Shaun Maguire, Nic Carter, Jeremy Giffon, Trump Stands His Ground, 20K/Month GPT

You're watching tvpn. We're live from the temple of technology, the fortress of finance, the capital of capital, the dojo of the dollar, the shrine of shareholder value. And today is March 5th, Wednesday, it's 2025, and we got a great show for you. The show starts now. We got a bunch of call ins. We got Jeremy Jiffon. Jeremy, we got Nick Carter calling in talking about CoreWeave, talking about crypto. I'm sure we'll. I'm sure we'll ask him a bunch of stuff. And we got Sean McGuire talking about, talking about the Elon empire. The empire that Elon Musk has built across every.

Sean, of course, has his tentacles all over.

Yes. And that should be the puppet master. The puppet master.

He's the man pulling the man behind the curtain. Yeah, yeah.

The wizard of Oz.

Exactly.

Yeah. We've given our studio nicknames. We need to give our guests nicknames. We've done this pretty well with David Senra, the godfather of podcasting the most powerful man in Miami. Yes. Anyway, speaking of very powerful men, we are breaking down, of course, politics. Trump gave a speech. I didn't watch it, but I did control F and C. Did he mention tech? Did he mention chips and. Yes, he mentioned tech twice. And he said Ronald Reagan wanted to do it long ago, but the technology just wasn't there. He's referring to. He's asking Congress to fund a state of the art golden dome missile defense shield to protect our homeland. I mean, golden dome. Let's go.

I mean, obviously referencing the iron dome. Yes. But has to do it one greater.

Hilarious. Hilarious coinage. I think I'm on board for golden dome. This does seem big for, you know, obviously politics. We try and stay away from it, but like the president is saying, we're going to spend money on something that technologists build, like tech.

So if you're in the golden dome business.

Yes.

This is a huge opportunity.

Exactly, exactly.

That's why we wanted to call it out.

Yeah.

I really also, I mean, just generally, you have to be pro dome. Just generally.

And I recommend everyone build a dome. Had a golden dome already.

Yeah.

So. So whenever.

Star Wars. Do you remember Star Wars? Not that. Not the show. No, not the movie.

Oh, right.

This is the Reagan. So Ronald Reagan, what he's referencing here says Ronald Reagan wanted to do it a long time ago, referring to a golden dome, but the technology just wasn't there. The idea was Ronald Reagan had this Star wars program. It got called Star wars after the fact. I think is a pejorative. They were like, oh, he's going to put satellites in space and they're going to shoot down missiles. Like, this guy's stupid. And then like, there was, like, a hit piece, and they're like, oh, he just watched Star wars and he wants to do it in real life. And now it's like, no, actually, we do need that. That's. That's a good thing. The same thing happened with Trump in the first administration where the. What was it? The star. What's the Space Force? Space Force, Yeah. So he comes out with Space Force, and it's, like, mocked. And there's actually a show called Space Force that's like the Office. It's like a sitcom. And it originally was making fun of Space Force being, oh, it's so ridiculous to have a Space Force. But then the people who were writing the show were like, wait, like, this is actually kind of cool. And it kind of turned into, like, a lot.

My team has a Space Force.

Exactly, exactly. And so, yeah, who knows? But Golden Dome, very silly name, but probably some technology that we want and will probably affect a lot of defense tech companies that we know and love and have mentioned on the show.

We should do a deeper segment on what is the difference between a generalized ICBM sort of missile defense system and whatever the Golden Dome is going to be, other than branding, because it's fantastic. Branding. Branding matters.

Yep.

Naming matters.

Yep.

I'm sure that Trump would like for his legacy.

Yep.

To have at least a footnote about the Golden Dome.

Yep.

But, you know, again, going back to this idea of we've had sort of missile defense systems for decades, and. Yeah.

So there are generally levels to this stuff, like ICBMs. We talked about this. The hypersonic thing like ICBM, they go in a parabolic arc. It's pretty easy to intercept them and shoot them down with just another missile. The Patriot missile and the. And the Roadrunner that Andrew is building, you know, the whole idea is a cruise missile comes in. You need to hit it with a. With a drone or another, like a Patriot missile. Very expensive, very, you know, not a particularly tradable system. And with. With the, with the dawn of hypersonics, like, you know, you need to upgrade your dome. So I, I think the question is.

Like, new Dome alert.

Yeah. Dome is a binary term that sounds very good, but realistically, it's like this massive gradient of, like, our capabilities versus things that want to penetrate a dome. But, you know, and Trump goes on to say, and other place they have it, Israel has it, other places have it. And the United States should have it too. Right, Kim? Right. They should.

Now that we are changing trade wars with our neighbors, we should, you know, be ready for anything.

This transcript is so funny. I saw a bunch of transcripts. Like the top search result is like, is like Trump's transcript and fact checked. And like you can barely read what he's saying because, like, every single line is fact checked. And I mean, I mean, reason.

And we do the same thing for the New York Times as well. But, but it's one sentence. True. Some. Yeah, true song.

But. But I mean, just listen to this transcript. They should have it too. So I want to thank you.

But.

But it's a very, very important. This is a very dangerous world. We should have it. We want to be protected, and we're going to protect our citizens like never before. It's like it really comes across in the actual.

The voice.

Yeah, yeah, yeah. The voice comes across even in the transcript. Just a couple sentences. So the other tech thing that was mentioned, of course, was SoftBank, Oracle, OpenAI. Trump is tying this to his America first policies. He claims that we have had 1.7 trillion of new investment in America in just the past few weeks. That money hasn't been spent yet, but it's been the vibes of 1.7 trillion have happened.

That's actually an area where I would want to fact check. Not because I don't fully believe it, but I would just like to see it listed out.

So he goes on and he actually lists this down a little bit. He says the combination of the election and the economic policies that people love. SoftBank, one of the most brilliant anywhere in the world, announced a $200 billion investment. OpenAI and Oracle, I love how everybody's.

Collectively just agreed that they're sizing down the 500 billion.

No, no, no, no, no, no, no. That's the funny thing. So he says SoftBank is putting in 200 billion. And then he says OpenAI and Oracle. Larry Ellison announced a $500 billion investment, which they wouldn't have done if Kamala had won. And it's like, well, I'm pretty sure that OpenAI, Oracle, 500 billion includes the 200 billion for SoftBank. So you're kind of double counting that. And then he says Apple announced 500 billion. Said, I can't spend it fast enough. I got. It's got to be much higher than that.

And of course, that was contradict.

And of course that was like, you know, money that, that Apple was like, already spending to some degree. But Generally, like, this is good. Like we want, we want companies to invest in America. Like, this is great. And, and so it's funny because it's hyperbolic, but it's also like directionally correct. So like I'm, I'm kind of down. I believe they're building the best, their plants here instead of China. And just yesterday, Taiwan Semiconductor, the biggest in the world's most powerful in the world, has a tremendous amount, 97% of the market, announced 165 billion investment to build the most powerful chips on earth right here in the usa. And we're not giving them any money. Trump loves the deal.

Loves the deal.

And then he goes into the CHIPS.

Act, says your CHIPS act is a horrible, horrible thing. We give hundreds of billions of dollars and it doesn't mean a thing. They take our money and they don't spend it all the time. All that meant to them, we're giving them no money. All that was important to them was that they didn't want to pay the tariffs. So they came and they're building their building and many other companies are coming. We don't have to give them money. We just want to protect our businesses and our people. And they will come because they won't have to pay tariffs if they build it in America. So it's very amazing to get rid of the Chips act and whatever's left over, Mr. Speaker, you should use it to reduce debt or any other reason you want to just transcribe this iconic, iconic paragraph.

Yeah. And so that's pretty much, that's pretty much the entire coverage of tech and chips. CHIPS act, you know, obviously passed during the Biden administration, but was saber rattled for years. And, and you know, the people always make the argument that like, America is more divided than ever. But with regard to China, Trump and Biden had a lot of similar policies.

Yeah.

And you see 100 year marathon, Michael Pillsbury, that book really lays down, hey, China wants to be number one. They're not going to stop. It's not that they're going to necessarily just go to war, but they are putting a plan in place very slowly to build and build and build and become, you know, the number one. Everyone wants to be number one. Why not? And so Michael Pillsbury kind of lays this down. And there's also this other book we have over here, America Against America, that was written by a Chinese researcher in the government came over here and was like, I think America is going to destroy itself basically culturally. And like, we can kind of just like sit back and let it happen. Very, very weird. Like kind of blackpelling book. It's very odd, but at the same time there's a bunch of interesting things in it.

But, but every time we're about to destroy ourselves, we find a trillion dollars in minerals.

True.

It's true. Yeah. And so, and, and so, you know, Trump comes in, starts talking trade war, starts talking, you know, competition that actually carries forward into the Biden admin. And Biden says, like, yeah, like I hated. I'm very opposite of Trump aesthetically, and I'm opposite Trump on many things like immigration, but, but on China and on tack and on a lot of these things, the two administrations are pretty similar. And so the CHIPS act goes out and now Trump is thinking about renegotiating, which. Which is supported by Ben Thompson at Certecary. Because right now this CHIPS act is structured as incentives to build plants, not supply side, not demand side. Basically.

Ben should give free subscriptions to anybody in the White House.

Yes.

Seriously, you want it, you can have it.

Yeah. Because his, his analysis is going to be. Yeah, hopefully, hopefully informing folks. But you know, I don't really, when I tune on, when I turn on C Span, I'm not really watching for the politics. I'm not really watching. What are you watching for, John? I'm watching for the watches.

That's right.

And so that's why we want to talk to you about Bezel. You can shop over 22,000 luxury watches fully authenticated in house by Bezel's team of experts. It's if you love Trump, if you hate Trump. His watch available on Bezel. He has a Vacheron Constantin Historiques square dress watch. It's gold.

Fantastic.

It's a fantastic piece. But you know, Obama known to wear a Rolex. Pick one up on Bezel.

If you're a Obama guy, big rolly guy.

Yeah. So whatever your political persuasion, there is a watch for you that will signal who you, who you stand with on Bezel. And this is what you.

There are 83 separate auctions live right now. So if you're in the, you know, it's almost, it's fun.

I do think one of the, one of the most fun things about, like, watch hunting is finding these ones that have stories. Like, I'm pretty sure Obama has a Rolex that was given to him by a Navy SEAL who was on his security detail and saw him wearing one.

Watch and was like a Casio or something.

Yeah.

Like, you gotta, you gotta, you gotta take this, Mr. President. Like, you gotta step it up. Something like that. And I was just like, boys like.

These guys giving a watch you love to somebody is incredible highest.

And then, and then of course you can track it and you can see like, oh yeah. And then Obama wore it for like the next 25 years or something. Or like 10 years. And you're like, that's awesome. And so, and so I think when there's a, when there's a rich history about a watch like that, like some story like the Rolex Presidential day date, like the gold, the gold Rolex worn by several presidents. Sure. You're not getting one that's worn by a president unless you're going and finding one for like a million dollars.

Yeah.

But even just having a story that you can tell around a watch is, oh yeah, this is the same model that was worn by Reagan.

To be honest, I wouldn't be surprised if Trump had some type of little scheme at some point where he would just take your day date, wear it for like 30 seconds, take it off. You know, he'd do like a hundred and he'd be creating, you know, it could be the Trump edition Rolex. So he just puts it on. It's worn by Trump. You know, they stamp it on the back on. They go.

It's ridiculous. Anyway, let's move on to some tech news. Let's go to OpenAI. Sam Altman told investors that he might launch agents costing between 2,000 to 20,000amonth to do tasks like coding and PhD level research. Nick here says Lowell Lamau, even this was something. Didn't we predict this? Yeah, I think I wanted a 20k agent.

Yeah, we, we had talked about this months ago.

Yeah.

One of the things that I think we were more fixated on was if you just gave me the best model and let me have unlimited usage of it, that would be worth around $2,000 a month.

Sure.

In my opinion.

Sure.

What they're saying. And the pushback here has been for $20,000 a month, you could probably get four separate PhD level research assistants and get them all the 01 Pro subscription for the cost of. So this agent can't just be a PhD level researcher. It has to be vastly more effective otherwise. Because 20k it says 2 to 20k. Right. So we don't know exactly.

What's fascinating is like there are now PhD level researchers that are in the training loop in the RLHF because at a certain point, like you can't just hire me and you to generate training data for advanced mathematics.

On the humor side, on the humor.

Side, yeah, Call us, Sam. When you need a humor eval and you need us to generate thousands of jokes. Yeah, we can do it, but it's going to cost you a thousand times more per, per, per token. But, but it is interesting to wonder, you know, like, do you wind up putting those folks in the loop at some point and do you just say, hey, yeah, at 20k a month, like 20k a month for a software product, you usually get like an sdr, you get like, like some sort of implementation person. And so it wouldn't, it would be funny if like you're at 20k a month and like you also get like a human who you can just like act like a human. You can text and call and be like, hey, yeah, like I need the a, I need the AI agent to do this. And then, and then they're using the AI, but actually you're, you know, interacting with.

Yeah, I want to know how real this is. Right. There's just been this. OpenAI has been at the leading edge and they continue to impress. Even though many people said four or five was a letdown.

Yep.

