, wait, I'm goated.

Anyways, you know, we're all, we're all rooting for open AI. We love for profit corporations. And if you're looking to invest in for profit corporations, head over to public.com investing. For those who take it seriously. They got multi asset investing industry leading yields trusted by millions. Go to public.com I was hoping they, you know, they say they have multi asset investing. I'd love to be able to buy a slice of a non profit multi asset investing. Invest in non profit so that when the for profit conversion happens, you cash in. Because that's what nonprofits to me are really about, cashing in.

Well, Microsoft owns part of OpenAI.

That's a way to get exposure legitimately and they own like a significant chunk. Yeah, for a while I was thinking about like another lens.

You got to buy Some of teams though too.

Yeah. No, no. There is an interesting investment thesis for the mag 7 which is basically like if you want exposure to self driving cars, virtual reality, artificial intelligence, autonomous delivery, like what should you buy? You could go around and buy slices of startups that may or may not work. Or you could buy Google which owns Waymo, Facebook which owns Reality Labs and Oculus. You could own Microsoft which owns OpenAI. You could own Amazon which is probably going to figure out delivery. Tesla Humanoids, right? Like the Mag 7 have pretty good exposure to the next trends as you list them out. Of course they're not pure plays and they're not going to and it's hard for them to do a thousand acts. But it's not not a crazy idea to just go and buy Mag 7 if you're looking to index on the next the next wave of 20 year tech trends. Who knows, some of those might be disruptive though. Some of them might be get get rocked. But so far looking pretty good. Anyway, the information dove deeper they say Elon Musk might have won the battle over OpenAI's future, but who's to say Sam Altman isn't going to win the war? Today's announcement by the Chat GPT creator that it was abandoning plans to move its for profit arm out from under control of the nonprofit, conceding to Musk's legal efforts to stop the restructuring could end up cementing Altman's power at OpenAI. After all, OpenAI will remain a company whose controlling shareholder is the nonprofit. So whoever controls the nonprofit controls OpenAI. The information writes. OpenAI's existing board will decide who sits on the nonprofit's board in the future. Is this like a one hand washes the other scenario? What's going on here?

Well, the key thing to note here is that it's in this article shares it's reasonable to assume that Altman has a lot of sway with the existing board, which is very different from the one that fired a lot 2023. So they already did effectively turn it over and align it with the for or the public benefit.

There's 10 members including Altman on the current board which will appoint the nonprofit board. So makes my head spin. It's hard to imagine that OpenAI that as OpenAI gradually added new members to the board, Altman let anyone anti pathetic to his ambitions join. The only potential wrinkle in Altman's effective control of OpenAI going forward is whether state regulators get a say in nonprofits in nonprofits administration. Barring that he Remember, this was why they.

This was the real reason that it seems they halted the conversion is because Delaware and California were not seemingly not too excited about what was happening.

Who is the current board? Adam. Is Adam on there? Okay. Yeah. So Brett Taylor, who's the chair? Adam d' Angelo Sue. Desmond Hellman. I haven't heard of, of her. Need to dig in there. Zico Coulter, retired retired U.S. army General Paul Nakasone, Adebayo Unglesi. Nicole Sigelman, Fiji. Simo, who's from Instacart, Larry Summers, as well as our CEO, Sam Altman. Seems like a pretty good crew. Seems fun, stacked. OpenAI is governed by the board of the OpenAI nonprofit currently comprised. So it seems like there's some switch here. This is all very confusing. Anyway, let's get some, let's get some. Let's get some venture capitalists on that board.

Yeah, Jeff, void of that.

Jeff Lewis would be great. I want Thrive Capital on there. Masa, all of them good people. You know, SoftBank actually fantastic board governance. You know, people talk a lot of trash about SoftBank for a variety of reasons. When they come into a company, they run it like a public company. Yeah, it's like compensation committee, conflicts committee. Like, like the board meetings are full day events. Like very well organized because they take that stuff. Professionals, professionals.

High risk risk on professionals, extremely risky.

But professionals, which we respect here. At least one OpenAI investor seems satisfied with the outcome and told us that it's the best outcome given the parameters that exist right now. Especially because of the scrutiny and pressure that came with Elon Musk's lawsuit opposing the restructuring. The investor added that even as they awaited the potential IPO in OpenAI's future, large secondaries, which OpenAI has already done, could be another way for stakeholders to cash out. Keeping control in the hands of the nonprofit board is significant. The restructuring endeavor began because the nonprofits nonprofit board's goal to make AI for the benefit of humanity may or may not conflict one day with shareholders desire for hundreds of billions of dollars in profit. You might Remember that when OpenAI announced the restructuring four months ago, the board which oversees OpenAI's nonprofits said they couldn't adequately consider the interests of investors who were pumping tens of billions of dollars into the for profit subsidiary that runs ChatGPT.

Imagine you're an LP in one of the funds that has deployed billions and you send, you know, a little Saturday afternoon email being like, hey, by the way, what do we actually own of OpenAI? And then you as an investor have to try to, you know.

Yeah.

Untangle this. This web.

They had a name for it. It wasn't for a while it wasn't. It was like credits or something or like fun coupons. It wasn't shares, it wasn't stock, it was units. Units.

Units.

That's what they were calling it. Absolutely.

Well, that's an terminology standard with LLCs.

Absolute units.

Absolute units. Absolute dogs.

Yeah. No, no, no, it makes sense. Anyway, I'm sure they have to plan all these secondaries, plan all of these share sales. And what are they using to plan?

Well, I don't know if they're. I doubt they're using Linear for corporate governance, but I bet you they're using line purpose built planning for sure. And product building. LINEAR is a system for modern software development, streamline issues, projects and product roadmaps. You guys know already that we use Linear for running the show and we are very grateful to have them as a sponsor. We are also going to be live in person tomorrow in SF with Kari from Linear and a bunch of other people. So very excited for that.

Well, there is more news. We'll stay on OpenAI for a little bit. We have so many stories around this. So Microsoft, we'll go to the Microsoft angle. Sharon Gaffer Gaffare says Microsoft has not yet agreed to OpenAI's restructuring plans and is a key holdout. Negotiations needed to get the restructured deal done along with approval from state AG offices still being decided how much equity Microsoft gets.

So such a wild dynamic this is.

Microsoft has invested $13.75 billion in the startup and they remain the biggest holdout among investors as the ChatGPT maker tries to restructure, according to several people familiar with the matter. Now, that 13.75 billion, was that cash that ever transferred and was wired or were those credits or did they transfer and then go right back to Azure? Like I'm very interested in the anatomy of that deal. I hope that someone did into this more and actually asks for the structure because.

Yeah, I mean, I think it was. I know it was both. The question is, what is the actual ratio?

Yeah. And how does, I guess how does Microsoft feel about how much they invested? Do they feel like, hey, we're owed 13.75 billion at this valuation, at the last valuation, a new valuation of post dilution valuation? Or are they like, yeah, it was really only like 5 billion in total cost because we gave you credits at like a two to one ratio, right?

Yeah. I mean the thing here that's apparent is it's non Standard for a $3.2 trillion company to invest $13 billion in another company that then is restructuring. And it's unclear how much the $3.2 trillion company is even going to own of the underlying investment. Haven't seen this at least in my lifetime.

Yeah, for sure. There, there, there, there really has never been a tech company.

It's funny, they kind of have to just like get around the table and decide, okay, how much is Satya gonna get? You know, how much does Sam get? Right. Because a part of this, I think Sam needs to get some personal exposure because you know, if you check back to the Senate hearings, he at that point in time had nothing. Had nothing.

Then he got a bunch of yeah, it does seem like they're going to make a movie out of this, but it might be more complicated than the.

Big Short or might be like a four hour documentary.

It's just going to be Margot Robbie in the bathtub for six hours explaining how all these different entities interact because it's more complicated than mortgage backed securities and credit default swaps combined. It's going to make the 08 financial crisis look like two plus two. It's like by comparison this is so much more complex. Anyway, Microsoft is still actively negotiating details. Microsoft not commenting. They said In a statement OpenAI said, We continue to work closely with Microsoft and look forward to finalizing the details of this recapitalization in the near future. I mean the good news is like Microsoft and OpenAI are pretty well aligned on what they do best. OpenAI clearly is fantastic at consumer product development. Breaking through with the Ghibli moment with deep research like all this. And then in terms of the API, it's great to vend that through to Azure and that's the relationship and it seems like it's a really fruitful one and it's okay for them not to dominate on the Azure side or take some sort of like not really build that out. It doesn't seem like they're. No one would say, oh, they're just going to be a consumer tech company. They'll only be worth 3 trillion. It's fine if they just become just Google or just Facebook. Right? Like that. That's a pretty good outcome.

Just eat some percentage of search volume.

Yeah, seriously. Microsoft isn't the only party that OpenAI needs buy in from. They're fighting with the attorneys general of California and Delaware. OpenAI needs to do a fair market valuation on the nonprofit stake in the future for profit entity and is asking state AGs for input. Say how much are we worth? So this is kind of like a 409A situation or like a 83B. Let's figure out what the base valuation here is. But that can be so wild. I mean, imagine you're attorneys General and.

Trying to value OpenAI. It's like on one hand, you may have created a machine God, that's worth infinity. That's worth infinity. On the other hand, you're burning, yeah, five, $10 billion a year.

And the world's richest man is actively.

Trying to stop fundraising running to the ground.

So it's either infinity or 0 one of the two.

So funny, I'm using O3 and I asked, how much cash has Microsoft invested in OpenAI? And it's basically stuck in a doom loop of searching the web and then trying to understand it and then just looping back to searching the web. It's been going for about five minutes now.

Wow.

So we'll see.

Just 03. Not even deep research. Yeah, just 03 is just five minutes into this. Figuring this out.

Well, I just want a pretty specific answer.

This is humanity's last exam.

Yeah.

How much money makes sense of OpenAI. If you can do that, you've passed the real Turing test. Yeah. AGI is here. So Microsoft's approval will be key. Only open. Only OpenAI insiders, Microsoft and other early investors currently have a direct say in approving the restructure, according to two people. As a result, only that group gets to weigh in on the restructuring plan, the people said. But among investors, OpenAI is currently only negotiating with Microsoft. Microsoft has a unique relationship compared with other investors because of its licensing and revenue sharing. And they also had that AGI clause for a while, but I don't know where that went. Anyway, I'm sure Microsoft, OpenAI, they all got to manage risk, they got to automate compliance, they got to get on Vanta, do it, approve trust continuously. Vanta's trust management platform takes the manual work out of your security and compliance process and replaces it with continuous automation, whether you're pursuing your first framework or managing a complex program. I can't imagine anything more complex than doing compliance at OpenAI. And yeah, if you're working pretty much.

Any company, it's very painful. Start to finish support for 35 different compliance frameworks. SoC2, HIPAA, you name it. Head over to Vanta, head over there, tell them the technology brother sent you.

Fantastic. Anyway, founder of Windsurf posted big announcement.

Tomorrow as if the biggest announcement wasn't going. And doing a podcast with a double you know double collar on yeah.

And so the OpenAI agreement to buy startup Windserf for 3 billion is finalized. It's been reported in Bloomberg. We need a massive size gone. Katie Roof has the story over at Bloomberg. OpenAI has agreed to buy Windsurf, an artificial intelligence coding tool formerly known as codium, for about 3 billion, according to people familiar with the matter, marking the ChatGPT maker's largest acquisition to date. The deal has not yet closed, said the people, who spoke on condition of anonymity. Both companies declined to comment. Although the founders out there tweeting and wearing two polos the acquisition could draw could help OpenAI take on rising competition in the market for AI driven coding assistance systems capable of tasks like writing code based on natural language prompting. Bloomberg News previously reported that the two companies were in discussions. Windsurf, formerly called exafunction Inc. Has recently been in talks with investors including Kleiner Perkins, General Catalyst to raise a $3 billion valuation. Last year they were valued at 1.25 by General Catalyst and OpenAI came over and scooped them up. So I have a little bit of a history deep dive if you want to go through this for to break down Windsurf how they wound up there so 2021 ages ago in the AI race but you know a real over true overnight success here could have been starting a company in 2021. Yeah they could have gone into crypto, they decided to go into mi, they decided to go into AI. So their MIT friends Varun Mohan X Neuro Autonomy Infrast leading Douglas Chen, who had worked at Meta and Oculus as an engineer, incorporated exafunction Inc in California to abstract away the challenges of managing accelerated hardware like GPUs. Like a GPU, they're basically a GPU wrapper. It's funny like wrapper is in their DNA.

Yeah. So basically what they were doing actually have a post up from Bruno from X and at the time exafunction customers connect to the company's managed service or deploy exafunction software in a kubernetes cluster. The technology dynamically allocated resources moving computation onto cost effective hardware such as spot instances when available. And so anyways pretty wild to see.

Yeah so you can get cooked if you're on AWS and you're just using like the reserved instances and so sometimes if there's, if there's specifically, you know, cloud hardware that's underutilized, the price falls in the spot market and so you want to dynamically move over there. So they're using this like server side virtualization to do that. But it really is a GPU wrapper. Yeah, but it's funny because Kubernetes is also supposed to do that.

Don't stop wrapping.

Yeah, but they wrapped. They wrapped the wrapper. So they raised a $3 million seed from Green Oaks to build this.

I have another post up from Shai, who highlighted Shai Goldman. He highlighted that the round was 3 million on 23 million pre. Wow.

Pretty good on the dilution front.

Yeah, not. Not bad.

Watch out, Sotheby's. These founders are going to be bidding against a lot of you folks.

Yeah, VC is creating their own competition at Sotheby's.

Seriously, this is dangerous.

Of course, they went on to do many more rounds, including some that we'll get to.

Yeah. So A year later, April 28, 2022, Series A X of function emerged from Stealth with a 25 million dollar Series A. It says Green Oaks and Founders Fund. Is this really the FF company? I never knew that if that's the case, but I know Lee Marie was in and so maybe, maybe that got kind of confused. I don't know. Hallucination, who knows? It's a bold claim. And they had a bold claim to be the reference infrastructure for service serverless deep learning. And so they're going deeper into AI, specifically promising customers faster inference and lower bills by multiplexing workloads across thousands of GPUs. So there's a lot of idle GPUs out there there, and they're going to find and allocate your inference workloads to the lowest ones. This is a famous story from midjourney, which was like getting started around this time. Mid Journey, obviously, very inference. Very inference heavy in the sense that you generate an image and it generates four of these images and they diffuse very slowly and you're getting these little streaming updates while the images kind of depix elate. But famously mid Journey, because it doesn't matter that it's low latency because it takes a minute and so it doesn't matter if it takes a full second to stream you the final image. They would render those images on servers across the world because no one's watching Netflix in 3am in Taipei or whatever. So they would have a server farm.

Well, I'm sure a few people are.