I think many people just expected it. So it shouldn't have been a letdown. It should have just been sort of, like I said, expected. All that said, if Sam is out on the fundraising trail and he needs to be able to show how they're going to add $10 billion of ARR, that's not on their core consumer subscriptions. This is one way to do it. Right. When you're talking about two to $20,000 a month per sort of instance of the software, it's very easy to get these sort of big numbers very quickly. And so I don't think people with where 4.5 landed. I don't think people are in the way that Satya is talking. People are starting to adjust the sort of narrative around, oh, AGI is sort of three, four months away. You should invest now at this crazy valuation because we're going to control super intelligence. There needs to be a new narrative from a fundraising standpoint, in my opinion, you know, to continue sort of commanding these really, really, really aggressive forward revenue multiples.

And it's test time inference.

Yeah.

Like, it is like when I hear 20k, I hear like I'm signing up for that if I have a workload that, that needs to be done every single day. I'm not exactly sure what a great example would be, but I mean, it's like, you know, run through every, I'm an insurance company and look through every single policy, every single day for anything that's out of the ordinary. And you're dumping in millions of pages of text again and again and, and every time you query, you want it to actually sit there and think for hours. Something like that. And so the 20K, I would imagine that the cogs on that are thousands of dollars.

Just every day you're gonna say, write me a stand up joke in the style of Shane Gillis and feel free to take 12 hours, but you better come back with something good. Otherwise I'm unsubscribing.

Yeah, I was thinking about the jokes that people like. It's always like the public figures who say, like, be me, be Andrej Karpathy. And then it's like, oh, well, it knows everything about Andrej Karpathy so it can tell a good joke. I wonder when normies will be laughing at it because it needs to know stuff about them to actually deliver a good joke. Because a lot of it's be me.

Love DraftKings, love Netflix, love parlays.

Yeah, yeah, I don't know. Anyway, you had the question of like, you know, how real is this? Let's actually read the information to see how much info they have. Basically, OpenAI is now at a 4 billion ARR. Which is incredible.

Gerstner came out and said they're projecting to be at somewhere near 11 by the end of the year. So the growth is still ridiculous.

Yep, yep, yep. The term agents typically refers to AIs that can take actions on behalf of customers without needing a lot of direction. While OpenAI investor SoftBank has committed to spend 3 billion on agents from OpenAI this year alone.

Yeah, yeah, yeah. Wait, like, so, so you saw Sam Lesson had that billboard on the 101.

Yeah.

That said AI is not your moat.

Yeah.

He came out and he said, I think it was basically on this digital billboard, which hopefully they got through ad quick, otherwise they probably, you know, wasted money and did poor targeting and all that stuff. But he was saying that, you know, $3 billion of their sort of projected revenue is from SoftBank, which SoftBank is using those projections to lever up, raise tens of billions of dollars of debt to then invest in OpenAI. So it feels, it feels hard to believe that that SoftBank will actually spend $3 billion on OpenAI agents this year. Just because I imagine that. Yeah, that's. Even if, even if the technology is there, imagine rolling that out across your entire.

I mean, $3 billion, that's got to be on. That's like in the same ballpark as Snapchat's. Cloud bill with Google, like, it's like. Or Netflix's infrastructure on aws, like they're probably higher than that, but like in the billions, like to spend billions with a single, like, you know, software, like, you know, tech company contract as a customer, like.

Yeah. So that's just to put it in perspective, I just ran a few very basic calculations that would be them spending $250,250 million per month this year, which sounds crazy, but it's, it's, it's like $8 million a day on agents.

Yeah.

And I think that of course OpenAI would let them spend that much, but SoftBank is also running a business. Right. And they need to be, there needs to be some real value exchange. Right. And I think if you just turn on a bunch of these agents.

Yeah.

And you're spending $8 million a day on them. Very unclear if they. Yeah.

I mean, what's like the craziest steel man you could possibly think for this? Like, what if you, what if you just say, okay, SoftBank has a ton of employees doing a lot of different things for this year. We're going to spend this massive investment, but we're going to effectively have every Single employee at SoftBank is going to have an AI agent shadow them. So every task, every email, every slack message that goes To a human SoftBank employee is also going to go to a, like a mirror world metaverse representation of that employee that's run by an OpenAI agent. And so you have information on, hey, SoftBank, you know, had customer service people. Let's see what the response from the human was. Let's see what the response was from the AI agent that was shadowing them every single day. And then do the same thing for the analysts. Do the same thing for the cfo. Do the same thing for the, you know, Masayoshi.

Yeah.

Do, do it for everyone. And then you can, and then you can see, okay, this is actually a replaceable job or this is a job that, that we got twice as much done and so there's actually twice as much value. So we should expand, keep the workforce. But yeah, it was a good deal. I don't know, it seems, it seems really hard to, to, to manage that much AI agency. You know, it's like, it's a lot of, it's a lot of intelligence to just go say, hey, hey, go. Do you know what? $250 million of work a month. I have trouble. I have the O1 Pro. I have trouble, you know, dispatching enough valuable queries. On a regular basis to really, like, get the most out of it. And I talk to people and sometimes people will be like, oh, you only use chatbots like five times a day. I'm using it 20 times a day. And it's like, I don't know if that's like. I don't know that you necessarily want to, like, be maximizing your AI usage. And that's like the right metric. There might be something where it's like, work smarter, not harder. But if I had a 250 million dollar. It's like Brewster's Millions. Have you seen Brewster's millions? The kid inherits, you know, a billion dollars from his, like, you know, rich grandfather or something and says, you know, in order to deal with this sudden wealth, you have to spend, you know, $10 million this. This weekend or something. And if you can spend it, it's all this analogy of like, you know, you catch the kid smoking a smoking cigarette, you make them smoke an entire pack and they're like, this is disgusting. They never smoke again.

Right.

So Brewster's millions, the same idea, but with money. And if you Brewsters millions to me on AI agents and you gave me a $250 million budget and said, go and deploy this and try and get $250 million worth of value out of this. I don't know, it's weird.

It's this weird house of cards, right? And when you see it's completely unbelievable that SoftBank would effectively spend $3 billion a year on agents, would guess that their actual spend on agents in the past 12 months was under $100 million. Right?

Yeah.

I mean, so. So. So that level of growth is not super believable.

Yeah.

It is almost believable that Masa would say, I will actually spend this money with you because SoftBank is going to be such a big investor in OpenAI that, you know, you know, we're. We're willing to make this commitment and it's sort of like this crazy, conflicted transaction that has a bunch of, you know, Sam is out there using this revenue commitment to raise more money from other people. OpenAI is using that, you know, open A. Or, sorry, SoftBank is using OpenAI's revenue projections to raise debt against. And the whole thing is just beautifully conflicted and.

But no conflict, no interest.

No conflict, no interest.

So let's give a little bit more context than we.

But we'll be able to actually. So the beautiful thing is we will be able to see pretty much. Not exactly necessarily, but we'll be able to see if this spending spree from Masa on OpenAI was real. Because SoftBank's public company.

Yeah.

So we'll be able to check back on this.

So OpenAI is going to have several tiers. $2,000 a month gets you high income knowledge workers. Mid tier agents for software development can go up to 10k a month. And high end agents acting as PhD level research agents could cost up to 20k per month. In the long run, OpenAI expects 20 to 25% of the company's revenue to come from agent products. The person said, our past reporting has given hints about each of these type of agents.

And so, so here's the thing that's somewhat, you know, just one way to push back here.

Yeah.

Devon AI, which is built on top of OpenAI.

Yep.

Costs $500 a month.

Yeah.

There's been reports that it's super effective in organizations.

Yep. Well, it's 500amonth. And then there's also consumption based pricing.

Right.

It's the same thing with Claude code.

Right. But all that being said, is OpenAI's products going to be 20 or 10 times better than Devon, which is built on top of OpenAI.

Yeah.

Otherwise why would consumers use the OpenAI version that costs $10,000? Yeah.

I don't know.

It'll be interesting. Yeah. There's so many. There's such a dance here between what are they going to release in the API, how closely are they going to keep things held as secrets. There's actually another story we're going to get to about Ilya Sutskever and how he's being really, really secretive. And it's interesting because, yeah, I mean, if you're vending your best model, your model that costs, you know, $20,000 a month out on API, someone can wrap that. Maybe you don't want them to do that. Somebody can maybe distill that or reverse engineer it. And so there will continue to be the question of like House, how much can you lock these down? Or how long does the frontier last? It's an ongoing debate and it's fun to follow along. But of course there are, you know, endless battles over OpenAI. And of course there is the news that Elon Musk and OpenAI are in lawsuit. And there's a breakdown here from Rob. Should we go through this?

Yeah, we can go through this briefly. I'll cover it. So people are sleeping. This is this morning at 3am he says people are sleeping on huge news in the Musk vs OpenAI case today the judge finds that If Musk's donation gives him legal standing, she thinks it's very likely she'd want to block the their entire $100 billion nonprofit to for profit conversion. And there's a highlighted piece of a legal document here that says next. And similarly, if a trust was created, the balance of equities would certainly tip towards plaintiffs in the context of a breach. As Altman and Brockman made foundational commitments forswearing any intent to use OpenAI as a vehicle to enrich themselves, the court finds no inequity in an injunction that seeks to preserve the status quo of OpenAI's corporate form as long as the process proceeds in an expedited manner. So Rob says Musk is trying to stop the opening OpenAI business, effectively converting from a nonprofit to a for profit. To do that, he needs to prove that he is being wronged in a way that allows him to bring a case to the court, which is a legal standing, and the conversion to a for profit which gets cut off here. I'm going to try to see if I can find it briefly.

Yeah, this is, this is a very high bar to reach. Interesting. I mean, I'm still on like team, like, you know, hopefully they can sort this out. It does, it does make sense that it seems like such like a. Just an aberration of the legal system that we're in this situation where it's like, hey, there's this weird situation where we kind of have to switch from a nonprofit to a for profit. Can we just have a one time grace period where anyone who donated to the nonprofit just gets a stake in the for profit? Because that would be. That does feel like the most fair. But it feels like legally it can't happen or else there will be some crazy tax implication or it will get like blocked.

Yeah. But I actually messaged Rob about coming on the show. He responded back and he says you should have a lawyer on the show, not me. This is a super complicated issue. But overall my takeaway from the news today was that the judge is not shutting down sort of Elon's efforts. It seems like still very 50, 50 or maybe even 60 40. Altman OpenAI versus the Team Musk. So we'll see what ends up happening. But overall this is pretty wild to just be kind of playing out in front of everyone while OpenAI stays charging ahead. While XAI stays charging ahead. And yeah, it certainly is evolving.

Yeah, I wonder even like what the time we're going to get ASI before we get a legal conclusion on this case. Aren't we? Like this. This is going to drag out for years. I mean, the Musk pay package is still dragging out. Right. Like, we still don't have a conclusion on that because he's moving. He's moving Tesla from Delaware to Texas, I believe, and, like, reinstating it. And then the judge blocked it. And so the. The. The legal world just moves so slow. Yeah, it's rough. Anyway.

Well, there's one more line in here that I think is relevant. It says for this inquiry, the Court looks at OpenAI's certificates of incorporation in both Delaware and California, which lay out the charitable purposes behind the corporation. The documents state that OpenAI quote, was organized exclusively for charitable and educational purposes and not organized for the private gain of any person. And so this is basically what Elon's legal team is leaning on in terms of making their point heard. So see how it goes.

Well, whether you're Elon Musk or Sam Altman, you're not going to be building AGI or ASI unless you're getting a good night's sleep. So both of you, please go to.

8Sleep.Com tell them the technology.

Tell them the technology brother sent you, and turn your bed into the ultimate sleeping experience. Sam, Elon, this is more important. Surely you guys can agree on this. You guys can disagree on who owns what and what the companies are, and what nonprofit for profit. All that legal mumbo jump mumbo jumbo. We can put that aside and we can agree that we all just want to get 100% sleep scores. And that's actually what's really.

I did not get that last night.

I did not either. I got an 84. What'd you get? 96 roast.

I'm back. I'm back.

I've been having a rough time.

Two hours of deep sleep.

Two hours. I got 79 on quality and 644 and time slept.

You're sick as a dog, but your energy is way back.

It is. I'm pumping it up.

I'm pumping it up. I. John was asleep at the wheel yesterday.

I think all the drugs I'm taking, mostly just Robitussin. It's acting like I'm high. Like, that's great.

That's fun.

I'm definitely letting it loosen. I think the answer is, like, if I'm low energy, just, like, let it fly.

Yeah.

Because it makes for a better show, for sure.

Absolutely. All right. Eatsleep.com/cvp and go check it out. Yeah. Back to the show.

Back to the show. Let's Talk about Ilya Sutskever. Absolute goat raising up billions of dollars. I mean, we've talked about that before. Apparently it's impossible if you're an open a co founder to build anything but a foundation model company.

Yep.

We would love to see someone start, I don't know, supplements company. Yeah, I want to see, you know, him build, you know, I don't know.

Or just economic. I'm going to go into bodybuilding. Yeah, go golden retriever Max.

Start a podcast. Do something. I don't know, it's just like, it's like all they can do is just foundation model, foundation model, foundation model. Every single day, every single machine.

God is real.

Yeah. And what's weird is they don't want to do it together.

Never.

They all want their own companies, they all want their own foundation models. But you know, apparently it's profitable. The company's already worth 30 billion. Raising multiple billion.

Is that the fastest?

It's gotta be. So anyway, not to say the value.

Has already been created, but Ilya Sitzkever, has a company ever gone from 0 to 30 in less than a year?

I don't know.

So Ilyas itskever, He was an OpenAI researcher. He's the one who found the transformer paper and you know, Google built it. But he really recognized that it was very important. He, you know, built the early GPT architecture is very foundational in the development of LLMs and AI and he left OpenAI last year. Now he has a new startup called Safe Super Intelligence and it's become one of the most valuable companies in tech thanks to his reputation. And SSI says it doesn't plan to release any products until it develops super intelligence. And I was thinking about this. So they raised like $1 billion or 2 billion at, at a 30 billion or they're in talks to raise funding at 30 billion. What would be hilarious if Deep Sea comes out, they're raising like 2 billion.

Journal just front running this and saying his startup is already worth 30 billion. Because it sounds more interesting.

Yes, yes, definitely.

Okay.

It would be so funny if it was like, you're okay. Hey, we need to train this foundation model guys. Like we need like 3 billion. We'll just give you like 10%. So we'll do like 30 billion valuation. But it makes sense because like it's going to cost us 3 billion to build a frontier model. DHC comes out and they're like, we actually only need like 300k to do this. So we're still going to do the 3 billion but it's just going to be secondary and don't worry about that and. But we're going to keep the valuation where it is because it's like the same tech. That would be genius.

He's just in it for the cheeky secondary maybe.

I mean there is a, there are group of. Groups of people who are basically like, you know, the, my ASI AGI world is just like I need to just corner some resource because like, like yeah, there will be abundance but there will only be so much beachfront property. And so like I must get my beach. But of course, like this isn't what Ilya is doing. He's. He's like obviously a true believer in AI and tech and also fantastically wealthy from OpenAI. I forget what they get. They get credits or.

Yeah, I would have stock. How much would you pay to get the full hour long pitch for ssi? Because arguably just getting, you know, I'm sure he's pitching relatively few people for this, right? He's not trying to go talk all over the town. Yeah, a lot of people are conflicted already. He probably is talking to Sovereign.

Isn't this a. Nat Friedman was the first group I thought.

And it's Gross.

Daniel Gross.

Daniel Gross is the CEO, right? Yeah. So, so, but, but getting the pitch from Ilya would be almost priceless.

Totally.

I would lover up TVPN to just get the full pitch and then be able to act according to, you know, who knows? He seems like somebody who would give a very authentic pitch versus just, you know, promising, you know. You know, he's sick of the sort of like high flying deal making approach of his former co founder.

Totally. Yeah, I mean he seems, he has that, that quote and I was like a Guardian piece or something or SF Chronicle video where he's like if, if, if the AGI doesn't go well, the entire earth will be covered in solar panels. Have you heard about this half Earth concept? Half Earth. Ha. F. F Now I think his name's Peter Half. He's a scientist.

He named it after himself.

Yeah, it's a coinage. It's like great. No, this is like a real thing. So, so he's like a, he's a, like a futurist, a scientist. And he has this theory, there's like this four quadrant model for like how, how, how the earth can develop and accumulate more and more energy over time. Like eventually like we're growing our energy consumption at like 2% a year and if you just play that out exponentially, eventually we will be consuming all of the energy that is, that hits the Earth from the sun. And then eventually you go into Dyson Sphere. That's like Kardashev too. But there's a question about, like, do you go black or green? Green is like, we focus on, like, bio and, and you have like, you know, more trees that are like, you know, absorbing energy. Essentially. Black is like solar panels. And then the question of, like, do you stay on Earth or do you go into space? There's like this, there's like this interesting four quadrant model, but half Earth is basically like, yeah, just, just black out the sky. Put. Put solar panels everywhere. And then we'll all live in pods and most of the humans will die and we'll basically just be AIs. And that'll be like our progeny. And like, maybe that's good for some people. Some people like it. Some people are into that. Not me. They've never been part of that whole world.

Never been big into that.

Never been big into that whole, that whole Matrix world. But he's, like, worried about this, like, genuinely, like, we're kind of joking, like, dancing around it. But Ilya definitely is, like, if we mess this up, we're just going to be continually optimizing, like, more computers, more computers, more computers. And then you're going to have solar panels everywhere and robots taking over.

Okay, here's a question, and I don't mean this in an offensive way at all.

Yep.

I have many family members that struggle with hair loss.

Knew you were there.

Is it possible that Ilya doesn't just shave his head because he wants to reach AGI before his hairline reaches the back of his head?

It is a funny choice, but I think it's just because he's, like, focused on other stuff. But, yeah, I mean, put on.

If you're looking at the mirror every day and you've got billions of dollars to deploy, and your arch nemesis Sam Altman is racing to achieve AGI, and you look in the mirror and you say, look, I maybe have. I maybe have like, 18 hours. I maybe have 18 months before to, you know, of a hairline. You know, I gotta grind harder.

This is disrespectful. I won't. I won't abide by this. I think he looks great.

No, he looks. He's an absolute Chad.

He just, he just cares so much about the AI.

Yeah, No, I think that's the right approach.

Yeah.

And he wears a suit. An absolute dog.

True.

An absolute dog.

Yeah. No T shirts over here.

No, I love it. I'm excited to see this round get confirmed and I do think this will be a first of speed running 0 to 30 billion in less than a year.

So check this out. He's running this place. It's a super tight ship. He's very, very worried about security. And so most AI startups work hard to get attention, hoping it will attract employees and investors in a highly competitive space. SSI operates as secretly as it can out of its offices in Silicon Valley and Tel Aviv. Its bare bones website contains little more than a 223 word mission statement. Candidates who secure an in person interview are instructed to leave their phone in a Faraday cage, a container that blocks cellular and wi fi signals, before entering SSI's office. One of the knowledgeable people said most of its staffers aren't well known in Silicon Valley because the company is looking for promising technologists whom Sutskever can mentor rather than experienced people likely to jump between employers and take what they have learned with them.

I love that approach.

Still, Silicon Valley's top investors, including Sequoia and Andreessen, have poured money into the company. The latest financing, which marks a six fold increase from its 5 billion valuation, is being led by Green Oaks Capital. SSI is in talks to raise funding at 30 billion, but doesn't include the new cash it's collecting, which were previously reported by Bloomberg.

Were Sequoia and Andreessen in the same round? I guess close. You don't see that much.

This also feels like super conflicted with. Aren't they both in XAI too? And they're doing the exact same thing.

Like hey, there's no conflicts in the foundation model world. Everybody's in everything.

I think it's, yeah, it's commoditizing, it's drilling, you're drilling oil and like yeah, I want a piece of that well. I want a piece of that well. Like let me get it all, I guess I don't know, it just seems like, like there would be some discussion of like hey, like can you not invest in a direct competitor? But you know, maybe it's dramatically different. Like when you think about the Pro, just the product pacing. Like Grok 3 is very much like productizing so fast. Like Grok 1 was immediately out and then, and then embedded in Twitter and you can. And we talked about like maybe it gets embedded into Teslas and it becomes the way you, you talk to your car. And it seems like Elon really, really wants to like be product focused. Like there's the voice mode and like all the different features yeah.

And we've talked about this. That what he sees in OpenAI is not benchmarks.

Yep.

It is consumer adoption. Right. OpenAI will probably reach a billion, you know, a billion users at some point this year. Somewhere close to that.

Yep.

I think even potentially they could be in the territory of a billion like monthly actives.

Yep.

Which is so, so difficult to reach.

Yep.

And Elon, even if Elon got every person on X to download the Grok Grok app, he wouldn't even come close to that. Right. There needs to be sort of organic adoption.

Don't sell them short. I do think Elon could potentially unseat Jimmy John's for number.

That's true.

Unseating OpenAI for number one is a tall order. But never bet against Elon and just never bet. Never bet for.

It's hard to bet against Jimmy John's new hot honey sub. But if anyone could take them down.

It is so funny that the number two app in the app store and.

They hit number one.

I think they hit number one.

They went number one.

They went number one dot Jimmy John's for some reason is the number one app in the App Store amid all this, like, you know, billion dollar, you know, lawsuits and fights over. Well, this is why Ilya staying out, he's like, I'm not going to release an app until it can just immediately be better than Jimmy Jones.

Yeah.

I'm not even going to try and go consumer.

I'm wrong. They hit number one in food and drinks.

Okay.

Is still on top.

Still undefeated.

And which is why, it's just why investors are willing to pay over $300 billion.

Well now, now, you know, normally people would be like, oh, if you have Jimmy John's on the show, no longer a tech show, but now there's a tech angle. So let's get them on. Yeah, let's get the founder on. Jimmy Johns.

Let's do it. Jimmy himself.

Anyway, we got, we got five minutes until our first guest is joining. Let's close this out. You know, everyone's interested in following all the different co founders around the world as they build ever increasingly large foundation models. The interesting thing is like, you know, we, we heard about. What was it? Test time, compute, reinforcement, learning. Deep, deep research. All these different tools that were, they clearly came out of Ilia projects early on at OpenAI. I forget what they were originally called. Q Star. Remember Q Star? And it was like, it was. This was leaked that Ilya was working. What did Ilya see? Right? And so clearly he has some ability to put Together kind of new algorithmic paradigms. And even though he was. People get so obsessed with like, is deep learning hitting a wall? Are you, are you scaling pilled or not? Whereas Ilya was like, he was both scaling pilled and went super long the transformer and deep learning architectures, but at the same time was working on reinforcement learning approaches on top of, on top of like the big massive scaled LLMs to bring those two together. And I think they were kind of, I guess they were kind of obvious. I was watching a video earlier from two minute papers from five years ago from OpenAI and it's talking about this, this, this game that they did where they had two AIs on each team. So it's 2v2 game. Hide and seek.

Yeah.

And, and, and they, and they play and over like, you know, a million rounds they simulate and like they just can't even find each other. And then they get into like 10 million rounds and then they start learning these strategies of like, well, if I go in this, in this box and then I put like a cardboard box here and I lock it in place, then I'll start winning. And then, and then the other AI figures out that if they take this like ramp, they can surf up over it and eventually they start like figuring out how to like break the physics engine of the game and do all this cool st. And, and that reinforcement learning obviously played into when OpenAI played Defense of the Ancients. Dota 2. Do you remember this at all? They did like a 5v5, basically like Starcraft. I mean they also did Starcraft, literally. Yeah, they did. They started playing video games.

Was the thing that would actually go viral.

Yeah.

And also live streaming it.

Yeah, yeah, yeah, yeah. And so those types of. Well, the Pokemon thing's interesting because it's not like they didn't train on Pokemon.

Yeah. It's just Claude Also running a 247 experiment is also very powerful and that you're putting it all out there. Yeah, it's verifiable.

Yeah, yeah. So anyway, I don't know Ilya. I hope he comes up with something cool. It's. It does seem like he's very aligned with safety.

Oh, to me, this is my boat. This is my Ilya is my bet on what Keith was getting at when we had him on the other day where he said what comes after the LLM.

Yeah.

Seems like every other former OpenAI early team member co founders working on language models.

Yep.

It's very possible. Ilya is working on whatever comes after that.

Sure.

So very Excited to see that play out.

And yeah, like as the, as the AI brain develops, you have the, the memory core, the reasoning core, like Ilia.

And he set the expectations so incredibly high, saying, I'm not going to release a product until I presumably have something that's significantly better than whatever else is on the market.

Right, Yeah. I wonder how he would even release that. Like what, what would the product instantiate itself as?

Like, we need, we need. It's just another app. There needs to be the information. Needs to send the tech paparazzi, you know, outside of the Tel Aviv office, snapping the picture of Ilya walking out, you know, all secretive. That's the content. I want the information.

Great. Well, you know what he has to do when he launches this company?

Billboard. Billboard.

Because he's gonna have. No one's gonna know about this. Yeah, it's one thing if you're in YC and you launch immediately and you grow organically and you're going direct and you're posting, but.

And you go viral for your.

He's gonna have this fantastic product, but he's gonna. No one's gonna know about it. So Ilya's gonna have to buy some billboards. And where does he do it? On Ad Quick. So Ilya, I know you're listening. Go to adquick.com, start working with the team now also, I mean, honestly, Ilya, you could do some recruiting ads on there, throw some billboards up Waterloo, Tel Aviv, Silicon Valley. There you go, the billboards up ssi, we're hiring, but go through Ad Quick because they make out of home advertising easy and measurable. Ilya, I'm talking to you directly. They you can say goodbye to the headaches of out of home advertising. Ilya Only Ad Quick combines technology, out of home expertise and data to enable efficient, seamless ad buying across the globe. So Ilya, go to adquick.com tell them the technology brother sent you and good luck with your new venture.

And so we're going to have Jeremy come on the show right now. And you know, our main advice for Jeremy was to start buying one on one billboards. Start buying one on one billboards saying you got a special situation, raise too much money, call Mr. Griffon.

What's up, guys?

How you doing?

Good, how are you?

Have you bought a billboard yet?

I'm working on it, I'm working on it. I think all the management fees need to go directly into billboards. That's what I've learned from the most profitable podcast.

That's great. Break it down for us. What would go on the billboard? What's your business? How do you pitch it to people these days?