Yeah, but. But you get the idea. Like the American servers were used for the international audience and vice versa. And so this idea of like load balancing is a tale as old as time, but clearly there was a market here and so they raised their $25 million series, early pilot customers in autonomous vehicle Envision startups reported 10.5-10x hardware utilization gains. But Mohan worried the business world would be the business would become a commodity once every model looks the same. Interesting. So they make a decision to pivot. In mid-2022, GitHub Copilot's debut, the founders concluded that their application layer, AI, not infrastructure, would capture real world value. They chose to pivot, betting that their low latency serving engine could run an AI code assistant cheaper than anyone else. Fascinating. So like in the at this time like the meme was don't build a wrapper, the value accrues to like the hardware layer, like go deeper. And instead they're saying hey, we're actually set up really well to serve a fantastic AI code assistant. And cost competition is going to be very important. Even though we've been in this era of like free VC money and, and people are paying a ton of money for everything. But having that, having that margin and that optimization really makes sense. And it seems like that's increasingly the way you break through is like the, you know, deep seq. It's like the optimization on the inference optimization on the hardware, like the economic gravity of these projects really does matter. So In December of 26, December 6 of 2022 Codeium launches on Hacker News. Neil On Hacker News, Mohan announced, we just released Codium to open up access of generative AI to all developers for free. Tried in the browser. The extension shipped for VS code, JetBrains Vim and more. Trained only on permissively licensed public code and promising to stay free forever for individuals. Very cool. So speed wins fans in early 2023, they're getting up and running. Kleiner Perkins leads a $65 million Series B at a $500 million valuation. And the, and the, and the response on Hacker News is very good. Folks are saying that it's noticeably faster than Copilot and Edge, credited to Exafunction's GPU scheduler. And so I mean this is the time when everything was scaling up and like, I mean even ChatGPT at the time would be like, oh, I'm cooked. GPUs are melting.

I'm cooked.

I mean I'm sure you saw this at the Studio Ghibli moment. There would be tons of times when it would just be like failed, broke.

Connection or just come up with an excuse not to do the work.

Lazy, lazy. And so so by mid year, Codeium had grown organically to tens of thousands of daily users. Without a dollar of marketing spend. Super viral, especially in tech, Twitter and X. Kleiner Perkins leads that series B, joined by Green Oaks and General Catalyst. Kleiner Perkins partner Lee Marie Braswell called Codeium a secure personalized productivity multiplier, deployable on PREM or in the cloud. We got to have LM on. She's great. I worked with her for a couple months, months at Founders Fund and then she got poached over to KP.

Very dramatic poach alert.

Yeah. So they hit 300,000 developers in just 15 months. VentureBeat reports that Codeum already wrote 44% of new commits for over 300,000 developers and dozens of Fortune 500 teams.

So the CEO at this time gave a quote to TechCrunch saying, Even though we've barely made a dent in the BE money, this lets us ramp, ramp R and D and make larger strategic bets. You love it just scaling.

So they add self hosted and air gap deployments, which is pretty cool. SoC2 type 2 compliance, post generation license filtering features that won cautious customers like Anduril, Zillow and Dell. And so yeah, if you need to be in a. In a. I think there we've been talking about this with like some of the llama instances going into the DoD like at a certain point, like, like you can't just be like calling out to some random server all the time. Like it has to run locally. And so having that optimization background makes a ton of sense. They become a unicorn in 2024. Just what, six months ago, August 29, 2024, they do their Series C. TechCrunch breaks the news. $150 million Series C led by General Catalyst valuing them at 1.25 billion. Total funding hit 243 million. GC's Quentin Clark praised the customer follow ethos and multi IDE coverage. They're at 700,000 users, then a thousand enterprises. Crazy. And they were the leading independent rival to GitHub copilot at this time. To signal a bigger vision, the company rebrands the product and domain as Windsurf. Which is interesting. Yeah, this happened in November of 2024 and that's where like. So it felt like Windsurf kind of came out of nowhere, but we're now like three rebrands into this. Like it goes from exafunction, which you might have heard of Encodium, which I think I'd heard of a little bit. And now Windsurf came out and then it was like the big launch and everyone's talking about Windsurf being So great. A new AI native IDE bundled inline completions, whole repo search and cascade. An autonomous agent that plans multi step fixes and starts working before you ask. Aimed at keeping devs tens.

Windsurf is beautiful name.

It is.

Windsurfing kind of fell off, used to be a thing.

It's moved on to what's the new one?

Kiteboarding.

Kiteboarding. Maybe next. Next AI company.

Yeah.

One editor unlimited Superpowers marketing touted Windsurf tab, a command palette that surfaces context aware snippets, test docs and deployment from a single keynote keystroke heralding the ide wars of 2025. We will never forget those. They hit 40 million.

I mean we're still in the midst. We're still in the midst of them. I mean obviously the other news out of the last 24 hours is that cursor's raising 900 on 9 billion.

9 billion.

That was initially rumored to be at closer to 10.

Yeah.

But has a bunch of big names in it too. So the IDE wars will continue, the show will go on and yeah, I don't think Windsurf is not going to back down in terms of their ambitions. I am interested to see if they just sort of continue the Windsurf brand or it's just rolled into ChatGPT completely.

But also interesting that they have this freemium Funnel converting to $12 to $60 per seat Enterprise contracts. Like even doing per seat here instead of consumption is very interesting because that doesn't necessarily like you have at least in the short term, like real variable costs like you could sell in a bunch of seats and if you have some power user who's just sitting there hitting tab every two seconds like they could burn through a lot of inference cost.

Yeah.

But I guess that they're just really comfortable with their infrastructure scaling.

Yeah. And to be clear, I don't think any of their investors at any point over the last two years was really saying, hey, you could focus, you should really focus on margins. Just because it was clear that inference costs were dropping off a cliff during that entire period.

Yeah, yeah. I mean we've certainly seen that with the new O3 models. And it seems like something that's very, very solvable in the midterm as you bake the models down, distill them, optimize them, all those things. So certainly something you don't need to worry about. It's just interesting that they're betting on this per seat model. I wonder if that will stay because you could imagine that the margins could be the same and you could still do it on a consumption basis. Like AWS is mostly consumption based, not seat based.

Yeah, it's one of those things. At some point if companies are hiring less engineers but doing significantly more engineering work, you would want to effectively be an index on engineering labor, AI led.

Number of lines of code changed is what you want to be comped on or just raw inference cost, right? Just pay me 10x what it costs to inference this thing. I don't know. So there's some funding rumors. We talked about this. Kleiner Perkins is thinking about doing another deal but they get this acquisition agreement for 3 billion observers called it a distribution grab. Owning windsurf hands OpenAI direct channel to millions of programmers and keeps a prime asset away from anthropic Microsoft's GitHub and cursor. Before the sale, Windsurf and Cursor were neck and neck in the agentic IDE space, one analyst quipped. Begun the IDE wars have as Cursor sought a $900 million round to stay independent. Mohan credits weekly five hour walk and talks with his co founder for intellectual honesty, the habit that enabled a radical pivot from infra to app.

While competitors will meant Manitis Method.

Oh yeah, totally. Probably putting up 30,000 steps a day at that pace. That's great. So they still own the proprietary LLMs on a GPU virtualization layer. It owns end to end allowing sub to 100 millisecond suggestions and cost structure. Cheap enough to keep the individual tier free company filters out GPL and other restrictive licenses, blah blah blah blah blah blah. Anyway, pretty fun. Pretty fun. Really good outcome for everyone involved. Just a couple years. Billions of dollars created in value.

It's interesting. Again, this comes back to this kind of question for all these platform investors around needing a lot of wind surfs in a single fund to return a fund even potentially 1x. Right?

So I do wonder if the if Windsurf will get rolled into OpenAI and ChatGPT in any meaningful way. Because just from a product perspective, OpenAI has been kind of bifurcating the apps a little bit. So the latest Studio Ghibli moment happened via images in ChatGPT. The previous images project they worked on Dall e and Dall E2, those were separate URLs. So separate separate apps. They eventually brought some of that into ChatGPT. But Sora is still its own domain, it's its own editor. Basically because you want to be able to adjust the timing on certain cuts and add music and kind of a non linear editor just makes more sense for video editing. Certainly you don't want to be writing code just in the ChatGPT app, but you know they have a bajillion users, right?

Yeah.

Probably makes sense to bring to leverage that distribution one way or another.

My bet as of right now is windsurf by OpenAI, not OpenAI code.

Yeah. Not like, oh, a new button where I could click Deep Research or I could click Windsurf and if I click Windsurf, it just builds me an app. I don't know. I think that might be cool. It's already writing code pretty frequently. When you ask it to crunch some numbers, it'll go and crunch the numbers down. And tool use seems to be really important. So I could, I could imagine some of the stuff getting ported back, but it does seem like Windsurf just by itself is its own great product. You know, it's popular and so they will continue to do that. Anyway, we were tracking this on Polymarket, sponsor of the show, and Polymarket had will OpenAI acquire Windsurf before August shot to 99%. It was at like 60% yesterday, right?

Yeah. I think the big question was just trying to understand how. Yeah. If this actually was a rumor, it was at 80% a week ago. And then somebody bet big. No. And cratered. And then it just steadily ticked back up until yesterday.

Yeah. So. So yesterday when we talked about it on the show, that was May 5th and it was like noon, it was at 67% and we said that the Double Polo told the story.

Yeah.

And the Double Polo was evidence that the deal was going through that you wouldn't throw on the Double Polo if the deal with Sam Altman was going south. And we were correct. And there were a couple posts about this later in the show. We'll maybe dig through them. But anyway, cool to see that Polymarket, you know, predicted it and was. Was. Was edge positive right up until the, the final confirmation. I still think it's interesting because like, like, yes, they have an agreement, but there's so many things that could fall through post. So 99% maybe feels even a little bit too high now. I don't know. But yeah, who knows? Go and express your own view on polymarket. Check it out.

Oh, it's so funny.

What?

Varun is hilarious. So 34 minutes ago, after we started the show last night, he posted big announcement tomorrow. And everybody's just saying like, congrats, this is amazing. And the big announcement is WindsERT wave eight. It's so big. We're doing it over three days. Today we have a bunch of updates that make Windsurf the best product for teams. Windsurf can now automatically review your code, use internal knowledge sources, share a conversation, and deploy apps internally and more. So anyways, I guess I think this is great.

I think if you are getting acquired, it is weighty and it is important. But no, no, no. Focus on the. Focus on the company, focus on the product, focus on the customers. Like, job's not finished. Like, yes, you have new shareholders and you're. And you're aligned with the new team. You have more resources but continue to grow that.

Yeah, it's so interesting because now Bloomberg and a bunch of other media outlets have said OpenAI reaches agreement to buy startup Windsurf. But we haven't seen anything from the Windsurf team.

You would think that they would put.

Out an official statement or anything. Yeah.

So Dagesh Gupta said so the Windsurf guy wore a double polo on the YC podcast, which was last worn by the OpenAI guy when he was working on his first startup in the early 2000s, signaling that the rumors of acquisition are in fact true. And it's like the guy yelling in the girl's ear and Spore says, so this was correct.

Lol.

It actually did happen. Anyway, should we go through Apple's iPhone woes? The prices are going up. They're getting hit by tariffs. Apple won't be able to avoid price hikes for long. The iPhone maker is absorbing $900 million in tariff costs this quarter. But the hit to profit margins could still get worse. So Apple's latest earning report shows the company is choosing to eat the additional costs and not raise the prices. But Raymond Jane's analyst says looking ahead, our base case for Apple is to raise prices, which should help offset the tariff impact to an extent. Now, Apple hasn't raised prices in a while, even in the head of inflation, even in the face of COVID stuff. So I think they have some room to raise the prices. But they are getting expensive. Once you break the thousand dollar point price point, people, psychologically it's different. It feels more like a laptop at the same time. Spend more time on your phone than your laptop. Right. So yeah, I mean I'd rather, I'd rather have a $4,000 phone and a $900 laptop, honestly. Yeah, laptop can just kind of do whatever. I went to the Apple store to buy this laptop.

Yeah, I've always thought it is amazing that you know, a device that you can be so dependent on still costs less than $1,000.

Use it more than your car. Like, if it was $30,000, I think people would still buy it. Yeah. You know, of course there's like pricing dynamics of like, you know, will the iPhone stay competitive? Because it can't be 10x an Android. People will switch. Blue bubble is probably not worth that much. But, but, but in terms of like the value that's created in your life by your phone, like, it's, it's up there. Yeah, it's up there with the best of them. Yeah. The tariff impact so far is actually pretty slight. In Apple's earnings call, Tim Cook said The company expects 900 million in additional costs. That would add less than 2% to what Wall street was projecting for Apple's cost of sales. And so, so, yeah, the cost of sales is in the, what, 50 billion range to make all those iPhones. If it goes up by 0.9 billion, not that big of a deal. But few believe Apple's like, we're going.

To eat it for now and then next quarter we're going to raise prices and hopefully people forget.

Yeah. Given the prospect of tariffs, of Trump's sectoral tariffs, we think it is prudent to at least double the 900 million hit to COGS for a few quarters beyond June, says one of the researchers covering Apple. Concern about Apple's margins come in coming periods Helped drive the stock down nearly 4% following its earnings report. The stock lost a further 3% on Monday after the information reported that the new ultra thin Apple iPhone, iPhone Apple is planning to launch later this year will include compromises such as shorter battery life. Interesting. I didn't realize they were going thinner. That's kind of a cool story. I might do that. I don't know. I never run out of battery.

I would, I would. I'm not going to say. I'm not. I'm not gonna go as far as to say I would stand in line for an ultra thin iPhone, but I would hurry to get one.

So you. So you would be okay with shorter battery life? Yes, I think so too.

Well, because the battery's always good in the first six months.

Yeah, but can you imagine? So it's gonna be a thinner phone, but the camera bump will probably be like three times longer.

It's a telescope.

It's a telescope.

Imagine. Imagine a lens, little telescope.

People are like, yeah, the camera lenses keep breaking off because they're so. They protrude so much.

Ridiculous something.

Anyway, there's a chart here about the iPhone prices. They went up almost to a thousand. And they've been down for the last couple of years, hovering around 900. Can't really go too far anyway. Apple shed more than 350 billion in market cap since President Trump's announcement of new tariffs on April 2. Most other mega cap stocks have since recovered from their post tariff losses. Company's been making moves to offset the hit from tariffs. Cook said Thursday that most of the devices shipping to the US Will come from India and Vietnam. They're getting out, especially as it looks to enhance the iPhone with new cutting edge designs such as a thinner body or bendable screen that pose more complicated manufacturing challenges. Yeah, they got to spice it up. They got to do something different. I feel like there's, you know, people have the Mac, the iPad, the phone, throw something else in the watch, you know, just give me something in between. I was thinking about Polymarket has.

Will Apple release a new product line in 2025 at 24%? And will Apple release a foldable iPhone in 2025 at 7%? But bendable iPhone is not on the list yet.

Even if it's worse. I just want something fresh. I'm bored. I don't care.

Give me something.

Just give me something. Yeah, exactly.

Anyway, bendable iPhone is really funny because you can imagine people just putting it in their back pocket and then it's just sort of folding.

Yeah, I mean, I think bendable means foldable in this context, but.

Okay. I don't know, I was thinking more like bendable, like, you know, stick. If you try to actually fold a stick, you're gonna break it forward.

I want something new. I was hoping they do a tv. I still like the Apple Vision Pro. I'm excited for the next one.

How about a backpack? Yeah, that. That, you know, operates as a, you know, desktop, you know, computer and a very, you know, an extra large mobile phone.