My business is picking up the roadkill of the venture industry. The startups, the founders that have triumphantly built heroic and wonderful businesses and have gained millions and tens of millions of revenue and built incredible teams and have a wonderful product and customers that love them, but for the fact that, that they had too much capital forced upon them and it has buried them under an untenable prep stack and made the whole business not work and left them with no choice but to kind of get stuck in no man's land. And my mission is to help those founders get back in control of their businesses and run them as fully profitable, cash flowing, great businesses, as God intended.

Love it.

Incredible line. I would love to get a sense. We spend a lot of time covering the news and the news tends to fixate on these sort of colossal failures and these sort of moments of glory.

The humane AI pin, for example.

Yeah, yeah, exactly. You're seeing the behind the scenes, which is companies that raised a lot of money maybe three to five years ago that are now in a tough spot. You know, the venture world has sort of turned sour on them. What is the state of that market right now? How busy are you? I know you spent a lot of the last year raising your first independent fund. Now you're investing. What, what's kind of happening on the ground?

Yeah, I mean, look, it's, there's a lot of these, a lot of businesses, there's a lot of great businesses out there and a lot of like stuck founders and, and there's more and more of them and increasingly a lot more VCs who realize that there is a severe lack of DPI, especially in the middle of the portfolio right now. And so you're starting to see a lot of great companies that realize they need probably a different owner, an owner that's more in line with actually their goal and their vision. And then a lot of funds that are realizing, you know, gross IRR is all well and good, but eventually, you know, you can only eat the dvi and that's what the capital allocators are starting to demand.

Have you seen founders that get in that position? You know, they have the opportunity, you know they're talking to you or they get into position. Are they ever trying to sort of pivot their business into a better AI narrative and then go back to market and is that working or are they trying that and giving you a call back and saying yeah, you were right let's try to figure something out.

Well, yeah, so no, what they're doing is, what they're realizing is that they can pivot to AI and they don't need any more capital.

Interesting.

They can, you know, especially if the business was built, you know, four or five years ago, they can replace a lot of their costs with AI. I think, like, every organization should be thinking hard about whether or not they ought to cap headcount and just keep headcount at or below where it is and increase kind of headcount, so to speak, through AI. And I think, you know, I think we're about to see kind of the renaissance of the bootstrapper. And so I think a lot of founders are starting to think, like, you know, maybe I can grow my business 10x without more capital because. Because of these tools. And so, you know, yeah, certainly you can. You can slap an AI label on and go raise another round. You know, you certainly have the founder who's in the. Just. Just two more million bucks, bro. Like, just give me.

Give me two more million.

Like, I got this. But I think that's a mistake. I think, I think we're about to see, like, an era of incredibly, like, capital efficient businesses that, by the way, like, are not giving up growth at all. Like, you can still grow 100 a year, and if anything, you can do with a lot less capital. You can own more of the business. You can, like, you know, be a lot more in the driver's driver's seat and really kind of control your destiny versus kind of lever yourself up? As much as I know that, you know, you guys are always in favor of the max leverage option.

Exactly. Can you walk me through Anatomy of a Deal and how kind of the. The pref stack waterfalls out on one of these? What are the different kind of tension points for some pre seed investor who feels like they're sitting on a 10x or 100x and then someone who got in late and clearly kind of knows they overpaid and just wants to get something back. Like, how do you get all the parties around the T and then think about incentivizing the team going forward?

Look, I think one thing founders don't understand or maybe under appreciate is that your cap table ultimately really affects your business in a wide variety of ways that can be very subtle. And so at some point, maybe you go from having a venture fund as an ideal owner to being a neutral owner. I think a lot of funds are in that stage now where, you know, it's not detrimental to be owned by a venture back to be owned by a venture fund. But maybe it's not super additive. And then eventually you get to a point where it can start to become, you know, it can, it can detract because just the incentives really diverge. And so you know, what I'm really looking to do is like what I implore everyone to think about especially if they're not in a business that's like on a clear trajectory to raise, you know, need to burn hundreds of millions in capex for some kind of like nine or ten figure outcomes is think a lot about like who are my owners, what are their incentives? And like you know, when you get to the board meeting, is it actually, is there friction or are they actually behind me? And I think that's a lot of it and a lot of it could just comes down to like what is the cost of capital of your owners to get a little bit boring. Like the, the cost of capital of, of the people that own you makes a really big difference for like what they expect and what they demand of your business. And so you know, it varies like if you're an individual, if you're a pre seed or seed, you're happy to stay in the business. There's no reason for you to sell. You can roll over if you're like a growth bun that maybe did a somewhat regrettable, it's a little bit wake up from the night after the party, look around, there's beer bottles everywhere. Would prefer to just pretend that it didn't happen. Maybe you want to get your capital back. I think oftentimes the best outcome is just returning capital which used to be a norm and venture. And because of this bifurcation between huge outcomes and then very little liquidity in the rest of the portfolio is not happening as much anymore.

Every venture capitalist today says they're in the people business. Right. If you ask them to have any type of specific thesis, they just say we're in the people business.

That's founders.

Yeah. So that's generally the right take. That's senators point of view. Are you more in the people business or the value business? Because from my lens, you know, you're looking for value and in some cases you're more than happy to let the founder move on to their next chapter and bring in a professional CEO. Is that correct?

Yeah, I mean I think I'm also entirely in the people business. But it's you know, is the. I think everyone should be, you know, doing their highest and best use like following their calling sometimes for A founder that's. That's, you know, getting more control, more. More sovereignty over their business and really kind of being in it for the long haul, having way more optionality in terms of, like, financing. Sometimes a founder, you know, a thing that happens far too much that, that no one talks about is all these founders who are stuck. You know, if you have 20 million of revenue, 100 employees, and you've raised 100 million bucks, you can't just quit if you're the founder, even if it's clear that, like, what you were trying to do is build the next Decacorn or whatever. And so there's a lot of founders out there who know. And what's funny is the board knows and the founders knows, but they're, like, not willing to have that conversation. And so, you know, there's a lot of founders who are stuck. And for those who are stuck and would like to leave but don't want to abandon the building the business, I think, you know, there's more options than you think. And there's someone out there to get back to the people thing. There's someone out there who is the perfect person to come in and run a business like that. Because, you know, a lot of times the person who takes a business from 20 of revenue to 100 of revenue can be a very. You're looking for a very different type of person than someone who's going to start it from zero and grow it into a huge business. And so it's really just about. About like, is this person doing the thing that they're really fired up about? And if they're not, then like, do you really want to be in that seat?

Are there any common character traits or kind of either even background, career paths for that type of operator who's maybe not the founder, but the type of person that can, with pretty high confidence, take a business from 20 million to 100 million in your experience?

Yeah, I mean, I think, like, you know, my friend Andrew Wilkinson has written a lot about this. I think he really kind of has it right. You know, there's a lot of C suite and senior executives at a business that's two or three times bigger than maybe the business you're looking at. And those are people who are, like, highly capable, has at least some entrepreneurial spirit, but maybe they're not, you know, whether by constitution or circumstance, not interested in taking the full risk of a founder, but still want to be in the driver's seat, want to, you know, want to run a business. And there's actually A lot of talent there that can be unlocked. Who are people who, you know, can't do the go through YC thing, but still, you know, would love to have real upside, would love to, to lead a team and run a business and yeah, I think there's actually a lot of really untapped talent there. And it's kind of an underrated option. It's kind of like how it used to be. Like buying a business and running it if you're like an MBA or something was like an underrated path to quote unquote. Entrepreneurship. I think coming in and running a decently sized startup or business is kind of one of those now that not a lot of people consider. It's kind of like very. You're either a founder or you're just an exec or maybe like a Fortune 500 CEO or something. But there's a lot of middle ground and I think, yeah, I think there's.

A ton of untapped talent question going off of that. Search funds got very popular simultaneously it seems like venture, the world of venture capital discovered private equity and roll ups. And you know, we've seen a number of companies come out the gates and decide they're going to basically raise venture dollars to pursue a PE style strategy. That concerns me given that private equity is pretty good at what they do. Right. They've sort of gotten better and better and better over the decades around finding fairly priced businesses and making them more efficient and growing them and selling them. What do you think happens to these sort of software or AI enabled roll ups and are you salivating over the potential of buying them down the road once they.

Look? I think, I think raising venture to acquire businesses is a great gig right now, you know.

Yeah. So, so and here's, here's why.

Jeremy's like the grim reaper. Like if you're, if he's excited about something.

No, and here's why. Right. You can, if typically in private equity, if you want to go, if you want to go buy businesses, you raise a fund, you get 2 and 20, which is a great business model. You've talked about this before, Jeremy, sort of the perfect business model. But right now in venture you can go raise $20 million at 100, dilute yourself 20%. And so you actually have much better economics if you're going, you know, buying up these businesses and eventually I think that's going to stop. Right?

Yeah. And by the way, there's a hurdle in private equity and there's no hurdle in venture capital, which makes the whole thing even. Sweeter.

Yeah.

But you know, I don't think I'm a grim reaper. I think I'm a liberator. I think all these founders are trapped.

If you envision that you're gonna be doing a lot of business in the category in 10 years, it's probably a bear signal, right?

Yeah, I think it's more seeing exuberance. You know, once the, once the NBA's pile in, it might be over. I love him, but look, if you build a business to 30 million bucks of revenue, that's like a heroic feat. It's extremely hard to do. And then you somehow end up with 8% of it under a $200 million prep stack. That kind of sucks. And I don't think that's the outcome that anyone is trying to have happen. But it happens as a kind of a spandrel of the venture industry.

Yeah. The grim reaper is the bankruptcy attorney. You are the winged angel.

Yeah. And you know, if you really want to sell your startup for parts, there are people out there who can do that with brutal efficiency. And they're great guys. But like I'm a fund. I want to sell these businesses for more money in the future, which presumably means they get more valuable. You know, I'm not a. I don't want to just like, you know, make Toys R Us go bankrupt. That's, that's someone else's job.

But I mean, is, is over the long term, is your business correlated with like frothiness and volatility? Like in a post.

Froth.

Post highly, you know, high, high exuberance. Like that is a good signal for you, right?

Yeah.

There will be more opportunities for sure.

I mean I'm, I'm, I'm interested in dislocations and dislocations certainly happen more in bubbles or exuberance.

Yeah.

But I also feel like it's happening more frequently too.

Like it seems like it's a good model. It doesn't seem like, it doesn't seem like your fund is like a one time, like, oh, they just like unique case right now. It's like no VC is built to do these types of things. There needs to be a system to help process the ones that don't go perfectly. And that's where you can come in.

Yeah, well, look, I mean venture is 10x in terms of dollars deployed in the last 10 years. It's a completely different thing now. Just as the biggest firms are turning into PE shops, the biggest venture firms are functionally PE shops. There needs to be. There's basically one product in Silicon Valley and it's like primary growth capital. And once these markets become bigger, more liquid, more efficient, there needs to be different kind of different. Just as there is in Wall street or on the east coast, there needs to be different products offered to founders, like equity at different cost of capital and things like that. Again, a really basic fact I think people should understand is if your model is to have a lot of zeros in your portfolio, your cost of capital, that is the minimum return that you need is just really high, which creates all these knock on effects. I really think founders would like just take a crash course on venture. It would really help them a lot. Like a lot of really exceptional founders kind of get into bad situations that they probably, you know, probably like 30 minutes with like ChatGPT could, could really kind of help them avoid.

And like maybe you should make a wrapper. I honestly think you should make a wrapper and just like create the sort of product for these founders and say anytime that somebody raises around, send them your, send them your little wrapper and.

Say, hey, put your term sheet in here.

Understand what you're about to embark on and it's good lead generation for you.

Yeah, yeah. I mean I think a lot of the board members could use the wrapper as well.

Got him. Okay, another question for you. Historically you've been heavily focused on software. You acquired 30 some odd businesses while at tiny, we've seen new categories attract venture dollars. Like basically traditional manufacturing businesses that have some sort of tech enablement. Is there a world in the future where that is an area of interest for you or are you all in on software indefinitely?

Everything is of interest to me, Jordi. I just love. There's always interesting problems to be solved that can create pareto efficient outcomes in everything. Right now there's a lot of them in software adventure, but. But yeah, there will. As long as like humans are not even in the driver's seat but like involved. There's going to be like weird, you know, dislocations and problems to be solved. It's like amazing how many times everyone could be better off if they just like, you know, it's not something that they could, it's not something they need a third party to do. But oftentimes like a third party really helps. So yeah, it'll be, it'll be on all kinds of things. I mean right now you're seeing the ventrification of middle America and as all these venture cos start to buy up vets and med spas and all these things like Venture capital. I'm in the rural Midwest right now, and these people are going to start to feel venture in a very tangible, physical way. And I think that's really going to change. It's just going to change the industry dramatically.

Do you have any. You had a great heuristic for, you know, early detection of potentially great founders being, I think you put it as stripe account plus high school diploma or something. High school report card. Can you break that down and tell me any more about that?

Yeah. If you made money online in high school, I honestly can't think of an exception of someone who's done that and has not ended up being like quite a successful founder in one way or another. This is like the Internet. You can just go on there with a laptop and make money. It's like, you don't gotta do all this. You don't gotta, like, go to Harvard and do YC and all that. Like, you can just go on there and like, drop ship stuff.

Make something that makes money. That's the new tagline.

Yeah.

What's your. What's your read on General Catalyst exploring an ipo? It's been something that everybody's sort of waiting for the first sort of scaled venture fund to go public. Do you see all the big firms going that direction over time? Or is it going to be a pretty binary in that Andreessen stays private forever, blah, blah, blah, or what does that look like?

I think one very simple cut at this is just firms that are started by founders versus firms that are started by investors. Mark and Ben, their founders, you know, gc, it's also founders, you know, David Fialko, obviously entrepreneur. And so I think, like, I think it's kind of expected to see that when a founder starts a business, even if it's an investing business, they're doing it. Why does any founder start a business? To grow a big business. And, you know, I thought Mark Andreessen put it nicely on Patrick's podcast when he talked about, you can be a small partnership that generates some good returns and then kind of is relocated to some Wikipedia page. Or you can build Goldman Sachs or JP Morgan or Andreessen Horowitz. I think it's also in the name. I think they put a lot of clues there. I think founders, great founders, are just going to build big important businesses. And your business can only really get so big if you're actually in the business of true alpha outperformance. You can't, you know, build J.P. morgan or Goldman Sachs or something.

Do you think that there's potentially alpha in getting into one of those big VC firms, but not in an investing role and instead going into an operating role like what you see as an, you know, not a deal team, but operating partner at Bain Capital or Texas Pacific Group or any of the large PE firms. Assuming that, that these large organizations are going to wind up actually operating more like PE firms, not just at financial scale but in terms of their behavior. And so you're going to come in and reliably be CEO of portfolio companies that are coming in through like the PE arm, essentially.

Yeah, I mean I think it's like joining Goldman in the 70s or 80s or something. You know, it kind of depends. It's actually kind of funny. This is like kind of a niche thing. But it's very interesting to me how COOs at funds are either like incredibly important heroic individuals that the firm would explode without or they're like completely back office kind of like low status, no one likes them. And they're also either they're either like extremely well paid or really badly paid. And so like it's one of these very like high variance roles. And I think it depends a lot on the culture of the firm. And so yeah, I mean I would argue that like a lot of these firms, most of the people there already are not strictly investors in the sense that they're not making judgment based capital allocation decisions. There's research and all different portfolio support and all this stuff which is like not what you would have thought venture was in 2005.

Yeah, yeah, I mean, I guess you just mentioned the idea of like having a venture owner is not the right thing. And that's where you come in a lot of times. But in the future some of these firms might actually kind of set up shops to become better owners of turned around assets essentially.

Right, they already are. I mean Heritage, you know, Sequoia Heritage is, is significantly or maybe not significantly, but it's like roughly the size of Sequoia. Yeah, I know Andreessen is doing kind of a Sequoia Heritage thing. GC is like they're all just starting to be giant at some sense, at some scale you just become an asset manager and really what you're looking for are you know, different fee paying products basically and like you really can think about it.

Yeah, I mean, I mean I guess there's still, there's still like a decision to be made between like oh, it's like an index fund and they're just going to like blindly take a slice of everything versus they're going to like that's kind of been like the, the meme around Andreessen a little bit is that they're just like in a lot of deals. But then you have General Catalyst where it's like we're buying a hospital network and that's much more like private equity. And so I understand that like all of this is going towards like venture becoming asset managers. But there are still built, there still will be forks in the road. Like Goldman Sachs does have a proprietary prop desk. Right. But have as much high frequency stuff as Jane Street. Right. There are different strategies and so it'll be interesting to see like where each fund kind of goes, I think.

One more, one more question for you. Mostly a prompt because we were talking about this in a group chat, I think a week ago there was a chart showing some of these new AI startups and their sort of revenue milestones. You laid out very clearly somebody in the chat, I think it was Bill, said something to the effect of are we short all of these? And I've kind of like struggled with the question because people get a tremendous amount of value out of these products. They love them. It's real revenue, it's real customers sort of trading dollars. You fired back and basically said iron law of the universe, easy come, easy go. Maybe just kind of extrapolate that in the context of the current AI market.

Yeah, look, I mean I'm sure some of them will be great businesses, but it does seem very true to me that just, you know, anything that can be done very quickly can sort of be undone. And I think you're seeing that. I mean I think all the kind of like ides, the best one changes from day to day. The switching cost is pretty low. These are very smart teams. I'm sure they will come up with moats and stickiness and all that kind of stuff. But I think it's kind of like if you think about the Internet, it's like, yeah, everyone's going to rush to use the Internet enabled version of XYZ thing, but does that mean there's the idea that Google is the last search engine, not the first. Right. And so I think for a lot of the companies where it's just like, I wish I had an AI version of this. I want like the consumer intent is like I want to buy an AI version of this, not like I want pick your kind of specific AI business. And so, you know, I think some of them will make it out, but I think a lot of them will not because it's just that's Just what the rush is. It's like anyone who can add AI into this is adding so much value that I'll just like buy it. I mean, I know like I pay for, you know, whatever, six different models. Like that's probably not gonna be true in three years. Right? Like there's probably gonna be one that I'll end up paying for too. There's certainly no reason to pay for six right now. And so, yeah, and like if you, if someone launches a slightly, like an incrementally better model tomorrow, they're going to get my 20 bucks a month until like, you know, until they don't. And so, yeah, I don't know, I just, I can't think of an example where like a whole class of businesses have cropped up, grown immensely quickly, and then kind of just been the new equilibrium, you know, that that's certainly not what happened in the dot com. There's some exceptions, but in general that makes sense.

We have a couple minutes left. How are you thinking about building your firm? Are you hiring for any roles right now? What does the future look like for Gafon and company?

Yeah, it's very interesting. I've been dutifully studying all the great solo GPS of the venture world and analyzing their setups and everything. I think it's very compelling. I think asset management is actually a great AI use case. So yeah, I think the answer is yes. I'm hiring little tech pros, cons and exclusive here. I'm hiring.

There we go.

Breaking, breaking news.

Scoop, put the chiron up.

Yeah, yeah, I'm gonna make one hire. I'm looking for like an investment principle, someone who's super smart, wants to be kind of the first hire, keep the team really small, a lot of economics to carve up. And basically like I am, I have more deal flow than I can handle right now and what I need is help basically. And so yeah, you know, if it just wants to be someone who joins me and then me and me and them and 30 AI agents, you know, allocating capital. If that sounds appealing, reach out.

Well, we have our next guest on the line right now. Thanks, Jeremy.

Jeremy, thank you for coming on.

We'll talk to you soon.

We love you.

Love you guys.

Talk soon.

This is fantastic.

Nick, welcome to the show.

Hey guys, I didn't mean to kick off the last night and you're all good? What's that? Booted, wrapped up three minutes early.

You're all good. How you doing?

Doing great, man.

How did investment committee go? Break down? Introduce yourself. What is what do you do? And then what does investment committee mean on a Monday? Like, who's in the meeting? What are you talking about?

Yeah.

When? A little asterisk first.

Yeah.

Nick just started lent, and he's not. He's giving up nicotine, so. He said he's a little bit grumpy today.

Yep.

But we're gonna. We're gonna turn it around because we're gonna eventually talk about Core Weave.

Yeah.

Which is maybe a 2000 bagger or something. And I was trying to run the numbers on it, but how did investment committee go?

It was good. Yeah. And we're actually having. I am having a poly crisis with all the. The companies that I'm on the board of, so that part's not good. So never become a venture capitalist. You know, I don't advise it to any young folks listening. Don't do it. Okay. Go become an artist.

Become an artist.

Become an artist.

What kind of art should the fans do these days?

What's that?

What kind of art is not going to be immediately commoditized by AI do you like sculpture?

No. I think you just have to learn to work with the AI, you know? You know, look at Grimes, our model. Grimes, AI. But people still like her.

Yeah, that's true. That's good.

How often do you take a board seat when you make an investment? Is that. You sound like future investments. You may just write the check and say, you know, good luck.

No, don't say that. Can potential investors hearing that? No, no, no. I'm on five boards. I've. Yeah, I got capacity. You know, I got. I got some space still. So if I'm leading and it's my deal, then we try and lead most of our rounds. So I will. I do. We do end up on a lot of boards, for sure.

Would you want to be on a public company board at any point? Like, you know how Andreessen has visibility into Facebook, and that's just, like, alpha for the entire portfolio, right?

Yeah.

And I think they pay you as well, so that. That part's nice. So I don't get paid to be on.

Well, maybe. Maybe Core Weave will bring you up. We'll bring you back.

You know, I'm silently hoping that they're gonna make your pitch for me to be on the board.

Explain the business to us. And if they like how you explain the business, then you're on the board. Boom. We'll send them this clip.

Yeah, we'll tell them. The technology brothers collectively requested that Nick be added to the board prior to the ipo.

Yeah.

And get a little. Yeah, maybe, maybe. Let's just break down.

I'm a very small shareholder.

Right, Right. For now.

Could start building that position up.

When did you invest in Core Weave? What were they doing at the time and what are they doing now? It's obviously been in the news, but I would love to hear it from you.

What a fun story, man. It's my first angel investment ever.

What?

Yeah, it was a good way to start becoming an angel investor. So I have like over 50 now, but that was my first and I was not accredited at the time.

I mean, you know, Dirty dog.

You don't.

Dirty dog.

No one checks, okay?

That's the trick.

No one checks, sis. He doesn't check. I don't think I even had the money to make the investment at the time. I think I borrowed it from my dad.

Amazing.

And I had met these guys because we were mutual haters of Ethereum. So we, we hated Ethereum proof of stake specifically. And I think I found them on Reddit. And Brian, Brian Venturo, he was posting on Reddit about how proof of stake was stupid and bad because they were ETH miners, right? They were mining eth. They're against the transition to proof of stake. And I was as well because I just love proof of work and hate proof of stake. So we, you know, we were kind of strange bedfellows. And then they were just mining eth. That's what they were doing. They had a different take on it than other ETH miners. They used more high end nvidias, which became very important later versus other eth miners, used AMDs, more commodity AMDs. And then we became friends. I became friends with Brandon and Brian and mostly centered on our mutual shared disdain for Ethereum. Ethereum culture. They had a very weird relationship with Ethereum. It was all their revenue, but also they just had a very antagonistic relationship. And then they pitched me the idea of repurposing their GPU hardware, just like idling by mining eth and other proof of work crypto assets and then also on a more bursty basis doing rendering stuff like cgi. Yeah, yeah, yeah. Like blender. So that AI wasn't even in the conversation. Yep, wasn't even in the discussion. This is 2018 though.

Yeah.

No one except for Gorn knew.

Yeah. Was this related to the render network? Like the octane renderer, rndr? Was that a thing or is this just like general purpose rendering for.

Do you know what they pitched me was like they were going to go to the blender conferences and be like the industrial HPC farm for blender and movie studios and whatever. And that sort of happened actually. But then, so I did the first round in which they raised capital. I think I was the only VC to be on the cap table. I did it personally, actually brought it to the fund and we didn't do it because it was seen as outside of scope, which was a catastrophic mistake obviously. Worst mistake of my career by far. Like I just ran the math that I did the numbers on it and.

Yeah, it would have been big.

It was our first fun, fun one would have been a 50x fund on that alone.

Wow.

If, if we'd done that check. So that was, that was a miss. And then what was the scope, you.

Know, back then like, like why was it not in scope?

Because we had done some crypto mining stuff and then we got really turned off by it and we invest in more asset light crypto businesses. So mining we thought was totally out of scope. And the whole pitch was we're going to stop doing crypto and we're going to pivot into this other stuff.

Yeah, makes sense.

So then in 2019 they almost went out of business because Eth moved to Proof of stake. So all the proof work market for GPUs ceased to exist. But right around the same time all of these early applications for LLMs like transform based LLMs came out. Like GPT2 was actually really big for them because there were all these apps that were built on AI dungeon, if you guys remember that. I think that was 2019. You guys remember that? That was my first. They were like, yeah, this thing's going crazy. Like the usage is hockey sticking. And I played it. I'm like, oh my God. Wow, this is actually like amazing technology. That's when I realized the transform model was going to change the world. And so then that just hockey sticked. And they obviously did an incredible job of being creative about financing. And Magnetar was obviously a great capital partner to them in many different ways over the years. And technically they've delivered. It's not an undifferentiated product. It really is. They built with the idea of HPC big performant synchronous HPC clusters in mind when absolutely no one was doing that. None of the hyperscalers were doing that. And then now of course everybody's realized AI specific data centers make sense. But Cory was really the first to do it and so they've been able to lock in long duration contract. So they've done very well.

Very cool. Changing subjects a Little bit. I want to go back to. It feels like a year ago. It wasn't that long ago. What was the vibe at the crypto ball, which you were at once everybody started realizing that the, the incoming president had launched a token.

No one knew at the ball that the Trump coin had launched or maybe others knew, but I certainly didn't know. I think it launched around 9pm and I didn't get out of there till like 11:30. And really.

So there wasn't too late. There wasn't. People were at the ball just hanging out, having fun and people weren't realizing what had happened yet.

Nobody that I knew at the ball knew. I certainly didn't know. No one was talking about it. So I felt that we were rug pulled, actually.