You know what, you know what I was talking about?

If you could get, like, make just the largest mobile phone on earth.

There's all this.

There's all this.

There's all this boom about humanoid robots. I was posting about this on May 4, my birthday, Star Wars Day. May the 4th be with you. May the 4th. I was saying we should make R2D2.

Birthday on the show. Everybody's gonna be like, what the heck? You didn't. Yeah, you didn't for your birthday.

We did, we did. We talked about it on Friday and Thursday.

Right, right, right. We talked about the exact date. People were kind of left.

But. But making a. Making it R2D2 seems trivial at this point. Right. It's like a bunch of car batteries. The battery technology is definitely there. The LLMs are there. You throw a starlink on that thing or a bunch of cell signals, like a bunch of cell phones in there, you're going to have great connectivity. It's going to be able to follow you around. The wheels exist. Like all the tech is there to build R2D2. That would be a great product for Apple. Just $10,000 and it just follows you around and tells you the weather. And it's like, oh, you wanted me to play intelligence. You wanted me to play you too. I still don't have a good AI, but I don't know, something like that would be fun and different. Different. I'd be into it.

R2D2.

R2D2.

Put it in the roadmap.

Anyway, I hope that next time we.

Can get, you know, press passes to an. To an Apple event. We'll. We'll stay to wander Tim aside and we'll. And we'll say, hey, look, R2D2. Make it happen.

Yeah, for sure.

Got the fortress balance.

If we go to wdc, we'll be staying in Cupertino and we'll be in a wander.

That's right.

We'll be finding our happy place.

Find your happy place. Find your happy place. Wander just on a tear. It's actually insane.

It is.

Every single day.

They're in the conversation.

They're greatest of all time. You mean.

Yes.

That conversation. Right.

Book of wander with inspiring views, hotel grade amenities, dreamy beds, top tier cleaning and 247 concierge service. It's a vacation home, but better. Anyway, you were telling me about this, the bringing Hollywood back to America, this whole thing. Jon Voight has a plan. Which is. Isn't John Voight Angelina Jolie's dad? That's right.

Really?

Yep.

I have no idea because I've never seen any movies.

You haven't. So actor Jon Voight presented his plan to boost US Jobs in the entertainment industry to President Donald Trump over the weekend, laying out a proposal that included federal incentives for US theater owners to upgrade their facilities. Let's go Universal Basic.

Is that gonna fix issues? I mean, judging by the commentary that I hear from people in Hollywood.

Yeah.

The issues are around how much regulation there is.

Yeah.

And the costs associated with said regulation in making movies in Southern California and revitalizing theaters and making.

And the lack of technology podcasts in Hollywood.

Okay.

It's no way.

That's another. That's another short.

That's why it's falling apart.

The idea that all give him a chance to kind of get into this. But the idea that just fixing the bathrooms and old theaters is not necessarily going to revitalize, you know, the domestic entertainment industry.

Yeah, it's a tough model when you.

Have but nice bath. I like nice bathrooms.

I mean, you think about that. You think about when Hollywood was in its heyday and like, you didn't have all. Like, you couldn't watch anything in HD. Like, you had a TV with three channels on it and. And it was 12 inches and it was crackly. If you wanted to see something actually big and beautiful with good sound, like, you had to go to the theater. Now you can just buy a home theater with a projector for two grand and you can also just scroll YouTube and TikTok and streaming and your phone all day and there's unlimited content. It's impossible to keep up with it. And so the excuses to go to the movie theater are getting less and less. You really got to have a federal incentive for taking your boys out to the movies. Guys, guys. Movie night really will be the thing that will be the thing that brings Hollywood back for sure.

Yeah. You've been leading the charge here.

I have. I have highly recommend it. Just text everyone who's at who's in your town. Get. Get 20 tickets on Fandango that you can always refund. Refund the ones if people don't show up. Send out the text message to everyone, hey, we're going to this Jason Statham movie next Friday. Are you in? Are you out?

Be there, be there.

And then just, you know, Venmo everyone at the end and you're all good. Have a fun time. But other than that, if it's not like an event, it's really hard to justify going totally. Anyway, this. John Voight was was appointed a special ambassador to Hollywood by Trump in January.

Showbiz.

I feel like these special ambassadorships are just, like, up for grabs. Yeah. Could we become special ambassadors to technology podcasts? We really should.

Like, yeah, we can make it happen. So. Voight was appointed in January, and he's detailed the plan along with his manager, Stephen Paul and Scott Carroll, the president of the President Paul's company, to Trump at Mar A Lago. The proposals include changes to the tax code to encourage investment in US Films and job training initiatives. According to Carroll, the incentives for cinemas, if approved, could help buoy theater chains such as AMC Entertainment Holdings, Cinemark holdings and Marcus Corp. That have struggled since the pandemic with films being released online sooner than in the past. They're basically saying we're gonna bail out the theaters which have been struggling because of something that we're not addressing like it.

Oh, which is competition.

Films are being released online sooner. People wanted are happy to just watch them at home. Everybody has a massive hdtv.

Yep.

And there's just. And I would argue that films are just not, you know, often. Often, you know, for the last few years, films have been good enough to warrant going to a theater when you could just.

You want to bring back film to America. Here's an idea. Give Jason Carmen $300 million to remake the Terminator starring Sam Sulek.

Today. Today.

If you, if you do that, it will generate $5 billion in ticket sales.

I think that's very possible.

It's entirely possible.

So anyways, filmmakers who co produce pictures with foreign companies would be allowed to obtain credits for their U.S. spending for quote, unquote bad actors who take all of their production overseas. There would be a tariff that equated with the incentives they were getting from the foreign countries. Carol said in an interview, here's another idea. Trade war in Hollywood. The president loves the entertainment business and this country and he will make help us make Hollywood great again.

Yes. I recently watched Almost Famous. I'm trying to learn about journalism. Almost Famous is a story about a young boy who goes on the road to write a profile about an up and coming rock band for Rolling Stone. And it's a Bill Dungsberman. It's a coming of age tale. So he learns and he grows and he experiences all the vicissitudes of life on the road with this rock band. Learns the good and the bad and writes this fantastic profile. Eventually gets them on the COVID Spoiler alert. Alert. I think we need a Almost Famous style movie about Ashley Vance going on the road to write the first big profile of Elon Musk becoming a man, becoming, you know, a real, real tale of his journey on the road.

A true journo.

Yeah. A Bill Dung's room in.

Yeah.

Coming of age tale about Ashley Vance doing technology journalism. This is the next thing after we've made some tech. We haven't made enough movies about technology journalism.

Yep.

This could be big. This could be very, very big.

A movie directed by Julia Black.

Yes. Get her in and. But mainly it needs to be the story of Ashley Vance.

Yeah.

Telling the story of Elon Musk.

Elon Musk is Ashley Vance.

Exactly. Yeah, exactly.

Trump posted on social media Sunday evening that he would impose 100% tariffs on all films produced over overseas. The comments sent shares of U.S. entertainment companies tumbling on Monday as investor tried to discern what types of films the issue is it's the US Entertainment companies that have been paying to produce films overseas as investors tried to discern what types of films would be subject to terrorists and how the levies would affect the profitability of producers. In a press briefing on Monday, Trump said he planned to meet with entertainment industry executives to discuss his his plan. I want to make sure they're happy with it because we're all about jobs, the president said. California Senator Ben Allen said he would he was given a heads up about Voight's plan last week. He said he would support a tariff, but that should be crafted to encourage US Productions, not harm foreign films. I understand why the president was moved to want to take some drastic action, allen said. We have been facing so much damage as a result of other jurisdictions throwing all sorts of incentives at productions to pull movie production. That really ought to be happening here in the United States and that ought to be happening in la. And again, you know what I hear is just the dynamic, the logistical complexity of producing a film in la. Why are you laughing?

Because I have another idea to save Hollywood. Michael Bay directing the Ramp Movie the Ramp Movie the Ramp Movie business@ramp.com yes, the evil receipt personified like Transformers.

Yeah.

The corporate receipt malaise is has fallen upon the American economy and only Eric Gliman, played by Vin Diesel can revive the American economy by bringing by streamlining.

Your corporate cards Godzilla sized receipt.

Yes.

Stamp.

Maybe. Maybe we go Chris Hemsworth for Eric and Vin Diesel can play. Correct. Yeah, I think that's a good pairing.

That's right.

We need some AI generated movies to like kind of really sell this to Hollywood. But we'll get Michael Bay in there. It'll be high stakes. It'll make the it's time to tap in bae. Yeah, it'll. Yeah, it'll make. It'll make the Social Network look like a snooze fest. Yeah, it's gonna be great. Okay.

You're onto something anyways. We should have somebody on that actively has tried to make movies in la.

Yeah. Over the last Michael Bay on I heard he's making a really cool movie. I don't know if it's public yet, but it seems Sounds awesome.

Well, he's been making movies for a while so I'm sure you're not leaking anything too significant. But yeah, we should have him on. I'm very interested to hear the actual updated details on why if you're an American film producer, you decide, we're gonna go to Portugal or we're gonna go to London, wherever to actually make this, just because the cost to do it locally are so high. So. So I think that the local politicians, you know, state, county level, need to take some responsibility as well. Because I don't think you can just say, oh, federal government, you know, time to step in and make the movies. A part of the trade war.

Yeah, it was almost more important. We saw that with like LA versus Atlanta.

Yeah.

It's like kind of a domestic issue almost. I mean, Toronto is in there, but Toronto is not. This is not like some like massive global.

A lot of the people I know in entertainment have been spending a ton of time in Atlanta, like for years now.

Really?

Yeah.

Just generally. Right.

Yeah. I'm not, I'm not joking. I'm like, if you're actually. It's a big thing working to.

Yeah, you see it at the end of movies. It's like the peach and it's like made in Atlanta because they, like the folks in Atlanta were just like, let's build that business here.

Yeah.

And so they gave a bunch of incentives. Baby Driver was filmed there. There's a bunch of movies. Yeah. I mean, it's economic war on like the state level, which is interesting. Which is like totally fine. That's the way the states are supposed to work. It's just like Hollywood was mismanaged and just like played the game wrong, basically, and was like, oh, we're so cool and powerful and Lindy, like, we'll never lose this. So we don't need to play the game of tax incentives when in fact, like. No, no, they did. People would just leave. Like they didn't really have the loyalty that clearly the Hollywood policymakers and lawmakers like, thought that they would. Yeah, very rough and. Yeah, yeah, rough for. You know, I think it was Rob Lowe who was talking about it how he's always. He always has to go to like Toronto and he doesn't like it because he's away from his sleep. Yeah.

Toronto is the other one.

Probably away from his bed. Probably away from his Eight Sleep. Go to Eight Sleep dot com. What's new in Pod 4 Ultra? Well, Pod 4 Ultra has all the signature features you love about the Pod, plus new groundbreaking updates. They got a five year warranty, 30 night risk free trial, free returns and free shipping.

We are incredibly back.

Oh, yeah. I wonder how I did. I feel like I slept a ton last night. I'm feeling good. Finally back in it from the previous weekend and travel and whatnot. Let's see. Tuesday, I got a 93. How'd you do?

87.

87. Smoked you. Wait, is that two nights in a row? You said it was impossible. You said it couldn't be done.

Okay, to be clear, a little asterisk for the.

The.

For the sleep record.

Sick.

I did have food poisoning on Sunday.

Oh, skill issue.

And what happened?

Yeah, I'm good. I'm, I'm, I'm moving through this. I'm going to be, I'm going to be rock solid this week except going to SF tomorrow, but I think I'll be, I'll be on my eight sleep every night, which is good. Cool.

Awesome. I don't, I don't know if we have time for this next story.

We do have time for an Ad Quick ad, of course. Out of Home advertising made easy and measurable.

Say goodbye to the headaches of out of Home advertising Only Ad Quick combines.

Technology, technology out of Home expertise and.

Data to enable efficiency and ad buying across the globe.

Out of adquick.com size gong.

One of our best yet.

That is the best we definitely got to talk about.

Ad Quick is coming out with an AI product and we were in a cameo in the video.

Can't wait.

They're launching Ad Quick Copilot.

Yeah, Very cool.

To help you more efficiently by ads out of home.

Yeah, I mean there was a lot of stuff with Adquick. Like if you actually go through the process like they will, they will pull up the map, they'll pull up all their tools, they'll provide like kind of an idea and report and there's no reason why that shouldn't be a little bit more self serve. Even if you do wind up, you know, working with an Ad Quick representative directly, there's no reason why you shouldn't be able to just like investigate their whole database via an LLM. Like it's a very obvious. Anyway, anyway, let's do one or two posts and then we'll bring in our first guest. We have a great one. Evan McCann says with so many people going no alcohol, there is a lot of alpha in drinking. Again, you love it.

I thought this was hilarious. I saw it earlier and I laughed. Contrarian. Yeah, it seems like alcohol fell off but at the same time it's just so lindy. There's an opportunity to bond with other people that, that love alcohol.

Yeah.

So it's just, it used to be so common to drink alcohol that you wouldn't have a meeting and somebody Else says you want to have a drink and you say yes. And it's like, oh, we now have some common ground.

We gotta send this.

Can find some common ground.

Furious.

No, it's funny. I. I've just gotten to a point where I. I stopped enjoying alcohol. You know, I went from drinking maybe, you know, having a drink one day a week on average to realize to going down to probably one bottle of.

Dom every time you double.

That's right. That, that most of the drinks I've had were on the show.

Yeah. It needs to be tied to celebrations because then you just start grinding harder and harder and harder for everyone.

Yeah. But I think the alcohol decision, like you shouldn't let the timeline decide whether or not you like drinking alcohol because some people can have a few drinks and, and their body can process it and handle it. Well, for me, I. It's not for me perform better than they're drinking. Yeah. The Balmer peak.

I've been. Yeah. I've been very pro. Alcohol isn't a performance enhancing drug for sales reps. Yeah.

The power lunch methodology.

Power lunch. You gotta.

John Coogan power.

Loosen up.

Yeah.

Have a couple martinis with a client.

Yep.

Inc. A deal.

Inc. A deal.

This is how business used to be done. It's coming back.

Yep.

Coming back.

It's coming back.

Anyway, Corey Levy was over at Y Combinator. Game respects game. Very, very funny picture. He of course, runs Z Fellows, direct competitor to Y Combinator in many ways. Corey's running around the valley signing deals before companies get to YC sometimes.

Yeah. But I'm sure Corey's generally excited when companies he backs totally end up joining yc.

Yeah. This is one of those things where, you know, they are like, they are competitors, but the nature of competition is that sometimes, like when you see two competitors take a picture together, do some sort of collab, it's like they either hate each other or they love each other, you know?

Yeah.

I feel like, I feel like a lot of people experience this where like, yeah, you could be like competitors. But if you guys, if you understand your business at a deeper level than everyone else on the outside, you can wind up having like a beautiful partnership and friendship with someone who is ostensibly a rival. Because you actually see the world in a more nuanced way than someone just on the outside being like, oh, you guys are just capital and you guys just invest. Like. They actually probably view their businesses very, very differently and as very complimentary.