By.

Dr. Yeah. Because most, most conferences there's quite a lot of people just looking down at their phone the whole time because it's just not that. And obviously this wasn't a conference, it was a party. But the timing was hilarious. And then the lesson from this last Sunday was if there's ever anything White House related, you know, any type of crypto event, expect some big news. I thought you had some of the best takes on the strategic crypto reserve. Now that a few days have passed, have any of your thoughts changed? You know, you're generally not super excited about it. You had written a post before that you were sort of broadly against it. What's your thinking on today, March 5, and where does this all go?

Yeah, I mean, we're going to find out on Friday what they're really proposing or thinking. I think there's so many unanswered questions like how are they going to pay for it? Is Congress going to have to authorize? Are they just going to hold the seized bitcoin? Are they going to sell the seized bitcoin to buy Cardano? Is that really going to happen?

I mean, can you imagine, hey, portfolio rebalancing. You know, it's strategically important for the United States to, you know, diversify.

Yeah. I mean, really, like if, if they can't find some ersatz way to finance this thing without asking Congress. Because they can't ask Congress. Let's be real. Congress is not going to pass this.

Yeah.

So, okay, are they just going to use the bitcoins? They hold 200,000 bitcoins. Okay, we want to do a basket of bitcoin. Cardano Eth, Solana Ripple. So you're selling bitcoin to buy Ripple and Card. That's crazy. Come on. So we're Gonna find out what David Sachs has up his sleeve. I mean, I like the guy. I think he's smart. I don't know how this policy came to be. I don't support it, continue to not support it. Whether or not it's just Bitcoin or Bitcoin in a basket of other crypto assets, it doesn't make sense at all. There's no strategic purpose to owning crypto at the government level. Other strategic assets that we hold are commodities that are hard to acquire in a pinch. If we have liabilities, denominate those commodities. Right. It makes sense to have a petroleum reserve. It makes sense to have a tungsten reserve, uranium reserve, even medical equipment, agricultural commodities, whatever. Where is the industrial need to own Bitcoin or Ripple or Cardano? There is none. So I just don't understand the whole thing seems totally incoherent to me.

What? I think we share the same point of view on Sachs. Do you think that he is having real influence on the sort of messaging coming out of the White House related to crypto, or is he finding out live like the rest of us?

I think if it was just up to Sachs, we wouldn't be talking about the reserve. There wouldn't be a Trump meme, Coin, World, Liberty, Fi wouldn't be as big as it is. What other crazy stuff did Trump do with crypto? I don't know. I mean, I've just been hoping that Trump would just fire Gensler, put in Atkins, and then never say the word crypto ever again. We would be so much better. Instead, he's like hugging us to death, and I don't know where it's coming from. I don't know. Is it Baron? Is it Eric? Is it Don? Who's doing it? Who's the puppeteer? Whoever it is, they need to stop. Hopefully Sachs can put an end to all this nonsense. I just. It does not Trump to care.

It does feel like a distraction from the other side of Sachs's responsibilities, which is the AI stuff.

We heard a good take on the. On the strategic crypto reserve yesterday that I want to get your feedback on. Trump is known for coming out of the gate hot with a very aggressive proposal. Canada is going to be the 51st state. And then when all things are said or done, it's like a 5% tariff or something like that. And. And so the idea here was, hey, maybe the Cardano thing is, is this opening salvo of craziness that winds up just being pure Bitcoin reserve. Do you think that's strategically 4D chess possible. I know you're still against it, even if it lands there, but is it possible that that's what's happening here?

Yeah, certainly. I mean, the thing is, I kind of, like, disagree with the premise of the question, though. Like, everybody says this is an amazing negotiation tactic, but, like, I make deals for a living, right. This is what I do. If a founder came to me, they're like, yeah, you know, we're raising a pre seed. It's $100 million pre money.

Yeah.

And then they expect to get negotiated down to like 50 million or something. I'm like, this person is clearly insane and not, like, trustworthy, and I don't want to.

Especially if they have, like, a track record of doing just that and it's like, very public.

I mean, yeah, look, Chompy, he wrote the Art of Deal. He's a great deal maker. Whatever. I still don't like the strategy. I think it sucks. What about just, like, being a fair and honest broker, you know? Like, what about that?

Yeah, yeah, that's totally fine. Yeah.

Does Trump launch another crypto project while in office? Are you, Are you or you think he's had enough?

So he's launched, what, four, five, six.

If you count October Melania. Oh, yeah. Four since October.

So there's World Liberty five, there's Trump Coin, there's Melania Coin, there's the Reserve, which kind of is a project. There's probably something else that I'm forgetting.

Sure.

NFTs. He has Trump NFTs. Two editions, actually, not just one edition, to be clear, because I know a lot of beholders. The first edition of the Trump nft, they're very upset at the second edition.

Okay?

Dilution. And the Trump NFT market, everyone says.

Immutable, no inflated supply, no inflation. But when you issue a whole new batch, of course you're going to dilute.

That was the problem with Melania too. I mean, look, if we extrapolate this, it'll be 4,5000 Trump crypto projects by the end of his.

They're growing exponentially. The supply of Trump crypto projects is growing exponentially.

He loves it.

More broadly, where does crypto go from here? It's sort of. It feels like a very interesting moment because it's never had more attention, it's never had more users, it's never had more adoption. Right. Even, you know, with stablecoins, things like that. Yet the. The vibes are bad. Right. And the price action is not what you would expect out of the government saying, we're going to start buying these assets. Where does it go from here? How are you thinking about it as an investor?

I think markets are calling Trump's bluff. I actually don't think we're going to see a reserve. We might see a stockpile. I don't think we're going to see reserve where this. We're using government taxpayer funds regardless of how they're accounting for whatever accounting tricks they use. We're not going to see taxpayer funds used to buy crypto assets. That's crazy. Okay. That's not going to happen. Markets are calling us bluff also. You know, risk markets are going crazy. Everybody's warned. Everyone's worried about the tariffs and inflation and, you know, recessionary signals, whatever. That probably explains a lot of the sell off too. Vibes are really bad in crypto because the big L1 launches of this year have not done well. Meme coins are dead as a category, even though we have sort of like emerging regulatory clarity around it. The retail bid's just not there. I think retail got hosed one too many times. So they're just not willing to do that. They don't want to buy the $30 billion FDV hot new L1 that Andreessen put $400 million into. Right. Nobody wants to do that. Okay. There's no ultimate bias. There's just VCs buying from other VCs now with a lot of these launches. So in terms of liquid tokens, yeah, everyone's down bad. The cycle didn't happen. Bitcoin did. Well, nothing else happened. There was no alt season. Right. So the framework, people have this mental framework that's wrong, that cycles have to happen in a certain way and they happen every four years and it plays out like this. It didn't happen like that. In terms of actual usage of blockchains, it couldn't be better. Stablecoins are taking over all of fintech, all of global finance, all banking remittances, B2B international trade settlement, My corner of the world, little stablecoin corner, we couldn't be happier. But most people in crypto don't care about that. The coins matter and the coins might be dead for a generation. Here, here.

Yeah. How are you thinking about net new stablecoin investments? Where are the opportunities? Obviously anywhere that money fiat moves today is an opportunity for stables. But there's now been a bunch of teams sort of tackling these various areas for a while. Are you making a lot of net new kind of pure stablecoin bets or are you kind of letting your bets Ride.

Yeah, I have wires going out for two new stablecoin related deals this week so we're at 19 now in our portfolio I believe if you look at the stablecoin stack, I don't know how exciting this is but there's the issuers then there's the infrastructure like the intermediaries like the B2B stablecoin startups. Then you have consumer which are like fintechs built on stables. The issuer game is immensely competitive and the domain of banks and consortia now. So like. And also basically no one's ever created a new stablecoin issuer from scratch done well except for probably Athena I mean tether and circle dominate so that one's very hard. The BAB stablecoin payments like the PSP stablecoin based PSP thing immensely commoditized. There's 100 new bridges right. Everybody wants to be the next bridge. Very, very competitive, very hard going to be thin margin. I'm looking at weird stuff like the banking layer itself. How do you incorporate stablecoins into the core banking infrastructure? We're still looking at regional fintechs that have a stablecoin angle that's very interesting to me. So you're looking for moats that would be more of maybe regulatory driven moat or local network effect, something like that. But yeah, I think just like tossing checks into the next bridge that's probably a losing strategy.

Is tradfi even trying to move to stablecoins in a meaningful way like in AI, you know we see yeah ChatGPT is dominant but there's Gemini and Copilot from Microsoft and every hyperscaler has their products baked in. Is Western Union thinking about working on this or is it just, you know, just completely the trains just passing them by because it's more complex from the.

Regulatory perspective I would say the major fintechs, non bank fintechs are very interested. They're looking at making acquisitions.

Sure.

The remitters are either being disrupted by stablecoin remitters or building their own. So for sure that's happening there. The big credit cards like Visa and MasterCard are pretty active like the big global payments networks are. I would say all looking at it. The last bastion though is the banks. The domestic banks, foreign banks are actually doing a lot of stablecoin stuff depending like look at Sochen or Standard Chartered, tons and tons of stablecoin stuff. The domestic banks have PTSD from what happened over the last two years with Chokepoint. They are very afraid. I think many of them would like to be in the business of either issuing a stablecoin or being a service provider for stablecoin being the first mile, that's going to take a lot longer. They're still afraid that the Fed is going to destroy them if they so much just think about stable coins.

Is it shifting from regulatory arbitrage to more just like, I don't know, whatever. The inverse of fear, like boldness, courage, arbitrage. If you're courageous, there's an opportunity right now.

Now.

I think so. I think so.

That's good. That feels good.

There's still a reg arb actually. I mean, yeah, that's why people say regarb to mean crime. Like I don't see it that way. You know a lot of like fintech thought leaders on Twitter will say that. Like, oh, you're just doing reg arb. You're just criminal. Yeah, I mean stablecoin. There will be questions asked of how stablecoins work from a KYC and surveillance perspective. That's actually not settled. That's a big gray swan hanging over the industry. People don't really talk about much.

Grace 1. I never heard that term before. I like that.

It's like, like a black swan, but some of us know about it.

It's a known. It's a known unknown, not a. Not an unknown. Unknown.

Yeah, cool.

Like that. What other gray swans are out there?

Yeah. Grace one.

Give me your top five gray swans.

In crypto or just in general?

Yeah, yeah, in general.

I only know about crypto so I'm going to limit my.

Okay, that's fine.

Tether is the top gray swan in crypto. Some kind of.

Can you explain that? We have some crypto investors and entrepreneurs in the audience, but I would say it's not the average. Why is Tether so sketchy? I know the answer, but I'll let you kind of explain.

Yes, I mean, you know, they made. They did some like, let's say light accounting fraud in the past. You know, they were hacked for $850 million and then they. Or Bitfinex was. Or was it Bitfinex or Tether? Anyway, they co mingled their balance sheets. There was like a weird ersatz promissory note loan type thing. I mean, look, they're just an offshore dollar that serves people that want dollars on the blockchain in a kind of lightly surveilled way outside of the contours of the US banking system. So of course they're going to be used by a lot of insalubrious people. That was always going to happen. So the Wall Street Journal is always going to be digging up examples of kind of nasty criminal use, and there's only so much activity they can flag and freeze. So that's just. That's really it. Obviously the reserves have been questionable in the past. I think they're fine now.

And they're fine because even if they got hacked for 850 million, they just make so much money that presumably they would be able to cover the hole in the balance sheet. Is that the idea?

Well, that they could do today, yeah. I mean, the reason the hack was they were forced into it because no US bank would really serve them. So they had to use these sketchy banks and one of them scammed them, basically. So, like the sketchiness, it, like the fact that they've been sketchy is also kind of a function of the fact that the DOJ is like chasing them around.

Yeah, yeah.

Because they're like forced to use some of the worst intermediaries on the planet, if that makes sense.

Yeah.

Now they've been able to legitimate themselves. Now they have Canter Fitzgerald, they hold all the treasuries. Like, that's probably fine. But in the past that's. That's like the thing that people don't understand about Tether. The reason that they looked sketchy and they wouldn't disclose what banks are using, whatever they were untransparent was kind of because they were playing this cat and mouse game with the US government.

That makes sense.

Any other grace ones?

Well, if Coinbase's custody setup gets hacked or anything happens, they custody hundreds of billions. They were really the main custodian for all the ETFs. That would be absolutely catastrophic for the industry. I don't expect it to happen, but that's. That'd be bad.

Yeah. I'm curious if you're looking at businesses, there seems like some of the biggest opportunities in crypto outside of stablecoins are just solving the security problem. Right. And the lack of. Once funds are on chain, they can move around instantly. Right. We saw that with what's the firm that. That got hacked for? 1.4 billion. Two by bit by bit. We saw that with Bybit.

Right.

There's not a lot of recourse there, is there? Opportunities around security in the sort of broader, sort of like fat finger issue. We talked about this, you know, yesterday in the context of Citibank, you know, accidentally sending, you know, they're constantly sending money around, sort of not actually sending dollars, but like these dollars are sort of appearing in Client accounts. Are you looking at opportunities where software can kind of like solve for that or is that just going to be kind of ever present sort of risk in the industry?