Totally. Yeah. I think Z Fellows had an interesting wedge and it's just consistently found Great founders. Oftentimes, I think even before they have an idea which is. Is not the same model as yc. Even though YC will in small instances.

Yeah. Even though they are somewhat in the same way.

I would say it is the Z Fellows of Z Fellows is Z Fellows and the YC of YC is yc.

Yes, yes.

And those things can, you know, overlap.

Totally with an X. Yeah. Xyz.

Xyz. Well, we have our guests ready in the waiting room. Let's bring them in. They're both in the waiting room. So do you think it's.

I wanted to bring in Kelvin first. I will text Santi and tell him to give us 15 minutes and we'll go back to back. I think that's probably what's best. Let me text them and let's bring Kelvin in. And oh, FYI, Kelvin is with fai. We are going to do Kelvin first. Then. Then Santi. Santi 1245. Cool. We are good for that. Let's bring in Kelvin whenever he gets a chance and let's update his lower third. The Chiron is ever evolving. All part of our mission to bring media to Hollywood. I like that stinger. That's very Michael Bay. It's not. There's something about the Ashton hall one that just, just. It's hit so much harder. Yeah. It's just so much better.

So much better.

So much better. It's night and day. It's night and day.

Anyway, if you have any sound effects you want us to add to the soundboard, reach out to Ben. Yes, he's a soundboard maestro.

Yeah, we really got to get on, you know, TikTok, Instagram, whatever. Wherever these sound effects are coming from, we need to bring them into the show because the soundboard has been a delightful upgrade. And we have our first guest of the show. Welcome to the stream. Kelvin. How you doing?

Boom.

Gentlemen.

Good to see you guys.

What's going on? Great to have you.

Yeah. Oh, yeah. We're big fans of the soundboard. Expect the unexpected during this interview. Probably be the loudest, most interruption riddled podcast you ever do, but it's fun. Can you. Anyway, can you give us a little bit of a backstory on you and specifically the announcement today some of the work you've published and kind of just give us the high level overview.

100% some Kelvin, kind of prototypical Silicon Valley tech bro. Dropped out of Princeton, moved out to the Bay Area, was a founding engineer at two companies and then a few years ago just started reading about industrial policy and the history of technological progress in the States and recognized that we were in the moment of intense techno industrial decline and that this was in large part facilitated by bad policy choices. And so in a moving to D.C. got to meet people like Santi and we released this playbook after six months of long hard work. And it's a recognition of the fact that tech industrial decline is a policy choice and that we are in a moment where with the rise of China, with the stagnating industrial base, there are intense changes that need to be made to the existing system. And these this playbook has 27 very concrete ideas for how to change things across the industrial base, across national security apparatus and in the ways the government funds frontier innovation.

Cool.

I want to go into the playbook. But first give us a lay of the land. Fai, ifp, what are these organizations? What are their goals? How are they funded? They're non profits. Can you break all that down? And then we'll go into the actual playbook.

100. So there's four institutions involved in this project. FAI, IFP, American Compass and NIA. A lot of acronyms here. So FAI is the foundation for American Innovation, the Center Right Tech Policy think tank focused on tech policy, emerging technologies, industrial policy. IFP is a nonpartisan think tank focused on a lot of similar topics also around, particularly around metascience and energy as well. And then American Compass is more of a new right economic think tank. They do a lot other stuff beyond just industrial policies and family policy, workforce policy, et cetera. And then NIA is the new American Industrial alliance focused on a new trade group started by a bunch of really sharp builders focused on bringing the best of Silicon Valley industrialists to Washington dc. So this project is kind of a merger of all these different circles of the new right conservative folks all the way to your Silicon Valley builders.

Got it. So we, we talked to Kathryn Boyle yesterday about American dynamism broadly. We've talked to Chris Power about Hadrian and building manufacturing in the United States and oral and some of these themes. But what are some of the more concrete steps that you're advocating for on the road to re industrialization, Techno industrialization? What does that actually mean?

Yeah, so this was born out of a kind of growing bipartisan consensus over the past two administrations that something like some things deep need to change. And the Biden administration really took a pillar approach where they looked at things like chips, energy manufacturing and passed really large scale subsidy programs that have been unseen in generations. And the Trump administration has recognized that there's all these systemic Barriers to industrialization beyond just subsidies, things like trade policy, non trade barriers. And so where we wanted to fit into the conversation is there is recognizing that there is in all these different critical industries, whether it's critical minerals or hypersonic missiles, all the way to biotech manufacturing, there are unique industry nuances that prevent more things being built faster and cheaper. And so the goal with this playbook was to really channel this newfound energy both sides of the aisle in advancing these issues and elevate folks who could write up very specific policy proposals along these lines to advance specific areas. So the way we categorize the playbook is we have three sections. Industrial Power, National Security and Frontier Innovation. And we have senior domain experts right on each individual piece of this pie with proposals spanning from developing hypersonics testing infrastructure to demand side support for critical minerals production to how does the Navy improve its ship streamline its design of ships. There's all these really wonky things that this playbook gets into in each of these specific categories.

Yeah. Do you have an overarching theory around tariff policy versus subsidization of American industrialization versus kind of the doge stuff which is like do more with less efficiency driven. These three pillars seem very different. Very. They're typically advocated by very different political groups. We're kind of seeing all three happen right now. Do you think that all three of those are equal tools in the tool chest or are you advocating for leaning one way or the other? Yeah.

This is what we're most proudest of for this playbook is that it is a synthesis of all these ideas and it's a recognition that for all these deeply, deeply important questions, you just need all of the above approach. There is no one silver bullet. The things that are required to restore American shipbuilding are going to be slightly different than things required to do a more broad based re industrialization movement. But it's a recognition that for all these different topics, there's all these things from regulatory fixes to state capacity fixes to subsidies that are all critically important on shipbuilding.

Can you give us a concrete example of what the playbook looks like? Because I feel like shipbuilding is pretty. You can kind of wrap your hands around it like it's a pretty, pretty simple thing to understand. We make a certain amount of ships, we want to make more. But what are all the steps that lead to actually increasing shipbuilding capacity?

Yeah, so shipbuilding in America is in a dire need of a fix. We produce China produces 230 times more ship tonnage than the United States. This is due to a Lot of different factors, everything from, from constantly changing requirements from the Navy to a lack of workforce to a lack of automation in our shipyards compared to other countries. And so the particular proposal that Brian Potter and Austin Bernard vote for our playbook was around streamlining the way the Navy actually designs its requirements for ships and the way the Navy actually goes about planning the design of certain naval vessels. And it gets pretty into the weeds on those topics. And really this is, you know, this is not a comprehensive plan to restore American shipbuilding, but more of a. Of a specific deep dive into one particular underrated aspect of the problem that policymakers may not realize. And so there's a great bill out there called the Ships act, which gets into a bit of.

More of a.

Of a broad based plan. And so the goal with this particular proposal was how do we complement the existing discourse with things that are more. A bit more of the weeds that people aren't paying attention to.

To.

Can you talk about some historical precedent where America was maybe behind in certain areas and managed to turn it around? Like basically like. Obviously you lay out a bunch of very kind of proactive solutions for a variety of these issues, but I think it'd be helpful to understand at what points throughout history maybe we recognize that. That big changes were needed from an industrial policy standpoint and we actually effectively made those changes.

Yeah, the Cold War is the best example of this. Right. You had space race where the US Public and our intelligence agencies were shocked when Russia launched Sputnik. It was really a. It's really hard to just fully articulate how deep this shocked the American psyche and how fast we got our shit together. In a matter of less than a decade, we directed 400,000 people, 20,000 companies and 4.4% of the overall federal budget towards the Apollo missions. And we went from being laggards in engine technology, missile technology, space technology, to land getting to the moon first.

Is that part of the challenge though now is that we don't have sort of a stunning visual of industrial decline. It's sort of this sort of, of slower, insidious sort of process that's been happening that we feel in different ways, but it doesn't strike fear into the hearts of Americans in the way that Sputnik did.

That's 100% a challenge. There's some people who don't believe America really gets serious until there's a Taiwan scenario. But there have been instances, there's been flashpoints over the past few years. Ukraine Covid October 7 where people have realized actually like the Lack of a robust defense supply chain, the lack of the ability to make basic things like medical masks are problems that affect us today. And so that's where the work you guys are doing is so important. That's where the work of these think tanks and other folks like Kathryn Boyle, the whole American dynamism movement. The more we talk about these things, the more I think people will recognize the seriousness by which we need to take these issues.

Did you ever explore the idea of, I don't know exactly how this would like get implemented, but some sort of incentive for American investors to focus on reindustrialization? We were kicking around this idea of like the, the carried interest tax was kind of up in the air. Maybe that would hurt private equity firms and venture capital firms. But what if carried interest, that that deduction didn't exist for investing outside of America. But if you invest in America, all of a sudden you get a much more favorable tax treatment. Something along those lines or even just anything like. We've seen the American dynamism movement take hold in the venture community, but we haven't heard nearly as much on the private equity side. That seems like the private equity firms are still just kind of dollars and cents maxing. But what do you. Do you think there's any moves that could kind of lead to more just decentralized American capital driving re industrialization?

John, that's such a fantastic question. There are so many levers, it'll take a long time to go through all of them. But Julius Klein of American affairs wrote a great piece for us on that very question of how do you get large scale capital institutions to invest in industrial finance? There's also Sam Hammond wrote a piece on utilizing the SBA's authorities to incentivize investing in small medium sized manufacturers. One of the greatest strengths of America is that we have the world's strongest entrepreneurial base. We have folks like Jeff Bezos, Elon Musk, and we also have a state that is capable of doing a lot of things right. Tesla being a great Tesla and SpaceX being the prime examples of what can happen when you combine effective industrial policy, effective tools of statecraft with, with amazing entrepreneurs who are building the future.

Yeah, interesting. Some sort of. Do you have any follow ups?

Yeah, I'm curious as you guys were putting this together, what kind of announcements or policies maybe sort of shifted your priorities? Because over the last call it six weeks there's been a lot of announcements. You know, on Sunday we had 100%.

Tariffs on the 3D chess board has.

Been flipped yeah, we're just flipping the board constantly. But no, I'm curious, was there anything that you were, were wanting to write about or include in here but you felt like already had enough momentum or kind of coverage from existing policies?

Yeah.

To give the Trump administration credit, we had a few things written up around defense procurement reform, energy reform, where they put out some EOs and we were like, oh man, we'd add some more things. It's been a very fast moving administration. But you brought up Liberation Day the way that we viewed tariffs. And each of these four think tank partners have their own opinions, but we're all in agreement that tariffs alone are not enough. Right. It's a stick. But if you actually want to be serious about doing the very specific things like critical mineral production or shipbuilding or whatever, there's a whole host of unique tools you need to apply to that particular problem. So that's really what this playbook is trying to provide.

Are there any lessons that you think America can learn from how China does industrial policy? We've seen, you know what, 50 years of investing in semiconductors over in China, in the midst of never really being on the leading edge there, but now even around planning.

Right. So they plan in these sort of five year cycles, right. You know, let's make a five year plan and another five year plan and they just sort of compound.

I know, I know we're adversaries, but like what can America learn from China on the industrial policy front?

Yeah, it's a fantastic question. And I mean, to your point, there's been a lot of waste over there, right? You have all these state backed private equity firms that kind of are just black holes for, for capital. You've got, I mean, they use a bunch of slave labor which helps drive down industrial costs. Yeah, it's, yeah. Yet they do a lot of things that are really effective. Like, you know, I think a lot of narratives around oh, China is just a copycat or China is like using like, like using distilled models to produce beef seek. It's a lot of cope. I mean, you know, they're leading electric vehicles. BYD is an absolute beast. Drones, critical minerals, like these are, these are real advantages. And the number one thing I think is just we should be taking things as seriously as they are. Like Chinese leaders routinely come out, not just Xi Jinping and the top level CCP brass, but even like the, the CIA equivalent in China. And they're like internal think tanks. They're constantly producing essays and giving speeches on how the foundation for world power is Industrial power and technological power. They view this as. They view the lack of technology as the reason they suffered the century of humiliation. And so they are deeply serious and throwing their entire state behind developing the future of all these different emerging technologies. So, you know, the thing that we can learn is just the degree of seriousness by which there's a political class is taking.

That's a really good answer. Yeah, thank you. Yeah, well, we'll move on and we'll let you get back to your day, but thank you so much for coming on. This is really interesting.

Yeah, I'm excited to fully read the report myself and yeah, looking forward to. Let's have you back on in the near future.

Yeah, yeah, we'll talk to you soon, sir. Thanks so much. And next up, we're continuing our deep dive into the techno industrialist play policy playbook. If you haven't had a chance to go see it, it's all over X today.

Rebuilding tech.

Rebuilding tech. What a good domain.

WWW Rebuilding tech.

Yeah. So Santi Ruiz is joining and he can share the techno industrial industrial policy playbook. Is the United States still the world's leading techno industrial power? The answer is no longer obvious and that should worry us. Let's bring him in to have him break it down for us. How you doing?

Welcome to the dramatic entrance.

Fantastic.

What's going on?

How you doing? Good.

I'm good, guys, thanks for having me. Long time listener, first time caller.

Yeah, yeah. Great to have you on here. Yeah, we got a little bit of an overview of how the. The little what, think tank cinematic universe came together. The Infinity Stone came together. Yeah, the Infinity Stone of thinking. But, but. So let's just jump into it. What? What? What? What? What? I know that there's a number of policies, but which one do you think is the most relevant today? There's a lot of things in the news around tariffs. TikTok is still in the news, but has kind of faded. If there's just one takeaway that you want to dive into first, where should we go?

Totally. Let me get to that in one second. I just want to say Kelvin did not flog the hard copy of this thing.

You got to get that. But it's really.

But it's really nice. Like, really happy. I got a hold of it yesterday and I'm like really feeling my oats on this. So, yeah, we'll put up the drop shipping link soon.

Yeah. We love the printed word. We're big fans of the Wall Street Journal. We love. We love books. We love everything printed.

No. And the Playbook should be in print. I mean, it's more deserving.

Just dump it out of a C130 all over Washington, Capitol Hill.

Right now, while we're talking, a couple of my colleagues are doing hand deliveries to every congressional office. So we're trying to do that.

That's great. Yeah. So. So let's go into some of the top topics in the report and what we should be focused on first and foremost.

Totally. So Calvin pushed on this project and the. And the way he broke it down, which I thought was really useful, was basically, you've got like, your national security recommendations, which are what they sound like. You know, a lot of the defense acquisition stuff, a lot of, like, making it easier to do hypersonic testing, for instance, the US and like, really defense specific stuff. Then there's like a broader, you know, industrial base section called Industrial Power that's, you know, critical minerals. It's like Small Business Administration to make it easier for small manufacturers to get going. The chunk that we really own to ifp. And again, for context, I'm at the Institute for Progress. We're a nonprofit, sorry, a nonpartisan think tank. We focus on, like, largely kind of the innovation policy side of things. So our section was like Frontier Science and Technology. So we. Not all those submissions are by us, but a lot of it, we source most of those submissions. And a lot of that stuff is just how do we get more out of the marginal dollar that goes to R&D in the US whether that's a federal dollar, R and D through like the NIH or the NSF or the Department of Energy. In a lot of those places, we think there's some really actionable stuff where you just want more bang for your buck. Let's say whether or not agnostic of whether you should be cutting nih funding by 30% or whatever, I happen to think you shouldn't, but agnostic of that conversation. We think there's still a lot of ways that you can get more juice for those squeeze. So a lot of the stuff that we were really championing in that section was, for instance, one of my favorites is my colleague Caleb Watney, who's got a piece about XLabs for science funding. He's got a new proposal. Basically, the way the NIH and the NSF and to a certain extent DOE are set up, it's really easy for them to give money to universities and academies and really hard for them to give money to other kinds of institutions, even if those other institutions might be better at the kinds of basic research or the really kind of hardware intensive high, you know, basically high capex research that you might need for certain manufacturing fields or science and tech fields. There's. Caleb gets into it pretty well, but there's basically existing authority that all of these agencies can use to just start setting up another kind of grant structure. There's no, no statutory text needed, no EO needed. They can go do it right now and you can basically, whether it's the ARC Institute, cool biomedical stuff, that's frontier, where there's some of the best stuff is not happening just at Johns Hopkins or just at these academic institutions. So that's one example. I could go, I could go all day. But like there's a lot of stuff that we really try to be as actionable as possible. Most of this stuff either needs small statutory tweaks or president or the executive branch can say let's do it right now, or it's already kind of firmly in agency authority and it's just culturally or for whatever reason, it's hard for.

Them to get going. Yeah, we were talking to Delian about this last week, this weird dynamic where arguably the greatest new innovation in science and technology came from OpenAI's nonprofit arm, which is now spinning out this for profit and everyone's going to make a ton of money off of it, is the OpenAI, I guess. What lessons are the think tanks on this project or even just what you're seeing in Washington, what lessons are they taking away from that experience? Obviously it's very different than doing basic research on the human brain, for example. But at the same time, the government didn't really fund the transformer that happened at Google. They didn't really fund the development of the initial large language model that happened at OpenAI. And yet the research did get done. And so is that coloring how policymakers are thinking about allocating funding towards research and development in the modern era?

I definitely think it is. And I think you got like a really live debate right now in like if you want to talk about the meta science field. Right. People who want to do more science on science, figure out how you get the most bang for your buck. Basically, I we at ip, this is one of our, you know, our five verticals and the five areas that we work on. And you definitely get people across the spectrum, people who think like I've heard people say, you know, like all federal R and D is like down the drain. I don't think that's true. I think there's good evidence that's not. But like you get People in this space who thinks like you have to overturn the apple cart completely. I think the things that most people agree on is increasingly you're seeing institutions that are not universities being much better at leveraging AI. And I think that's like a pretty, pretty straightforward argument. You're seeing nonprofits set up often by, you know, folks on the west coast who want to be able to leverage a lot of cap. They have a lot of capital, they don't have a lot of the other institutional ties. But you can build kind of new institutions and you're not constrained by the tenure or, you know, whatever. And if you can affiliate with, with say at Stanford, then you get the grad student pipeline. And so you still get a lot of the great talent from the academic system. I think you've seen, you know, focused research organizations or fros, where the idea is, hey, we're going to spin up for like five years to tackle one basic research problem. We're going to see if we can just crack this. We're going to be a nonprofit because it's, we're not going to commoditize ever. It's just like, it's too fundamental. But it's either a basic research thing that we think is upstream of some like whole field that other people will be able to tackle, or it's we're going to build tooling. We're going to build like an open source, like map of the brain at a level that hasn't been done before. And we're never going to make any money on it. So there's no reason to do it as a for profit company. And we're going to move at a pace that you could never do at the academy. And so we're going to try and like build the institution to solve a kind of problem instead of starting at your institution and saying how far can we take this institution? And I think that instinct that like you want to just solve specific problems and build an institution that's like built to solve that, that rather than try and pick up whatever you have and squeeze it in. I think you're seeing a lot of that in Washington right now.

Can you talk a little bit about the replication crisis? Is that front and center? What is the replication crisis in your mind and is there actually a plan to solve it? Is it something we should even be trying to solve?

Yeah, we've got a great recommendation in here from our good friend Stuart Buck, where he basically pushes us to, to get more serious at the federal level, investing, replication. I think it's a really plausible idea. It's funny you ask that. My first on campus job over the I think my freshman summer was working at a place called the center for Practical Wisdom, which was like a psych research lab. And about halfway through that job I kind of realized it's a great name. But I think I kind of had to come to Jesus knowing where I realized this research is not going to hold up. Like two years from now. This is going to be poked full of holes like what am I doing here? Narrator.

It was not practical. It was, it was.

You know, I still, I still think you can probably study wisdom, but maybe not with like these small end studies about how people appreciate paintings. It was really bad. I, I tend to think that the replication crisis stuff is more of an issue for the soft sciences, for the social sciences. I think a lot of the hard tech that you really want to push on you can get a yes or no and replication's a lot easier. I also think there's a lot of, of basically regulations or like encouraging of gating of science. The federal government does that. If you loosen that, you can make it easier for people to try and replicate stuff and to basically crush bad science quicker. What you really want is just faster turnaround cycles. Like you're, no matter what in a thriving scientific ecosystem. I think you're going to get a lot of kind of crap. You know, the question is like how quickly does that stuff get flushed out of the pipeline by good replication or by incentives that make it easier? You know, if, if, if there's a, got a bunch of great scientists out there, you're still going to get, get random stuff that doesn't replicate, but you just want to kill it early and have strong incentives for that killing to happen.

How do you think about timelines for all of these issues? One of the potential disadvantages of our democratic system is potential turnover every four years or two years in leadership. And we were talking with Kevin about sort of China's ability to do a bunch of consecutive five year plans that are kind of building on each other's. So across the board, how do you make sure that these kind of policy recommendations can really, outside of getting immediate changes implemented, have sort of longevity because in many of these, you know, we're not going to fix shipbuilding in the next three years. We're not going to, you know, know set up, you know, it's going to be hard to set up a meaningful amount of special compute zones in a few years. Right. So maybe talk to the strategy around on for you Guys, in terms of making sure there's like, really these become kind of organic movements in Washington.

Totally. There's a few different pieces. I think one is go on a lot of technology podcasts. You know, that's really simple here. No, I would say, like, at the top level, like, you just have to sell this to the Democratic. To a democratic public. And you know what Kelvin mentioned about, like, you don't have the Sputnik case, you know, short of a Taiwan invasion, there's not one single thing that's going to, you know, wake up the American people to this where they weren't on before. I think you're just going to have to sell. You just have to, you know, keep, keep making this argument. I tend to be pretty bullish on, you know, the American people's, you know, perceptiveness about this stuff, but you just do have to go sell. A lot of the recommendations in here are around shortening timelines for stuff. There's a lot of things I think you can do to just make that kind of change in administration, Democratic accountability problem, easier by just saying, like, for any given thing that you work on, can you make it last two administrations instead of three? And what are the kind of technocratic things that are standing in the way of that? What are the actual regulatory things that make you take four and a half years to get your initial NEPA permit for infrastructure structure? And I think that's like a, that's like a real issue here, right? On a long enough timeline, anything is going to get vetoed. If you have veto points throughout the process, like, the longer the timeline is, doesn't matter what it is, someone's going to kill it. And so I think you have to have like an affirmative Democratic case for, like, look, people voted for this. We're going to do it and it's going to be done. And we're not going to let somebody 15 years from now kill it. And then there's a lot of stuff in here that I think is bipartisan, where it's wonky, it's technocratic. You know, there's not like huge rifts on critical mineral production. You know, it's not huge rifts on, like, the Small Business Administration should be really effective, but you need to exercise political will to do it. And so some of it is just like really fine grain technocratic stuff that we think, like, you have agreement, but you just need to push and you need a really clear action plan for how to do it. It's not enough to just say, like, hey, shipbuilding should Be better. You need to be able to lay out like here are two things you could do tomorrow of it and it's easy.

Yeah.

I mean staying on shipbuilding, what does the vibe feel like around advanced manufacturing automation, humanoid robots? Let's not worry about the workforce at all versus let's go into reskilling, upskilling. Let's bring, let's bring back the idea that you can have a beautiful American dream career in an industry like ship building.

Yeah. Thanks to our, our friends at American Compass. You know, I'm an ifp, Kelvin's at fai. American Compass is kind of the third part of this but they spend a lot of time thinking about workforce skilling and redevelopment. So there's a lot of stuff in the industrial power section here about like what do you actually do, what's worked. Because there's, you know, it's easy to say okay, workforce retraining but like it's, it's not an easy problem. There's lots of failures there. So what do you actually, what do we actually know about how to do this? That I think everybody on this, on, on this team, like the Venn diagram is there's not going to be as many manufacturing jobs five years from now as there were at a, you know, at the peak of American manufacturing. Even if we improve output and I think, you know, your listeners will know this, that you're going to have to lean on automation pretty heavily to kind of get close to competing with China. At the same time I think there's absolutely room to say like if you, if you re industrialize and automate, that's, there's going to be a lot of jobs now there. So I think the question is just like on what scale is that, is that, you know, a World War II level mobilization? I don't think anybody here is pushing for that. You know, that more than half of, you know, all able bodied men should be working in the factory. But I think there's absolutely room to, to push, to push that even without, even without looking at those kinds of orders of magnitude.

They should be in a warehouse in El Segundo. Honestly, maybe everybody.

Yeah.

Making it rain. What, what topic in the playbook do you think is going to get the least attention this year that will get the most attention. Attention, let's say in, in, in five years. Is there a specific area? Because right now it feels like, you know, a big, a big part of this is around, you know, is, is all encompassing around kind of industrial policy. But you know, we were in Washington last week and there Was, you know, it was very bipartisan event around re industrialization. So it, it seems like everybody is, is not, not everybody. But it feels like there's a general alignment there and a lot of attention there. But of these kind of subtopics where do you think is being kind of like underrated or under discussed?

Yeah, I think probably in the discourse there's a couple I would point to. One is just nobody wants to talk or think about pandemics anymore. You know, there's just like no energy to like re litigate Covid or to kind of think ahead of plan. I think this is one of these weird lessons of COVID is like Operation Warp Speed is a massive success in a lot of ways. We learn a lot about like what should we do to prevent future pandemics or to minimize them and then just nobody has any energy left to talk about that. So I think there's some really good ideas here on like pathogen sequencing and I think there is real movement in D.C. on like all purpose like viral vaccines and a couple colleagues put together some really good work here. But it's just like nobody wants to talk about that right now. I think that's like, you know, and I get it to be honest. Then the other one is, you know, the high skilled immigration fight is going to keep happening, you know, throughout, over the course of this admin. I think you've got different parts of the coalition, you know, in the admin that have very different views on this. Obviously even among the orgs that put this together we don't all agree on everything on immigration. But there's a proposal in here that I really like that's basically DOD could be much better tracking like who are the top scientists, maybe the top thousand scientists out there who we would like to have stateside and just like keep it, keep it tight, you know, this doesn't have to be lumped in with the O one or the other, just kind of general like high level talent. But if you have like a specific list of here are the people we would love to poach from China and then just like make it extremely easy for that specific list of a thousand or two thousand people to, if they ever want to come over here, you know, they've got the kind of golden ticket path and I think there's you know, if you think that science is like long tail, that the, the really the rock star scientists, the best founders, the best R and D people are just like a few orders of magnitude better than some pretty good people. This is not a kind of Crazy idea, I think like let's, you know, poaching the highest, highest value people would have an outside impact.

Awesome.

Makes a ton of sense.

We're, we're, we're staying on shipbuilding. But thank you so much for hopping on. This was great to have you.

Come back on again soon.

Yeah, we'll have to talk more and.

Yeah, congrats on the launch.

Thanks.

See you. Cheers. Bye. And we got Phil from Dirac coming in the studio. I want a sound effect for Phil.

The man I once called a child. I didn't call him a child.

You called him a baby.

No.

You referred to him as a five year old. You said he was a toddler. No, it was very rude.

Going too far, going too far.

We've all had a live stream.

I made a big mistakes.

Things make mistakes.

I've also mispronounced. Oh tons of now we're boys.

But now he's here in Washington for the third time. Join the stream.

There he is.

Yes.

Gentlemen, how is it going?

Welcome to the stream, Phil.

Welcome to the stream.

It's been ages. I haven't talked to you on stream in three days. It's terrible.

Yeah.

I'm so glad to be back.

Back in New York, dearly.

Yes, I'm in New York in the Empire, you know, in the Empire State Building.

Capital of the Empire.

Capital of the.

Good line.

I love it. Okay, quick intro. Explain like 5 what the company does and then the announcement today.

Absolutely. So we're Dirac. We have some software that automatically generates assembly instructions for manufacturers. We like to say it's all about context aware production planning. That's not the, you know, five year old wouldn't really get that part.

Lego instructions for show.

Yeah. Imagine Ikea assembly instructions but for planes, cars, engines and now boats. And that's what we're going to talk about today.

Okay, so what happens? Break it down.

So we're talking about yachts, right? Potentially transitioning Alcatraz to pleasure yachts, super yachts, Tax haven, yacht harbor. No, sorry, go on, go on.

All about yachts. Specifically yachts that can launch projectiles.

There we go.

Got it. Okay.

The best kind. Yes.

So unfortunately, shipbuilding for a while in the US has been slowly slow slipping. We at Dirac have decided to partner with a company called Fairlead, who is a, you know, captain of maritime production in the United States. They're based out of Norfolk, Virginia. They're, you know, sort of the Norfolk shipyards. They run the, they run the place. They work with all sorts of different primes. They're a massive shipbuilder. We're talking submarines, we're talking destroyers, we're talking carriers. Fairlead has selected Dirac to basically modernize a lot of their production, planning and infrastructure.

So amazing.

Really, really excited to work with them. They are super tech forward. Funny enough, they actually launched their own venture fund earlier today.

Wow.

So great timing for that. And you know, we're excited to announce this partnership. This looks like us basically getting into the forward deployed engineer game. We're gonna, we're gonna have a guy deploy down there, modernize a lot of their infrastructure and deploy the latest and greatest Iraq technologies to their production plan. So if we.

Yeah, yeah. So break down the, you know, the massive onion that is the layers of building an actual ship. On fairlead.com I see they have an aircraft carrier, but it seems like they build systems that go into the aircraft carrier. So are they like a. Are they under a prime contractor? Do they work for Lockheed and Northrop or someone else or Huntington? I can give you. Is that right? Yes.

So I got, you know, in the, in the interest of preserving Fairleads confidentiality, I won't say exactly which primes, but they have many, many, many prime contracts that they systems directly onto. I can tell you the biggest shipbuilders and maritime producers that are primes are hunting Ingalls Bollinger, General Dynamics Electric boat. These are potentially the types of companies that hypothetically fairly be working with. And so yes, they support.

And then what are they actually building on the ship? I see some systems, very variable speed drive switchboards, control systems. Like can you concretize this? And like break it down in more simple to understand. Like is it the steering wheel that they're building? What are they doing?

So basically when you get a boat, right, you're saying it's time to ship, right? We're going to make a boat which is, you know, a shipbuilder makes a hull, right? They make this. You know, they make, they start to make the hull. They start to make various systems within that ship.