Yeah. So first of all, yes, cyber is actually one of our biggest investment themes on a go forward basis. So we think that we've really underinvested in security in the crypto space. Second of all, what I think we should do is take crypto away from being a real time gross settlement network or like a bearer asset, a digital bearer asset, and then make it more like the way that traditional finance works. And I know people like Matt Levine are going to make fun of me for saying that, like, oh, you're just rebuilding traditional finance from scratch. My answer to that is yes, that's exactly what we're doing. Okay. But instead of working with fax machines or whatever and weird file transfer protocols that predate the Internet, we're going to do it on a better substrate and we're going to introduce this amazing thing called deferred settlement where you don't settle the payment instantly and in so doing you obtain scalability and you can obtain recourse. And so if we're able to create like pull payments as opposed to push, all crypto payments are push, that solves a lot of this problem. If we create a messaging layer alongside the settlement layer, that's what Swift is to wires, right? It's the messaging layer to the settlement layer. All of these structures that make traditional payments work and be safe and have recourse, they just don't exist in crypto. It's so weird. If you were to examine the crypto market from first principles, you'd be like, this is garbage, nonsense. I don't know how anybody uses it, I don't know how anybody trusts it. So yeah, it's just a matter of making it more like traditional payments and finance. So that's the next chapter here.

Yeah, I mean, I don't think it takes away from, we built an alternative to traditional financial rails and we did it in, I say the collective. We, we did it in 15 years and it's moving hundreds of billions of dollars around the world and it's showing a lot of potential. And I think it's totally, I think your point of view, and I share this view, is it's okay if we're just sort of making a better version of the sort of the legacy systems that are more Internet native. So. So I love that point of view. Do you have anything else, John? No, it's great to have you. We'd love to have you as our resident cryptographer.

I don't actually know any cryptography.

Well, it sounds cool. And that's what this show is about.

Cryptographer sounds like we're gonna have him on the show to like crack people's passwords.

It should be a fun segment.

Yeah.

Just don't ask me to explain a CK proof. I can't do that. I'm sorry. It's impossible. No one can actually.

Did any of the. Did that Dark Forest game go anywhere? People were pretty hyped on that for a while.

Was that like a paradigm thing?

I don't know. I. It was like a couple years ago it was ZK proof, powered like multiplayer online game and the in the map, like the fog of war is cryptographically sealed. So I can, we can both run open source installations. We know we're not hacking. It was kind of a cool idea in general crypto.

Rule of thumb, if you ask me. Did that crypto game go somewhere? The answer is no, and that's true 100% of the time. It's never not been true.

Yeah.

So there's the answer.

Yeah. What's interesting is that, I mean, did. What was the thing you were talking about with GPT2AI Dungeon.

Yeah.

I haven't seen like we have GPT 4.5 now and I'm still not seeing like Dungeon Games, AI generated games taking off. I see a lot of demos of people building games with Claude code, for example, or cursor, but I don't see a lot of people saying like this game is completely AI generated and it's a mobile game and we're printing millions of dollars and it's working. It's like a business. It seems, seems like somebody should have run with that or maybe they have and I just, I'm not aware of it.

It.

I don't know if you've seen anything like that.

I think AI Dungeon was popular because people could just like jailbreak the game and then just use it as a conventional LLM pre chat GPT.

But now you can just.

That's why it was so magical.

Interesting. Yeah.

But yeah, maybe this is the new benchmark. Like can you make a good game with AI soup to nuts? Like that's what Levels did. Right. Have you guys played his level?

Have you played it?

Yeah, yeah, I've, I, I've seen like a weird amount of time playing.

You play a lot. That's great.

There's a, there's a PMF or Die Blimpad.

Yeah.

Oh, you guys sponsoring It PMF or Die has a. Has like a blimp in the game.

I think they did some corrupt bargain or something and they didn't.

No. I think. Patty. Patty, one of the players was so excited that he just said, I'm paying for this out of pocket.

Oh, wow, that's great.

Very cool.

Well, I think this answers your question. You can make a game with AI. Yeah, it's just. It's a horrible, like, Flash game from 2000, but it's.

Hey, you're playing it. It's got your attention.

Yeah, I know. They got me.

It's a Skinner box. You can make a Skinner box.

You're. He's a whale. On. On the. On the Peter. On the Peter game.

On the levels. IO flying flight simulator.

That's great.

It's a terrible game, but it's fun. You guys should get on there.

I will go.

I will go play. We'll see you in that.

We'll see you in the air next time we're on. We'll all be playing at the same time.

See you in the metaverse.

Yeah.

Thank you for coming on, Nick. It's great.

Take care, guys.

Bye.

Bye.

It's a lot of fun.

Apparently levels is up to 58k mrr in about 14 days.

It's really ripping with this.

According to Heskel in the chat.

That's pretty good.

It's funny because he has enough of an audience now that he can almost release anything and hit metrics like that.

It's great.

It's sort of potentially warping the. The actual pmf. Who knows though? Do we got a couple of posts we can run through before?

Well, I mean, we're inviting a Sequoia investor on and you know, Sequoia is an investor in Ramp, so let's do a Ramp ad.

There we go.

Time is Money Save. Both easy to use. Corporate cards, bill payments, accounting, and a whole lot more all in one place. Go to ramp.com to get started and what? Oh, there was news about Core Weave. They're buying weights and biases for 1.7 billion. Did you see this deal?

I saw that.

Have you ever seen weights and biases before? I used to watch these videos called Two Minute Papers and they had this very interesting professor who would have all these funny turns of phrases and he would always say, hold on to your papers whenever he saw some AI breakthrough. And he had a bunch of other funny little phrases that everyone would. He had a great community of just giving you updates on, like what the latest AI papers are and weights and Biases was like his main sponsor and so that's where I know about them. So weights and Biases. They have paid tools start for $50 per month for professionals have attracted data scientists from various industries including pharmaceuticals and medical imaging firms. It's more than 900,000 users hail from companies including OpenAI, Siemens and Salesforce. And so Corey is expected to IPO at a public market valuation north of 20 billion. But its concentrated customer base has prompted investors to debate how long the company's growth will continue. Weights and Biases competes with major cloud providers and other AI focused firms such as databricks Hugging Face, which sells similar tools to developers. Weights and biases are common terms in the AI field. The revenue, the information couldn't figure out the revenue. But they raised at a billion dollar valuation in 2021 and then 1.25 billion slight up round in 2023. Company started in 2018 and previously raised 250 million in funding from investors including CO2 Management, Insight Partners and Felicis. I can never pronounce that. Felicis. Is that what it is?

Yeah.

I don't know.

I think you got it at the end.

Well, let's see about Sean. Let's see. It's 12:31.

I'm here, I'm here. What's up, boys?

Amazing. Welcome to the show.

What's going on?

I'm scared to pop on with you guys. You guys are wittier than me and that's unfair territory.

Well. Oh, I love the T shirt. Counter Strike.

I figured you guys would like this one.

What versions of Counter Strike did you play throughout the years?

So I was unbelievably hardcore. And I mean, I played early, but primarily 1.0 to 1.6.

1.0. I started in 1.3.

I was multiple Cal invite teams. Cal, baby, me too. Did you?

Yeah, yeah, yeah. Like God's exiled angels. Gxa.

I don't remember.

And then ek the elite Knights. For a while. I was on a bunch of these teams.

I remember. I remember Dek. I was. I mean, I'm probably older than you.

Yeah.

I was on a team called htp High Tech Performance One called neg.

Did you ever go to the cpl?

Of course I went.

I could never go. I could never convince my parents to go because I was too young.

Yeah.

They're like, what are you doing? Lugging a CRT monitor out to Texas? No way.

Exactly. In the Volvo station wagon.

Yeah, yeah, exactly.

I and I would go to a lot of tournaments and SoCal is the best.

Oh, yeah.

Yeah.

I remember going to a Best Buy. A tournament at a Best Buy, and we played against this team called. What was it? It was. These guys made one of the protocols. Before 3G, there were. There were two different wireless protocols. I forget what they were called.

I mean, we. So, I mean, I think. I can't remember the timing. Like, maybe one G was primarily tdma, like the original time division multiplexing.

That's right.

G was primarily like, cdma.

Cdma, yeah. So we. So we go up against these GU multiplier, and their. And their clan tag is cdma. And we're like, why? What. What does that stand for? Like, it is like, you know, cool guys. You know, like, everyone. Everyone's name usually is like, something. A reference to, like, the game. And their name was like. Was like, this reference to this wireless protocol. And. And we were like, why did you name it that? And they're like, well, we're the guys that invented that. And, like, we're the founders of the company or something that did something. And they absolutely destroyed us. They were really, really good.

Yeah.

Yeah, exactly. They were like Qualcomm engineers, basically. And so they destroy us. And then they felt so bad because we were, like, 13 years old that they gave us the prizes, which was great.

So there was. When I first started playing, the best team in the world was called tso, and they were based in Seattle.

Yeah.

And they were like. I never met them because I was a little kid at the time, but they were described as, like, telecom engineers, too, but they were more on. I think more on the infrastructure side. Like, they, like, they would actually build the early fiber connections and stuff.

And so, I mean, that was how you got an advantage back then. Like, if you could figure out a better Internet connection.

Networking, baby.

Yeah. You would have an advantage every time you played.

And so I got an F in algebra 2 in high school, but I. But I got a networking education through Counter Strike.

Like, yeah, fantastic. Well, I don't know.

Are we live, guys?

We. We are live. We've been live. Anyway, Sean McGuire's here, partner.

Yeah, we're live. I'm so glad we didn't have, like, a Zuck moment. Zuck was on Rogan and he was like, pretend he was larping like he does, like, bow hunting. And, like, Joe Rogan started kind of pressing him on it. Like, John's asking you, like, oh, you got the Counter Strike T shirt. And, like, you actually could, like, back it up. But, like, of course, like, of course, Zach, like, actually didn't have much Substance behind it. I'm sure he had it. But no, we are live super excited to chat.

There's been so many Elon companies. That's what we want to dive into today.

Yeah. First of all, we are a show built on X. Take a little victory lap on X. You invested in take private.

Yes.

$800 million, guys. And no victory laps until it's exited in liquid. But there has been some saving of democracy that's been happening.

No.

And the bigger thing is, I think around six months ago, there had been legacy media like the Washington Post basically reporting coming out and saying Sequoia Capital and listing out all the investors have lost $20 billion on this investment. That obviously wasn't true. It had been I think marked down internally and some of the other asset managers had. But it was more so to kind of reprice it for employees. And you know, it's very clear that the. In many ways there's two things happening right now. One, people that used to be active on Twitter that are less active now, they message us and they say like, I'm sharing this guy Social. Yeah. But I think the app has never been better. Yeah. And it's very clear that it's here to stay. So I just wanted to give you a little bit of credit there. You never lost conviction.

I mean, no, on. On that. I never lost conviction. A lot of people did. And I think that I. People. People are very susceptible to believing what they're told. And I think if you believed what you were told by the media, then you could understand, oh, it's dead. But if you use your own eyes, it's like the whole thing is operating. No major outages, Elon at the helm, product innovation happening and like lowering costs. I honestly found that to be one of the more interesting cultural mass hysteria moments that I've lived through. And that was just so easily disprovable in a short period of time. But teach their own.

I'm glad you only had to wait a couple years. It would have been more excruciating if it had taken much longer to get kind of repriced. What? There's so many different things we want to run through today. I'd love to get your, you know.

Can we run through your guys style? Like how are you guys so well dressed?

Well, we had to come out the gates. We were calling ourselves, you know, we tech bro is sort of derogatory. We were the technology brothers. We needed to kind of reclaim it, elevate it. And part of that was dressing, dressing nice. But you've had your sort of hands across the whole sort of Elon empire. Obviously have a relationship with Elon. I kind of want to go company by company and just kind of get your take on them. Right now. I think Tesla's very interesting. It's, it's not a pure play humanoid bet, but it does feel like Humanoids are the next platform for Tesla. Especially when you look at sort of of generalized EV sales for Tesla have not been performing great. Talk about kind of the humanoid market. I'm sure you've gotten pitched by a lot of these companies. I'm guessing you're still a Tesla shareholder. But talk to us about that category because it seems to have everybody's attention for good reason.

So let me. First of all, I work at a registered investment advisor. None of it is financial advice. It's my own entertainment. Also not the opinion of Sequoia Capital. This is all as a dumb individual, Sean McGuire nonsense. And also, we're not investors in Tesla, just given the timing and that it was public. We're major investors in SpaceX, invested in Boring Company, investors in X, investors in XAI. And I love all those companies and very optimistic about all of them on Humanoids. Look, I strongly believe it's going to be a real thing and a giant thing. I think the only question is the time frame for when this happens. I don't think it's an if, I think it's a when. I really struggle to make a forecast there in part because there's such a complex supply chain and there's just so many different subcomponents from, you know, my partner Konstantin has spent a lot of time thinking about rare earth metals and like getting, you know, different rare earths for making magnets all the way to, you know, like scaling manufacturing to, you know, like just every individual sensor, every actuator. Like I really just don't know the scaling of it. But I was at the Tesla Optimus event. I can't remember when, maybe, you know, months ago or whatever. And it was more impressive in person than I think people realized online. Like you. I was just standing there in the street. It was at some studio in la, and you're just standing there and all of a sudden like 20 robots started walking out from behind a curtain right next to me. And I think it's one thing when you just see a robot video, it's another thing when you see a tethered robot Demo to see 20 untethered robots walking in a progression. And I genuinely couldn't tell at first if it was human in a costume or a robot. And I had to figure out, how do I verify? Because it's about the same size, and they were maybe 20ft away when they first started coming out. And so I was scanning down the whole thing, and then you look at the waist, and the waist was very narrow. It's like, holy crap, that really is a robot just walking. And they were out there for about two hours interacting with the crowd. And.