Ship.

But really they're making like a shell. What then happens is a integrator needs to come in and actually soup it up with the latest and greatest technologies. You know, they need to integrate software, they need to build in the control panels, they need to build in all the infrastructure and weapons platforms that then go on that boat. So it's different as an industry in the way that you produce than automotive, than aerospace, defense, than a variety of other industries. But when you look down like the types of things they make, these are sub assemblies that you can see on any type of platform out there in the world. So that's, you know, shipbuilding as an industry, if we were talking about it, relative to other industries, they build obviously the largest systems on the planet because those are basically boats. If you look at automotive as an industry, automotive is fairly modern. You have, you know, automated facilities. Automotive is fairly modern. Aerospace and defense as an industry is probably like around 20 years behind automotive and then shipbuilding is probably 30 or 40 years behind that. So, you know, CAD as an example of like a metric or a CAD adoption, like 3D modeling, adoption as a metric for, you know, how advanced the industry is. You know, we're talking automotive. Everything is catted up. Aerospace and defense. The majority of things at this point over the past 15, 20 years have been catted up in, you know, digital representation. In shipbuilding, you know, these, these guys are largely on paper processes and they make, you know, some of the most advanced pieces of technology that come out of an engineering department. But they're still looking at 2D drawings. They're still, you know, basically drafting pieces of paper. And that is what we're getting maritime off of. Dirac is leading the charge captaining that ship.

What does that design kind of work and production workflow look like from a, you know, prime to a fair lead to the actual, you know, manufacturing engineers that are making, you know, individual parts. And I imagine oftentimes the initial designs or the specs, you know, you're actually, I'm sure that Fairlead and Fairlead's engineers realize at different points like, hey, this worked in, you know, a CAD file, but it doesn't actually work in practice. We need to re engineer it. What does that of collaboration process look like on a general level without going into detail on Fairlead's specific process?

It looks a lot like a lot of plane tickets bought to Virginia. A lot of folks flying from DC from Norfolk to dc, from Boston, all over the place to California. It is very difficult for these folks to communicate technical designs digitally right now. And it is because it is all on paper. And so it looks like a very angry phone call down to, you know, the, down to the docks and saying, hey, we got that, you know, system that you guys shipped. This thing don't fit. You know, what's going on. So that's, it's unfortunately very frequent occurrence. You know, if you look at the average delivery date for a submarine, on average submarines are three years late. So it is really, really bad because the Barrel that everyone in maritime is staring down at at is the year 2027. It's a big year that the Navy, that the army, that everybody in the US Military is focused on because that is the year that we think, you know, some crazy stuff's going to happen with China. And so, you know, they're thinking, okay, it's 2025, we need submarines yesterday. You know, the annual production rate for submarines in the United States is one and a half submarines a year. China makes more than 10 month. And so, you know, that is a terrifying number because the. As you were kind of asking about.

Red alert, I mean that, that is staggeringly bad.

That's terrible. Can you give me a little overview of the executive order restoring America's maritime dominance? What's in there? What does it mean for shipbuilders? You America broadly and. Yeah. Does it make you more optimistic about that problem?

Yeah, I would say it makes me more optimistic. The administration, current administration, is taking it extraordinarily seriously that we need to upgrade the modern American shipbuilder. These are things that are named like that statistic that I mentioned. Statistics like this are named specifically in the executive order and they're basically calling out the shipbuilding industry saying if you are 15%, I forgot the explicit number. But if you are late by a certain percentage on your contract, you will be fine. You are subject to losing your prime contract. Like it is not okay to ship late anymore. You know, unfortunately not too many JIRA tickets flying around shipbuilding. But they ought to use something like that at least when it is a. It is a staggering problem that at least the EOs take it seriously.

Yeah. Yeah, that makes sense. Well, thank you so much for hopping on. Do you have anything else, Jordy? Now we are going to hop over to talk to Fair Lead, the other side of this deal and get Jim's take over at Fair Lead. Put the screws to him.

We're going to pitch Fair Lead on it, on leading a 100 million on 1 billion posts into Dirac. So wish us luck.

Good luck.

Well, thank you guys. I will catch you later.

We'll talk to you later.

Great to see you.

Bye.

Cheers.

And let's bring in Jim here. The other side of this. We don't talk. I mean fairly is not a company that a lot of people know about private it. It's been in business since the 80s. But fascinating to hear from a true industry insider who actually works on building these ships. You know, we've seen in Silicon Valley a lot of companies kind of nibbling around the edges, there's some, you know, a bunch of shipbuilding companies that are building smaller systems or more swarm based systems. Not really focused on the exquisite systems. Well, Fairlead is the company that actually works on the big ships, which we need to keep making making even as the landscape of the battlefield changes. They do ship repair, maritime manufacturing, privately held, veteran owned, controlled by its founder. They're still in founder mode. You love to hear it. Jerry Miller.

And they have a new venture fund. Sounds good.

They do.

I'll have to ask about that.

It was founded by a former U.S. navy Surface Warfare officer and longtime Hamptons road ship repair entrepreneur. Owns the company through closely held Miller Grant Group. In 2014, he rebranded his previous Miller IMG field operations as Fairlead Integrated and has remained chief executive officer and majority shareholder ever since. In 2019, former General Motors CEO Dan Ackerson bought a small equity stake and serves as a minority owner and strategic advisor. But no outside corporation holds a controlling interest. Anyway, is Jim available? Let's bring him into the studio. Welcome to the show, Jim. Nice to meet you.

Thanks so much. That was fantastic.

Fantastic.

Yeah, we appreciate.

And privilege.

Yeah, we're really excited to have you here. Would you mind introducing yourself and a little bit about the company? I gave a little bit of an overview, but I'd love to hear it from you.

Certainly. James Blum. I run corporate development on behalf of Jerry Miller and Fred Pasquin, the president who's running Fairlead. And we rebranded into. We now turned it into Fairlead. And we have Fairlead Ventures which we announced earlier today and very privileged to also partner with Dirac. And that announcement is coming out any moment, as I understand from Phil and the team. And at Fairlead, we're very privileged. Systems engineering, shipbuilding, ship repair. And then we work on many different advanced energy challenges that the Navy has experienced and wants to integrate and adapt and apply. And we're very fortunate to do what is known as commercial office ship shelf integrations and then harden those integrations from the world's leading industrial manufacturers and integrate them into Navy platforms. And very fortunate to support the Navy and work on a number of different key initiatives across the maritime industrial base. We're part of the submarine industrial base and we're fortunate to work on some very important problems around Columbia and the Virginia class. And then we do a lot of modular shipbuilding around aircraft carriers and all the latest designs that are coming out and applying a lot of different autonomous capability to our aircraft carrier fleet. And then we're working very closely to destroy our class as a platform and working on applying a lot of next generation energy capabilities so that they can be integrated for lasers and hypersonics. And then lastly we worked very closely across the USV space really applying very, very hardened military industrial systems engineering and critical engineering capabilities where you're really classifying a new set of software defined infrastructure that's being cascaded and fused across all the different platforms so they're more intelligent, more capable and can move faster and address a different lethality capability that we faced in the past. And now we're upgrading and, and ready for maritime dominance under the executive order that our president just announced. And then obviously with the Office of Shipbuilding coming in, shipbuilding and ship repairing our maritime dominance is something fairly that we're privileged to support the community around. And so we're really augmenting, accelerating what I would call that next generation of hardening and that capability of engineering that is critical from a quality control and speed perspective. And so that's what we're doing at Fairlead and that's why we created a venture arm and we actually launched it and announced it today. And we're, we're backing emerging managers who are leading in autonomy and AI and applying autonomy and AI and machine learning some of the most important weapon systems for, for our country and as well as to afford us that capability and deterrence and in all the key theaters that are commands are responsible for addressing.

So that's very cool.

Are you very nice to be here on.

Yeah, thanks.

Yeah. So it's a new fund folk, is it? Would you call it a fund of funds or you're investing in managers or, and direct investments or what's the focus?

Yeah, the two ladder. So we're really focused on backing the best emerging managers in the space and the best founders and entrepreneurs in the space. And that's what Phil from Dirac is an amazing example of. And we're just really fortunate to bring that kind of model based manufacturing, you know, into shipbuilding now. And one of the key areas in the country which is Norfolk and Hampton Roads and Virginia beach and then Portsmouth. So.

How is one of the, you know, Phil is extremely eloquent and brought up a bunch of interesting data points on the 15 minutes he was on before this, one of which was the United States, what we make one submarine a year at this point. China makes around 18 months. He said China makes around 10amonth. How does Fairlead think about, you know, helping close that gap over time?

Well, we, we believe that the key is workforce development and Capacity and capacity planning. And these, these systems are very exquisite and they require extraordinary amount of testing and certification. And when you want to apply next generation technology, you really have to have all those processes in place. So we're fairlead just sought out from a strategic planning perspective. We've really as a brand and under Jerry Miller and his leadership and Fred Pasquin, the entire company is dedicated to quality and to speed at scale. And so for us we sought out a company that was doing exactly what Phil and Pete and, and you know, Jared and Trevor were building and build OS for us was the software that was going to help us really accelerate that schedule and address that schedule risk you were just talking about. And so yes, the Navy and all of its partners across the defense primes, you know, across the Navy, all of them, shall I say the heritage primes and the new Neo Primes, everyone has to actually help, you know, General Dynamics and Electric Boat move faster, as fast as they can. And they're an extraordinary company with extraordinary leadership. There's a lot of technology and complexity that has to be integrated. And then quite frankly, from a shipbuilding perspective, that production schedule is very challenging. And so getting it to a position where it can introduce effectively two to three more subs over the next five years. So instead of deploying and introducing one new submarine or two in a year to up to three to four, that's going to take a very, very, very heavy lift from the ecosystem. And Fairlead is part of that ecosystem. And we think with Dirac and that modernization and that single source of truth that they bring will help a firm like ours cascade and accelerate that entire industrial defense supply chain base around the Columbia and Virginia class submarine manufacturing process. And we think we should be able to advance it rather significantly. But, but what we're talking about is really moving in production schedules from 2035 and 2030 into 2028 and 2027. And so that's a very hard lift for the Navy for its defense primes. And we're certainly with Iraq going to impact that schedule. And we think we can have a major impact not just on the acquisition in delivery side, but also the sustainment side, which is a really important part of actually if you look at how all wars are won, they're one through sustainment in their supply chain. And we're really excited about that. Fairlead it really, you know, reducing and mitigating a lot of that risk across the suppliers and across the defense industrial base, supporting the great submarine industrial base that we have in our country. And that leadership.

I'm sure a lot of founders of new defense tech companies would love to do a deal with you or sell into a company like Fairlead. Can you walk us through what it takes for a young company to to prove that they would be the right choice for a partnership with a more established company like yourself? What's the conversation like? What are the KPIs? What are the things that you need to do to de risk taking a chance on a new and growing company?

Sure. Such a great question. Succinctly they believe in a center of excellence. So from our perspective, the leadership team and their engineering and, and I'll just say all cross product, all cross engineering and the founders have to be committed to addressing the ecosystem. So what that means is literally actively having the forward deployed engineers be there to build up that new system of record with their own solution engineers and architects. But they literally will advance training for that workforce development, that workforce education, and then scaling that over a 36 to 48 month time frame. So as a result that then means that the founders and their founding investors have to have patient capital and capital from what I would call remarkable venture firms and growth equity firms and then they have to have a deep capital base because this is a very, these are three to five year sprints and the goal is to move at scale and then you're training and upskilling and then really attracting extraordinary engineers, the best builders in our country, to actually a new space for them which is maritime. And this maritime industrial base for maritime dominance that requires a heavy lift on an earlier stage firm. So what I would say is almost series B and series C firms in general are a better fit. Unless of course they have leadership that is extraordinary and deep. Deep like Dirac is and is a great example of significant expertise in the maritime industrial base, significant understanding of how electric boat and General Dynamics actually works as a company. And then having that knowledge and then being able to scale that knowledge, build that into a software system which is a platform that we can integrate all of our shipbuilding and our systems engineering and a lot of the product development that we do and then applying a lot of autonomy and AI stack into that from a work assembly perspective, from a modularity perspective, that's a heavy lift for most companies. Dirac had all of the above and then they had extraordinary founding investors behind them from a seed perspective. And then they have a broad ecosystem support across what I would call the venture capital stack and that capital investment partnership that they have have cascading across the country both on The east coast and the west coast meant for Fair Lead, they had mitigated all the risk for us, as well as the maturity of their software and then their commitment to build a center of excellence in Norfolk around that entire community where the Navy has a very significant amount of its fleet and manufacturing base. So those were the key requirements. And I would say that they also had the technology roadmap that we thought was absolutely core and critical to us. And we'd been seeking that out for the last nine months under Fred Pasquin and Hal Barsch. And Hal really leads that for Fair Lead, and he was extraordinarily impressed with the presentation by Jared and Phil. And just to give you a sense of how fast we move, we moved from meeting them to actually executing a purchase order and agreement and a commitment to getting trained and scaling this fairly in under three and a half weeks. That's how fast we move at speed and scale. So for any founder out there who believes that they have the next piece of the core tech stack stack that will unlock capacity, scale and autonomy and manufacturing for our country, yeah, we can. We. We move much faster than most industrial enterprises. And we're. That's why we have a venture arm that we launched because we wanted to back that with capital in addition to being a great customer and a great partner.

That's very cool. How. How is Fairlead thinking about opportunities to support various primes on smaller autonomous systems? It's obviously very different manufacturing process to make 1000 of something versus 1 or 2 of something. And I know it's top of mind for you guys.

Yeah, so well said. We began this journey back in 2015, really trying to back as many new founders who were really early in that smaller autonomous stack. And so we're just been very fortunate to back a number of new and traditional players in the USC space. And we really started really on the system engineering and systems integration side. And we found now that we're really, we've been very lucky that the Navy has been so supportive of us that we're now working with many of the. Of the brands who are out there who well published, doing an incredible job raising capital, doing an incredible job pioneering in what I would call smaller hybrid fleets and introducing new port capabilities, terminal capabilities, the full value chain. And so we're just really privileged to work with, you know what I would say the new entrants to the space as well as the heritage players. And what we're fortunate to say is we're working with almost all of them and we really take it from a software and systems engineering perspective and what we're really trying to do is harden the navigation systems, harden the autonomous systems from cyber attacks, building out mission modules and payloads to support them. So as you know, many of these remarkable firms with remarkable capital are really building out these new vessels which are, we need and are fantastic. But actually integrating the, that from a payload and cargo and high value, high consequence cargo perspective, that's actually an extraordinary, shall I say, level of systems integration that's needed. We just feel like we're, we're just a natural partner and the majority of the industry has selected Fairlead. So we look at everybody as a partner and we're really looking at their timeline to accelerate production so they can field and get tested and certified and introduced into theater and so that we can address that schedule risk that our President and you know, our new Secretary of the Navy is very, very adamant about reaching. And so we just feel that we're just, we're very lucky to be in that position to have the resources, the assets, the real estate, the capital, the leadership, the expertise. And quite frankly we're now actually a prime ourselves. So been very, very.

Congratulations.

Yeah, thank you.

Well, James, if you've ever, if you ever to want, want us to come podcast on a submarine, let us know.

We're here, we're here.

Ready gentlemen.

We'll introduce you to all the right people.

We're all congrats on the launch of the new fund, the partnership and it's, it's great to chat with you guys. Thanks for all the context.

Yeah, we'll talk to you soon. Have a great one.

Bye, thanks so much. Take care. Bye.

Cheers.

And next up we have Jacob from New Limit talking about life extension. Can you live forever or are we trapped in this mortal coil? We will get to the bottom of it, but I mean new and it's a fantastic company. Brian Armstrong has been working on it for a while, brought in a fantastic team and they've been doing a bunch of stuff. So Jacob, welcome to the stream. Thank you so much for joining. Could we start with a quick introduction on yourself and the company just for those who don't know?

Yeah, absolutely. Thank you all for having us today. So my name is Jacob Kimmel, one of the co founders and president of New Limit. So New Limit is working on a new type of medicine to try and extend human health span. By which we mean we want everyone to have more happy healthy years in their lives than they do. And we're doing that using a technology called epigenetic. Reprogramming, which we're happy to unpack if it's of interest.

Absolutely. Let's unpack it.

Let's unpack it.

Okay, let's unpack it.

I think it's an interesting. All cells in your body, our human audience, there might be some AI agents.

Listening that not so much.

The ape audience could also benefit if there are. Who are currently tuned in. So all the cells in your body, they've all got the same DNA code, which is both kind of trivial. You probably learned that in third grade biology class. But it also presents a wonderful question, which is then, well, how do your eyeball and your kidney and your tongue all do things? They have the same code base, it turns out, on top of your DNA, and some of the proteins your DNA wraps around, there are these chemical modifications called epigenetic marks. I think of these kind of like the control flow in software, where not every line of your code base runs every time a user hits your application service. You use some pieces of the code base sometimes and others other times. And so your genome's the same way. Some cells express some genes from your genome, others express a different set, and that's how they do different jobs. Turns out, as you age, all of those epigenetic marks get, for lack of a better technical term, messed up. And so your cells stop using the right genes at the right time, they wrong genes at the wrong time, and then they degrade in function. And that actually precipitates a number of the diseases and pathologies that arise with age, including things that we don't even traditionally define as disease, but that you or I would certainly like to remedy. About our experience, if we could, is.

The plan here kind of traditional FDA biotech? I'm thinking like some research, then some mice, then dog or monkey, then human trials, then phase one, phase two, phase three. Main your public company at some point, or is there some sort of Silicon Valley twist or different approach that you're taking?

Yeah. So all of the downstream steps are pretty much what you just laid out. Ultimately, in order to get a medicine into humans, we have to first do no harm, demonstrates the medicine has some reasonable chance of actually helping a person, and then definitely isn't going to harm anybody when you put it into them. So we do need to do animal trials. We then eventually need to go into the clinic. I think the Silicon Valley twist here is that the problem we're trying to solve with how to basically rewrite these epigenetic marks back to the state you had when you were young. Really wouldn't have been tractable even 10 years ago without basically two technologies. One of those is something called single cell genomics. So we can actually take a cell out and sequence all of the different RNAs in that cell, basically telling you which parts of the genome it's using. And then the second, which I'm sure your listeners are intimately familiar with, is AI. There are more ways we could try and rewrite this epigenome than we'll ever be able to test in the lab. It doesn't matter how clever you are, the number here is about 10 to 16, which is a stupidly large scientific notation number to put it in context. Context, that's like 10,000 Milky Ways worth of stars. You just can't test all of those things. And so by building AI models, we're able to actually prioritize which of those potential reprogramming medicines we even go bother to test in the lab. And that helps us get to these discoveries much faster. I really don't believe this company would have been a good idea prior to the advent of really those two key technologies. And so we're trying to parallelize as much as we can early on so that by the time we get to that traditional drug discovery recovery process, we have a lot more conviction around the molecules. We're moving forward than you might otherwise.

Can you talk about the different kind of milestones that you sort of that have been important to you and important to your partners to get to this point? It's much different than SaaS where maybe you get funded with an idea and a deck like every company does, and then you get to maybe 2 million of ARR, you do a series A, 10 million, you do a series B. I imagine the milestones have been much different here, but I'm sure you guys hold yourself to a pretty extreme standard around continuously de risking the opportunity and getting more and more convicted. But I'm curious what those have been to date.

Yeah, absolutely. I love. Larry laid that out. One of the challenges with our technology and biotech in particular, is the milestones are not quite as legible as revenue for quite a while. While, you know, famously some of the largest companies in the space took quite a while to even get to revenue positive. And the reason these bets still make sense is that once they do have revenue, you really start printing it. There are more drugs making a billion dollars in revenue today than I believe there are software companies. So it's truly amazing to recognize like a single product is transformational. So all of these hard bets make Sense, even if there is a challenge in making them legible along the way. So for us really there were a few key milestones. When we started out and built the company company, all of this was effectively an idea. There was an existence proof where we knew it was possible to reset the epigenomes in old cell back to a younger state. But that existence proof was something that in the field we often call a tool compound. It tells you that the biology is tractable, but it's not a medicine you can put in people. It basically works, but it's not safe. And so we had to first build all of the technology that would even let us run these experiments in the first place, just to give you a sense of scale. Before New Limit people had really only tried 19 different ways to do this sort of epigenetic reprogramming intervention, make an old cell. Sixteen of those I had done in my old lab with my own hands. Really this was just a gold mine that not many people were digging in. It's just me there with a tiny little plastic shovel. And so at New Limit now, we've tested about 14,000 of the different ways to reprogram cells. So we like to say it's 6 or 700x depending on if you let me round up or not. And we had to build all the molecular tools to even let us run those experiments. That's sort of step one. Step two is we then have to demonstrate that actually, even though I'm really proud of the fact that we can run 14,000 experiments, it's way smaller than 10 to the 16. We can build AI models that sufficient to actually prioritize which of those experiments are worthwhile. If you imagine that there's some gold out there. But there's just way more places to dig than we can really exhaustively search through. We need a way to find where those spots are. So we were able to demonstrate that. Now we think we have the most performant models in the field. Really, thanks to just the scale of data we've generated, we have hundreds x more than anyone else does to train these types of tools. And then one of the final pieces, and this was the last unlock that actually came a lot faster than we thought it would, is we demonstrated what in the biotech community is known as a preclinical proof of concept. In layman's terms, we've built some molecules that are actually drug like. So you can actually hold these in your hand. You can hold a tube of a potential reprogramming medicine, put that into an animal model of a disease and rescue the function of that diseased animal. We can do this both with human cells. In our case, we're able to take old liver cells from old people and restore them back to a youthful function so they can regenerate and repair a tissue the same way young ones do. And then likewise, we can put these molecules into animal models. And one of the diseases we're going after is alcohol related liver disease. It's some damage that accumulates from alcohol exposure over the course of your life. Not only can we make those cells regenerate better, but we can also just make them much more resilient to alcohol. It turns out that as your hepatocytes, the cells in your liver age, they become much more vulnerable to alcohol when you drink. That's an experiment that I've run in my own life. And so I believe the data, even without having to replicate it as many times as we have, and we're able to really improve the protection they have against that type of damage in a way that we think will really benefit patients. So to our knowledge, that's the first time anyone's really made a reprogramming medicine you can hold in your hand that demonstrates efficacy, something you could imagine carrying forward to the clinic.

Can you talk about the Series B? $130 million. I want to know about the round, but then I also want to know about what do, what's the use of funds look like? Because you're doing a lot with data. Are, are you going to spend $100 million in some sort of large H100 cluster to crunch all these numbers? Are we at that scale yet or is it just R and D salaries for engineers, scientists? Are there legal folks putting together FDA applications? Is it a bunch of money on lab mice and monkeys? What is the shape of the business right now?

Yeah, absolutely. So everything you noted is some expense. Thankfully, our compute budget quite isn't 100% million yet, though I hope we're generating enough data to get there soon. Jensen can donate some GPUs if he's.

Feeling so kind, if he's serious about living to 150.

Yeah, I mean, it's in his own interest. He wants to be running Nvidia for the next 40 years. Why not?

It's true. So, yeah, the breakdown is basically the main costs are R and D. So we have three different therapeutic programs. Our most advanced program I told you a little bit about is focused on restoring function in the liver. We hope that can treat some advanced liver disease diseases and eventually Treat everyone whose liver ages. You probably have a little bit less energy than you did when you were 21. I certainly do as well. And that turns out that as we age, over half the population starts to suffer from something called metabolic syndrome. Basically, your rates of heart attack, diabetes, obesity, all shoot up. We see this in our own lives with people we know. It's just we don't necessarily have the language for it normally. And we think those medicines can treat a very broad swath of patients. We also have programs in immunology trying to restore youthful functioning, your immune system and more recently vasculature, trying to basically fix your kidneys and other portions of your body that degenerate when you can't really move blood around the way you did when you were younger. So that's a large fraction to spend and then another big fraction of the use of proceeds is really to take these molecules toward the clinic. As you start to do experiments in humans, things get very expensive very quickly. It turns out even just manufacturing the molecules that you need to go take into a clinical trial has to be done in a really controlled way. You're also just making very large mass masses of this stuff and it's very expensive to do. And so because of that, we try and really upfront de risk everything before we get to the clinic. But we still have to reserve a fair amount of the dry powder to make sure we can do right by the patients we're going to treat, produce the drug in a sufficient quantity, run a sufficiently sophisticated trial to really gain evidence that it's working.

Can you talk about the pressure of working on technology that so many people are effectively betting on and some somewhat, in some ways feel is inevitable? I feel like if you ever have a conversation with a loved one, you know, a parent, you know, you're talking with, let's say a parent, they'll be like, you know, maybe a young person is thinking on a 30 year time horizon and the parent is thinking, well, I don't, you know, I don't know if I'll be around then. And then the, you know, younger person says, oh, well, by that point, like we'll have life. You know, I just feel like there's something that comes up constantly that people like, maybe, maybe it's, you know, maybe it's misguided, but I, I feel like there's a general sense of inevitability with these types of technologies. But when you guys are actually doing the research and, and fighting these, you know, fighting to make this possible, I imagine that at times you know, maybe in the fullness of time you feel like it's inevitable, but it's, it's certainly not pre programmed. Right. It's something, it's knowledge that needs to be earned. But it's a much different level of pressure than somebody raising $130 billion for a SaaS company where if it fails, there'll be another note taking app that just sort of takes a place and we'll still be able to organize words on digital documents.

Maybe for your note taking app, Jordy, My note taking app is going to be entirely unique and irreplaceable.

Yeah, exactly. Yeah.

I love the way you set this up. It is something we think a lot about, which is the responsibility we have as we're trying to work on these tools. I think in biotech you don't get many opportunities to take really audacious bets like this. The science we're working on is earlier stage than you typically start a biotech on. There's more inherent tech risk here than many companies are willing to underwrite. And so because we do have one of the unique opportunities to take a shot on goal at potential, essentially a thousand x outcome, a medicine that everyone would benefit from over a certain age, we feel really a burden to make sure we're using that opportunity responsibly. And so I do think there's some amount of stress in it, but I think it's mostly positive in that I feel like often we're presented with challenges where if we shrink the scope of the ambition a little bit, we can make our lives a lot easier. Well, certainly we can get to the clinic faster if we're comfortable just having a medicine that maybe doesn't apply to as many people or doesn't work quite as well and having that burden to say no. We've got one of the few rare opportunities in the industry stewarding this capital, working with these generational investors to be able to try and take a shot where really we are optimizing for the scale of the outcome rather than just the marginal basis point on probability of success, which I think is something that biotech has maybe forgotten over the past decade or so. We've really optimized on probability of success, but forgotten the other half of the EV equation. If you multiply even a high probability by a low potential value, your outcomes aren't generating as much as you'd hope for anyway. So it is something that's always front of mind for us.

Makes sense.

Yeah, makes sense. How do you think about the potential impact of, you know, Improving the function of, of the kidney by 10% could have some, you know, additional externality in the body that could actually have a really meaningful, you know, so how do you think about kind of complex systems and the sort of unintentional impacts of certain sort of effects of the type of drugs that you guys will develop over time?

Yeah, it's a great question. It's something like we don't have a super rigorous framework for this because these questions are very difficult to answer quantitatively. I guess in short, we like to treat those sorts of knock on benefits as pure upside. So before we launch into a program, we need to convince ourselves that just making, making one type of cell in your body younger is actually going to benefit you. And that might sound kind of obvious, like, sure, I definitely want younger cells, but if you then step back and realize your body's hundreds of complex cell types, they're all interacting well, maybe one really solid link in an otherwise pretty busted rusted chain isn't actually all that useful. So we try and find examples where if we're able to demonstrate that restoring function in just that one cell type helps individuals, especially helps humans, then we can get conviction around starting a program and then everything in terms of the KN benefit. Maybe if you had a healthier liver or a healthy kidney, we'll see benefits elsewhere in your physiology that maybe we haven't even directly measured yet. We just treat as upside potential for the long term.

Yeah, I have a little lightning round. I want your reaction to three kind of biotech projects or technologies and I want to know if they're at all relevant to the work that you're doing, if you're learning anything from them or just the overall impact on help, health and just general biotech community. Ozempic, Crispr and Alphafold. So I'd love to start, I guess with Ozempic. It feels like, you know, it's not necessarily life extension, but in terms of health span could be very good. But how have you been processing the story of GLP1s and are there any learnings from GLP1s that you've been able to kind of port back or update your philosophy or strategy at new limit?

Absolutely. I think you're entirely right in saying these are some of the first healthspan medicines. I'd argue we actually have a couple already. This isn't a totally crazy idea. There's another type of medicine called statins, they're a bit older. One brand name, one's called Lipitor you've probably heard of. And between statins and GLP1s. Now I think it's quite plausible if you do some napkin math, we're adding several healthy years to the median American's life, which very few other medicines can claim. If you look at the rest of therapeutics development over the past decade and add up how many years could you, I, Jordi, maybe expect from these benefits, it's actually quite small. So I think one of the things we've internalized is that the scale of the opportunity really does matter. The difference between the GLP1s and many other drugs is that the axis of biology they're targeting actually is dysfunctional. In a huge swath of the population, at least 30% of Americans, probably more, depending on exactly how you define that. And when you strike those really high value opportunities, I think exactly as you alluded to to, you start to see knock on benefits maybe you didn't even plan for initially. You know, GLP1s are now showing benefits in cardiovascular diseases. There's early data in neuro that I don't think anyone quite understands, but does seem to be reasonably consistent. And so trying to keep your mind open.

Stuff specifically like the addiction and gambling stuff you're referring to?

No, no. So there's actually even neurodegeneration. So there are some early trials that like Volks on Glip 1s are having lower sort of slower disease progression for all Alzheimer's and Parkinson's. As of my understanding right now it's like basically no one knows how that works, but you're starting to see results like that pop up. Just because these drugs are out into such a broad swath of the population, we're kind of running these experiments without even necessarily intending to. And so I think really we've just tried to internalize the scope of the opportunity really matters. And then also the strategy they took to get the GLP1s into the market is one we're trying to emulate. They didn't start by trying to dose the GLP1s into a huge population regulation. You start with a small patient pool that you know will benefit a huge amount. So in the case of ones certain type of diabetics, eventually you expand to the obese, eventually you expand to other indications from there. And so we're really trying to mirror that stepwise with our programs where maybe as an example, we're starting an alcohol related liver disease for our metabolism drug. But that's the beginning, we don't intend it as the destination. And we've tried to plan this in such a way that with, you know, steadily increasing quantums of capital, you can unlock larger and larger opportunities rather than just trying to shoot for the home run right off the bat.