And.

Even if there was some human control for safety reasons and all of that, the vast majority of actuation and all that was autonomous, meaning whenever you have multiple actuators, you get feedback loops of all these things, and no human can handle all the feedback loops intuitively. And so you need controllers that smooth out all the noise and can map whether it's a computer brain that's telling you the goal or a human telling you the goal. It's almost like with computers, how you take raw electronic signal in transistors and then you have very, very low level programming, and then you have assembly, you have firmware, you have operating system, you have higher level systems, you have all these things up to queue. It's like everything other than maybe the final operating system level was everything else was basically autonomous. And it was insanely impressive. And so it really gave me the sense of the stuff that's coming soon. I just don't know exactly when. And look, I think that for Elon, he is playing all the different angles of AI from foundation model with XAI to robots, which I think will be the ultimate embodiment of the combination of if you have AGI plus a robot, that's what lets you do kind of of unlimited amounts of physical work. In the real world, the foundation model lets you do kind of unlimited just knowledge work. And then neuralink lets you merge with the AI. If we actually get AGI, that's way smarter than us, maybe existential for humans, but if we have neuralink, we can merge. Just put the AGI in your own head and now you're, you know, an enhanced human.

You know, we're back to playing Counter Strike. Me and you are going to be piloting Tesla optimus robots against each other, shooting each other.

Like, no human can beat the chess systems, but it's still fun to play chess. You still have Magnus?

Yeah.

You know, Heroku or whatever his name is. You know, we're still going to be tearing each other up at the land, and I think we'll be doing more of that with AGI.

Do you think it's a good time for me to put a deposit down on a Tesla Roadster. I think. I think Elon might deliver it. I think we've been waiting but all the pieces are falling into place. I think it's going to be an epic car when it drops. What do you think?

I. I actually have to be careful around any public company. No comments.

I just want to do it with the car. Yeah, let's move on.

So let's talk some of your recent investments. So talk to us about Reflect Orbital. That round got announced last year. Very exciting company. I would love to hear it in your words and then a little bit more about your investment thesis and I'll stop there.

No, I love it. So for anyone that doesn't know Reflect Orbital is an absolutely insane idea where they want to eventually have a constellation of satellites that reflect sunlight back down to the earth with kind of two phases of the business. The first phase is lighting applications. And that's everything from like imagine in the LA fires, firefighters operating at night couldn't see anything. It's genuinely really difficult to bite a fire at night. And being able to just put sunlight and let these guys operate almost in daylight, say in a disaster when there's a hurricane like the hurricane in North Carolina or anywhere there's lighting or let's say you're doing a giant construction project somewhere away from the city, that the lighting costs on these projects are insane. And to basically let you extend the hours, especially like in northern latitudes, let you actually do construction more year round. But the longer term market is energy. And I'll come back to the thesis and why I was willing to make this investment in a second. But basically the real the crazy big business someday is extending the useful time of solar power plants by a few hours a day. So depends on the latitude and time of year. But a solar farm will operate with decent efficiency, call it 10 hours a day on average, something like that. And by having this reflector ring that basically lives the satellite is a ring that basically the constellation is a ring that basically only lives at the interface of of sunrise and sunset all the time. Like just the light dark boundary. And you can basically extend the very high efficiency time of a solar farm by call it two hours before sunrise and two hours after sunset with this constellation. So anyways, insane idea. I have been involved in the space community for, I mean officially for over 15 years. Unofficially for most of my life. I've been obsessed with space since I was a little kid. Like literally obsessed. I can show you guys a sick ass video from I Think sixth grade where I did this thing called satellite amateur radio experiment Sarex, which was this NASA program Sick Retro 90s video where we learned ham radio and talked to astronauts aboard one of the shuttle missions using ham radio. Got to ask a question in sixth grade, it was pretty fun. But I've been obsessed with space my whole life. I've been skeptical on basically every space business model I've ever seen. The Reflect guys, there's a few things. One, I just profoundly believe that the launch costs plus a bit like mass to orbit over the next five years or so is going to increase. The costs are going to drop dramatically and mass orbit is going to increase dramatically. Basically guaranteed. SpaceX is driving this. If you look at the mass to orbit curves from the last five years, it's incredible how much more mass to orbit we're doing every year because of SpaceX, the cost hasn't dropped as much, but I am very confident it will with Starship. And so we're getting to the point where you don't want to start making space investments once all of that has already happened, then you're late and it's competitive. You want to be a little bit ahead of the curve. So I think we're at a point right now where for the right businesses it makes sense to be building them right now so you can actually be alive and be scaling a little bit as we hit. You can ride with the cost and mass curve. For what it's worth, I passed on Varda early on. Obviously you guys are very close to Delian. I love Delian. I think I should have done the investment and I'll tie it to Reflect in a second. But with Varda, in my opinion, in the seed pitch, I think they had the wrong micro idea but the right macro idea. And it was my mistake for wanting both the right micro and the right macro. And to unpack that with Delian, the first product that they wanted to do was ZB Land. It's like this ultra pure, you know, fiber optic cables that are really good for very long range communications. To me, I just don't think that's a very important product. And I actually think that with Starlink we're going to need very little like zblan in the future. Almost all long range data movements in the future is going to be done via Starlink. In my opinion. I think starLink has like a 10x plus cost advantage per gigabyte for long haul data transmission compared to say putting a fiber optic cable line under the ocean. And that's where you actually get a benefit from something like Zebuland. And so anyways, Dylan and I love to argue with each other. Super smart guy. That's one where I had been so involved in space for so long that I already had a thesis on Zeb Land and already thought it was irrelevant with Starlink and kind of passed because of that. But the macro I think was 100% right of just like we're at the right time in space and just building a cracked team of space engineers that know how to get things in and out of space. I think that will be very valuable and I think it's hard to imagine what Varda's revenue driving product will be in five to 10 years, but I think they will have a successful product. I just can't tell you what it will be. So anyways, back to Reflect. I've met with so many space teams in the past and these guys just, they're just insanely good and they know and they have like a real. Even though it's such a crazy idea, I think that the best businesses in history oftentimes sounded like insane ideas early on. You know, Sequoia was a seed investor in Airbnb. Hearing the stories of how people thought about Airbnb in the beginning, it's crazy. People were like, I would never share my house with a stranger. They're going to trash it, they're going to use all my toilet paper, whatever people's issues were. But lo and behold, it worked. And Sequoia made a huge amount of money from this and they couldn't even raise the seed round. If you look at SpaceX, when Elon started it, everyone thought he was insane. You're an Internet entrepreneur who's going to build a rocket company. Isn't there NASA for that? I think so many of the best businesses in history really sounded insane. But for Reflect, I think they have. If you really go through the details, and I think very few people have actually gone through the details, they have unbelievable macro tailwinds. So one is just dropping launch costs. Second is I don't think you could have done this business 10 years ago even if the launch was there because we hadn't yet built enough solar power plants on the ground. Something that's nice to reflect is we don't have to build any of the ground infrastructure. 0 There's already 10,000 solar power plants. As we get to leverage that anyways.

You go is the idea that Reflect would take basically a rev share on the incremental sort of energy generated that they are directing at these farms. And is that why?

Yeah, exactly. Right. So basically like for one of these solar farms, there is almost no additional operating expense to operate for these extra colleges. Four hours throughout the day. It's literally just like silicon photovoltaic sitting on the ground that is otherwise just not generating power. And so it's like either they get no revenue from this or if they work with us then they get revenue. And it has, it actually has very profound ramifications to the kind of financial engineering of running a solar farm. A lot of these utility style infrastructure projects, you basically juice it where you use debt to build the solar farm and your revenue, if we call it the first 15 years, is just paying off all the debt you have. And then years 15 to 30 is a lot of profit. But it takes a long time to get to the point where you've paid off a lot of the loan and it's generating lots of profit. It if you can increase the useful time by 40% whatever of a solar farm, that's all profit upfront, it's insanely valuable. And so it's like an incredibly easy sale. It's one of these things, if you build it, they will come and they will come in droves. But the point is this company would have made zero sense before there were 10,000 solar farms on the ground because they building each of the. We couldn't build those ourselves, it'd be too expensive. But now that they're there we have just this very clear value prop and like I think these guys actually know how to build the satellites needed and anyways it's an easy idea to shit on and it may not work, you know like the. But I believe in taking risk and I think these guys what I'm excited.

For, I want to sit outside at night, have set up our podcast set. You come over and the sun just appears. We do the, you know, the first podcast under Space sunlight question for you. I think you know, it's easy to see why SpaceX is a multi hundred billion dollar company in many ways that feels cheap at times being the sort of toll road to space. At the same time, how do you advise, is there sort of any type of fear among founders that are building space faring startups that what if SpaceX does this right so you're benefiting from the sort of decreased launch costs but at the same time there's this risk that somebody that has even more access to space can sort of enter that market. How do you think about that risk?

Well, I think that, I think that Elon has probably the biggest target on his back from a like regulatory perspective of anyone. And there's no one I'm aware of who takes like antitrust and all that more seriously than him. And like I, I think that those guys are going to like, they're just building the best launch platform in the world and they probably will launch some other business. Like obviously they launched Starlink, but I.

Think there's also a fixation on Mars that's sort of helpful. Right. It's actually great that the attention is going there and there's plenty of opportunity in sort of low Earth orbit for the rest of us.

Yeah, so I think that's a great point. Look, I think SpaceX is deeply incentivized to have a giant ecosystem of lots of great independent companies in space. And I think especially in leo. And yeah, you're right, I mean I think the focus is going to be Mars in addition to starlink which is already operating and scaling and on Mars. I'll just say I was, I mean I've always thought just from an intellectual perspective it's the coolest thing ever to build a city on Mars. But when I wrote my initial SpaceX investment memo, it's like, yeah, I don't really understand Mars from a business perspective, but I think this is going to be such a great company, irrespective of it, that it's fine, it's a side quest from a business perspective. I am now at the point where I really believe that Mars is going to be like one of the biggest profit generating endeavors in history. And not necessarily, not like from the direct like building a city on Mars, but from all the indirect things. And I'm sure you guys remember all the myths about, or you know, all the stories about NASA in the 60s and 70s, all the spin off technology, microwaves, materials and all these different things. I think the spinoff technologies from having a clear goal of building a self sustaining city on Mars which will have about a million people or whatever in 30 years, the spin off is going to lead to space, probably going to have to make sure that nuclear power plants in a box get built that are very safe and reliable and you can transport in space. It's going to have to perfect vertical farming, going to have to perfect surgical robots, going to have to, or just have optimists do it, but it's, there's so many things going to have to make fabs that you can assemble in space that can remake whatever leading edge silicon you need. Going to have to do mining. I think all the second order effect spin out. Businesses that are going to come from this are going to be outrageously valuable. And there's something about when you have a clear goal and all the constraints of being in Mars one, it lets you recruit the best people in the world because it's such a cool project and the best people will do more than the second, like a level below. And the second is just having all these constraints forces you to come up with solutions you would never come up with otherwise. And those are oftentimes better solutions. So anyways, I think we're going to. I think the company's just going to rip.

We have one more minute. I want some advice for new entrepreneurs. New grad wants to work with you, wants to work with Sequoia, wants to start a company at some point. What are you telling them? You had a little bit of an uncommon path, PhD, DARPA, algebra too. What do you recommend for new grads going into entrepreneurship in the age of AI, in the age of Mars colonization?

I mean, look, it's not what people want to hear, but I think that if you don't have a very, very clear idea, don't start a company just for the sake of starting a company. Go join SpaceX or go join Xai or Tesla or go join Varda, go join some amazing company for a year or two. I strongly think that's better than just trying to start a company for the sake of starting a company. And then when you do have an idea for a company, don't listen to what other people tell you. Even, even me. Even like VCs, you got to kind of think for yourself and a lot of the best ideas seem crazy.

Great place to end.

Great place to end.

You are a new resident space brother.

Thank you.

Great having you.

I'm fantastic.

Nothing makes me happier than being your guy's brother. Music from my ears.

Welcome. Welcome to the brotherhood.

Amazing. Thanks for coming.

We're going to play Counter Strike someday.

We should, we should. Let's have a late party.

Mid pod.

Yeah, yeah.

Have a good one. Thanks for coming on.

Thanks for having you. Fantastic.

I got a call right now.

Yeah, we got to wrap up. This was a fantastic show. Thank you for watching. Thank you for listening. Tune in tomorrow. We're doing more call ins, we're going to do more timeline, more deep dives. We're just doing more of everything. That's the theme of this year. More and better and more and better and more and better and more and better.

Moss.

So it's the live Moss lifestyle.

Live Moss lifestyle.

Yes.

Thank you, everyone.

Thanks, everyone. Talk to you soon.