Yeah. Do you want to stay on Ozempic before we move on to CRISPR and AlphaFold?

Oh, yeah, yeah.

I want to get your take on crispr. Obviously it was extremely hot technology about a decade ago. Is that at all relevant to what you're doing or have you learned anything from the story of CRISPR and how that was ultimately commercialized and baked into the rest of the biotech community?

Yeah, absolutely. Maybe two things. I think the impact of CRISPR is somehow both over and underhyped in different places. It's maybe overhyped a bit in terms of direct application to therapeutics. I think if you go out and look at sort of how the panoply of different CRISPR companies are doing, there are a few really stellar examples like Beam therapeutics that we really admire who have managed to make compelling medicines and others have struggled. And so maybe there's a bit of hype there, but it's really undervalued on the research side. So we've used a lot of CRISPR tools on the discovery end to be able to ask questions at the level of very basic discovery experiments. What happens when I turn this gene on or off? And maybe we're not going to take those CRISPR tools and trying to actually put them in a person as the medicine itself, but they certainly help us accelerate the discovery of other therapies. And so I think maybe the lesson to draw out of that is when we have breakthrough technologies, often people rush to what is the most direct application therapeutically, but undervalue, sort of the knock on benefits of that technology underlie and accelerating a lot of trends that are already at play.

And then similarly, Alphafold, massive breakthrough in terms of AI and bio didn't really move the biotech markets when the news broke. And my interpretation of that was that maybe protein folding as a cost center for biotech development was actually not that much of a cost driver in the sense that it's mostly just farmed out to grad students, maybe, but. But I'm sure there's things where you look at that and you think this is going to. This is on the path to really, really powerful AI and bio working together. But are you using Alphafold or anything like it, or are you learning anything from it? And overall, do I have a right understanding about the impact of Alphafold on biotechnology broadly?

Yeah, yeah, I think you've hit the nail on the head with the direct impact. You know, I think solving protein fold folding itself, like actually taking a sequence, predicting a structure, hasn't been rate limiting for drug discovery in the traditional sense. It is painful to get structures, but usually we work on few enough molecules that getting the structure is not really what slows you down for most traditional therapeutics. That said, I do think it demonstrates we're able to learn much more powerful representations of biology than others thought possible. And so maybe by analogy to crispr, the direct application. Is predicting structures itself really going to solve therapeutic development? No, probably not. Is learning the language of life in a way that we can actually work with tractably going to accelerate everyone in the ecosystem? Absolutely, yes. And so to your question of are we using something similar? We definitely do. So one of the tools that we've built, the models we've built, they try and take in some notion of the genes we're manipulating transcription factors. You needed to learn in embedding, just like in embedding in natural language processing, we're all familiar with that you can get from LLMs and then predict what will happen to a cell's age as a result. And it turns out some of the best embeddings we've found are from these protein language models that really are often deriving from that basic premise of the alphafold like tools. And so I don't think if you see the AlphaFold announcement you'd ever intuit, oh, solving protein folding is going to make it easier to find reprogramming drugs that make old liver cells act young. But it turns out that that underlying substrata of learning these representations actually does accelerate a whole of bunch, bunch of science you don't expect. And so I think often the headline is a little overhyped, but the subtext is underhyped. And probably more important.

That's an awesome framework.

How much does bio, the the biotech industry broadly pay attention to biohacking and citizen science? I know, I know you can't read too much into what 1x account is experimenting with on any given day, whether it's it be Cialis. Cialis or sleeping, sleeping nine hours a day. But the other one that was interesting recently was a guy, I think it was last Friday who was let, he let like hundreds of snakes bite him.

Oh yeah.

Over time. And so I feel like there, there, there is alpha in citizen science, but there's also, you know, 90.

You got to take it with a grain of salt.

Yeah. And it's also not, you know, they're.

Not Running with a spoonful of cobra venom.

This guy's a legend. I wish I had the opportunity to shake his hand because he's definitely underappreciated. So, writ large, I think the pharma and biotech ecosystem are actually fairly receptive to using the results of citizen science because it's so hard to get data in humans that even if the data you're collecting in a human is confounded in some way, yes, this person didn't do a program, perfect, controlled trial, et cetera, it can still be better than having no data at all. So I think the two qualifications on this, and maybe the examples you gave, are good ones. One is when you've got a very small trial, like a single person, but you've got a huge effect size, like this guy who's immune to basically every snake bite. That is such crazy outlier biology that I think that wakes people up to, okay, there's something to learn here. And the other is when you actually have this replicated, maybe unintentionally. So a good example I think you alluded to earlier, the GLP1s and sort of alcohol use, where this wasn't something that the companies were exploring, Eli Lilly or Novo. It's something that started getting reported anecdotally from patients. Hey, I just don't want to drink anymore. And so the amount of drinking I've done has gone down a lot. And so that was in a way, a type of citizen science, people recording, reporting on their own behavior. And now those are actually going through clinical trials in a more formal process. And so in the industry, people call this reverse translation, and it often and has a lot of fancy words dressed up on it. But the basic idea being that the place we can learn the most biology in humans is in humans, and we should try and treat every piece of data we have preciously. I think most people would appreciate there's probably not as much active work trying to collect this sort of citizen science data as there could be. But I think there are a few strong examples where it's really changed. Drug discovery.

No, this was fantastic.

This is fantastic.

This is great.

I do want to let you kind of give an overview of the company. Like I imagine you're hiring. What type of roles are you hiring for? What's next on the agenda for new limit?

Yeah, absolutely. So over the next few years, part of the use of proceeds here is actually to take these prototype medicines we have today and move those into the clinic, actually take them to patients for the first time. And so we're hiring across the board, really, for drug developers and therapeutic area leaders to help run some of our therapeutic programs. For scientific scientists who are skilled in functional genomics. And on the computational side, we have a growing ML team. And for those sorts of individuals who might be excited about AI and bio, the carrot we can offer is we have some very differentiated data. We have the largest data set of this kind by hundreds of fold. You really need to get to this scale before you can train meaningful models. And we think new limit is one of the few places you can go and expend your valuable talents where the predictions of your models are actually going to get tested in a real laboratory territory fairly soon. So you've got real contact with the universe. You're not too far from the metal. So would love to be in touch with anyone who's interested in any of those verticals.

Amazing. Well, thank you so much for taking the time.

Thank you. Fantastic doing this work.

Thank you for saving my life in the future because I'm putting the car before the horse on this.

We're making, you know, our. Our sons are about, you know, under five, and we're making succession plans with them. In about 30 years, they'll be taking over the microphones. But if we could get another, you know, 30 years at a new limit, that would be ideal.

So the world needs parallelized technology. Brothers podcast, it's required. So we'll try and do our best to make that happen. Thanks.

Great talking, Jacob. Congrats to you and the team.

We'll talk to you later. Thank you so much. Bye. Let's run through some posts and get out of here. Thank you to Brian Villa, who went ahead and started a list for following everyone that has ever been on tvpn. I thought this was really cool, working through the TPPN feed slowly to add all the guests they've ever had. He said, give me a few days so you can go follow this list. We'll quote post it, but we were thinking about doing this and he's just doing it for us. So thank you. Citizen journalism there. Citizen podcasting journalism. We think of ourselves as corporate podcasters, but corporate media. But it takes a village.

Thank you, Brian. Very cool. I'm gonna go follow the list.

Also, Eric Dungan says, I swear TVPN has me wearing a nice watch and buying Dupont Registry at the airport. If I start wearing a suit, then they won.

Oh, it's great cracking open the.

And he's got the wedding ring on. You know, he's family man. You love to see it.

There we go.

I can't clock what watch that is, but it looks fantastic.

Eric's a Day One fan.

I see.

Great having you as a technology brother.

And it's nice to have a. Have a watch on. It looks great. And Dupont Registry. I'm a big fan. I follow them on X pretty much every day. I'm always tagging my friends in it. This is a good option for you.

Yeah.

Anyway, we saw Greg Brockman at the Met Gala, which we talked about. This was a fun one. Brex had just launched a podcast with the. With the CEO of Deal, and I feel like they must have recorded this before all the drama. And then they just decided to release it once the things that kind of cooled down. But anyway, I invited Alex from Deal on the show. He got back to me, said he can't come on right now. Obviously things are a little chaotic, but we would love to have him. There's a bunch of interesting questions. I want to know about the business. I also want to know about the drama. And it's a live show, so we won't edit you, Alex, if you come on the show, we can't. You can. You can deliver whatever message you want. Always interested.

Yeah. This was.

It would be a great comeback story.

Certainly a dramatic way to launch a podcast.

It is a crazy episode one.

Yeah.

Like, of all the people like, this is. This is who people want to hear about. I wouldn't be surprised if this is a really, really popular episode. Although it seems like it's more focused on the run rate, the scale, the employees, the business. But hadn't really heard much from Alex before all the. All the drama. But interested to hear about him now, you know?

Yeah.

Anyway, Shweta says never underestimate the power of one. Good, good tweet. Roark's founders were $15,000 in credit card debt. Been there and sleeping on a friend's mattress when Matt's post went viral. 15 minutes later, Austin Allred wired 100k. By the end of the day, they closed 350k, leading to a $2.8 million seed round led by a 16z conviction.

Is crazy.

And Matt says, my jaw dropped. Roark lets you create entire iOS apps just by describing them. Zero code required. This changes everything. For app development, Roark blows Bolt out of the water. And yes, I invested in immediately after trying it. Watch this. It's crazy. What a great viral market entry. I mean, this really does seem like just part of the standard playbook of launching a seed stage company at this point is like figure out a way to go viral and just fill up that first, you know, batch of customers to test on.

Yeah.

Might not be a sustainable growth mechanism.

But it's almost getting to the point where if people, people will judge you as a founder, CEO, if you can't nail a sales call. Right. It's extremely bearish.

Totally.

Even if you're not every, every founder should figure out how to sell the product. Right. Because you have to sell employees, partners, investors, you know, the list goes on. Being able to manufacture virality, just figuring out how to go viral is. Is getting to the point where it's going to be required in the same way that just being able to nail a sales call is right.

Yeah.

So 100%.

Well, speaking of that watch we saw earlier, if you're looking for a new watch, you want to join the ranks of our listeners with fantastic watches, head over to getbezel.com your bezel concierge is available to now to source any watch on the planet. Seriously, any watch head over there. Keon Nucleus Genomics friend of the show is over at the Ramp office. Another friend of the show doing a little collab. All Ramp employees can use their wellness benefit to get the world's most advanced DNA health test. He'll be at Ramp's office for lunchtime.

Love how he just sets up a little stance.

Selling to Ramp employees. I love it. Nicely designed little partners. I thought that was cool. So congrats to him on that. Michael Dempsey says, okay, maybe the best venture swag of all time time from Hooty R and I. It has a badge here that says what has show capital Fund lp. What do you think of this? You're. You're the merch guy. You have put together some fantastic drops.

I like, I think it's pretty cool. I like the pockets.

I think it's a different look. You get a lot of T shirts and you know, T shirts are great, but you know, in the fabric, this.

Is fabric, fabric, very nice.

Something a little different.

It's, it's got some weight to it.

Didn't Bain Capital get in trouble for doing the car heart? And it was like a little larpy stored in stolen $. This feels like it evokes like the workman. You could, you could do some work in this, but it's not trying to be like perfectly aligned with another, you know, a cohort of individuals. And so it's just kind of its own thing. That was cool. So congrats to those and we'll close with this post from Paula over at XAI says who recently joined X. Yeah, Dramatic move. Congratulations. Paula Rambles. So who's building the app at the intersection of AI and astrology? I actually told Sean over at my first million when I went on over a year ago that this was a prime area for LLMs because there's something called the Barnum effect. Yes, PT Barnum. So the Barnum effect is essentially there are a set of phrases that you can say to someone that sound hyper specific, but in fact apply to everyone. So an example would be like, you have a complicated relationship with your family. Sounds like I know you, but really everyone kind of has a complicated relationship with their family in some one way or another. Yeah. Or you think you're meant for more. Sounds really specific. But really everyone thinks that they could do more. You know, everyone has these aspirations.

What's your. What's your astrological sign, John Matoris?

Bull.

Okay, I'm asking three. My. My business partner is a Taurus. How should I communicate? How should I backstab him today?

But yes, I think that astrology and AI go hand in hand in the sense that you can download this app and with only knowing a little bit, you can start producing really, really tailored responses that feel very specific to the individual. Create this. And even if you're just regurgitating generic life coaching advice, everyone needs to hear, you got this. You can do it. Your astrological sign says that it's fortune cookie.

Yeah, I mean, the thing here is I think there's an opportunity to build a relatively thin wrapper on top of whatever your preferred model is.

Yep.

I believe that many people will just use their preferred LLM for a lot of this stuff, Right?

Totally.

Like I just asked. Oh, three.

Yeah.

You might not need to go get a separate app, but at the same time, if you're. If you design the app with the right aesthetics, right functionality, you do the right customer acquisition on Facebook and then you purchase the subscription.

Like, I think a lot of it could be its own astrology. Apps that have raised venture have actually done very well. So here's my do's for communication with John today. What to do Open with appreciation. Hey, John, you know, I think you did a really great job on the podcast today. Everything from the prep to your zingers to your interview questions were fantastic. But we gotta talk after the show because present. No, no, no. Present. Concrete, tangible data. You know, John, the number of times you said hit the soundboard, you know, it meant a lot to me. So thank you for that.

You're welcome.

Jordy Frame ideas As we projects. You know, John, the prep you did for the show today, I think we did a great job. Common and again, I think this is like the challenge here is this is like probably very general advice, but it.

Can all work out. If you're into. Who knows? If you're into astrology, go check it out. Go. Let us know.

Funny it says here no, no rush deadlines, but you're actually very good with deadlines.

I love deadlines. Oh, we have to go live in 30 minutes and we don't have prep.

Done, done.

No problem.

Done. You're the kid who waits until the last 30 minutes.

Always do your homework same day in the hallway before you go into the class. Anyway, thank you to Ramp. Polymarket Public Numeral AD Quick 8 Sleep, Wander, Bezel, Linear Figma and Vanta for.

Incredible partners of TBPN more than sponsors.

And we're gonna be up in San Francisco with the Figma team tomorrow at config.

Config.

Config.

Config.

Config.

Config. Config. We're excited to be up in the bay tomorrow. If you were in sf.

Yeah. If you're. If you're at. If you're at config, let us know. Seriously, hit us up. We'd love to have you on. Talk about design, talk about figma, talk about vibe coding. Vibe designing. We're going to do it all anyway. Thank you for watching. Watching. Fantastic show. We will see you tomorrow. Have a great rest of your day.

Can't wait.

Bye.

See ya.