Loading summary
Transcript185 lines
- [00:00]
A
Foreign. Welcome back to Firewall. I'm your host, Bradley Tusk. It's a Tuesday episode, so with this is our friend and producer, Hugo Lindgren. Hugo, how are you?
- [00:15]
B
Good morning, Bradley. How are you?
- [00:17]
A
I'm good, I'm good.
- [00:18]
B
So you had like a kind of a crazy weekend, which we're going to talk about, like a 15 mile walk, 50 miles. Okay, we're going to talk about that in a second. But we do. We live in New York. We are New Yorkers. We are New York Knicks fans. Obviously, anybody who listens by podcast knows it. And Wednesday night, the NBA final starts. The Knicks. I'm just looking at Kalshee right now. San Antonio is a 64% favorite, which sounds about right.
- [00:46]
A
So, I mean, you know, I'm trying to balance a few competing things.
- [00:50]
B
Yes, me too.
- [00:51]
A
I'm trying to balance one. From a purely rational standpoint. San Antonio is a better team.
- [00:57]
B
How good did they look the other night?
- [00:59]
A
Right. And they have a guy in Wemby who's just like, you know, Lyle and I were talking about this. You can't really guard him because if he's hitting his three point shots, he can shoot over anybody. Without even jumping, by the way.
- [01:09]
B
No, without looking like he's trying that hard. But he.
- [01:12]
A
So if you can't stop him, it's really, really hard to beat him, as Oklahoma City, who was considered the best team in the league, just found out. So rationally, you can't really disagree with Kalshee or all the different odds, because they sound right. Although the Knicks did beat the spurs two out of three games this year. And the game that we lost was, I think. I think only by a point. I remember watching the game. I think Champagne had a shot at the very end to beat us.
- [01:37]
B
But just like we're a different team, I hate that we're. I take the we. But the Knicks are.
- [01:41]
A
No, the Knicks are for sure.
- [01:43]
B
But the Knicks are a different team. The spurs are different teams.
- [01:45]
A
Both teams are a lot better.
- [01:46]
B
Yeah.
- [01:46]
A
By the way, they met in the Finals for the NBA Cup. And then I don't know if that's probably never happened before, where the two teams that met in the Finals for the NBA cup then meet in the Finals for the actual champions.
- [01:54]
B
Well, how many years have they done the cup?
- [01:55]
A
Three or something? Is it three? But nonetheless, the point is these clearly are the two best teams in their conferences. Right. You know, because they keep making the last round. So rationally, they're better. I also know that as a Knicks fan and A Mets fan. I am conditioned to believe that my team will always lose. And I am convinced that if I believe otherwise, then they will definitely lose. So I have to believe in what
- [02:21]
B
is all about you, Brad, so that
- [02:22]
A
they can win, right? Of course. And so. And with all that said, I'm so excited. Right. Like, you know, we'll talk about this walk, but, you know, we did a. My friend Kenny, who I did it with, we walk 15 miles. He had a Knicks hat on. At least a dozen people, strangers on the street, like, called out, go Knicks to us or adapt us or whatever.
- [02:41]
B
Vibes the city. It's so good. I feel a lot like 99, I have to say, where the Knicks went
- [02:45]
A
to the Finals, I think it's much better. I don't know.
- [02:47]
B
It's better. But, I mean, the way I feel is that I remember when they got. I mean, that was a very improbable run. That was not their best team of that decade by far, but they got to the finals, and I remember thinking, like, you know what? I'm just psyched to play the Spurs. I'm just psyched that they got this far, and I do feel a little bit that way with this team. Like, they played so well.
- [03:05]
A
It's.
- [03:06]
B
The season will not be, like, a disaster.
- [03:09]
A
We can win this. We really can. And this is easily the best Knicks team of my lifetime that I've ever watched.
- [03:15]
B
Totally.
- [03:15]
A
And so, you know, so give me your pick.
- [03:19]
B
Just.
- [03:19]
A
I'm gonna. I'm gonna give you two picks.
- [03:23]
B
No, no, no, no, no. Two picks.
- [03:24]
A
No. Two picks.
- [03:25]
B
One pick.
- [03:26]
A
One pick Knicks. And six.
- [03:27]
B
Okay, I'm gonna pick the Spurs. And six. I. I want you to be right. I definitely think the Knicks can win.
- [03:33]
A
Yes.
- [03:33]
B
Like, it's. It's. There, there. They're. They have no chance, you know?
- [03:38]
A
They have a chance. Correct. We have an excited season. So the first we got to take, you know, the first two games are, you know, Wednesday and Friday in San Antonio. We got to take one out of the first two. You can't go down to nothing.
- [03:48]
B
Yeah.
- [03:48]
A
And then if we do, then it's like, okay, now it's the best of five series, and, you know, let's keep narrowing it down. So. But I am incredibly excited. One thing I'm trying to figure out is how early for game. Game three, which is the first home game. If Trump does go, how much earlier do you have to get there to get in, like, how much more security? So I really hope he doesn't go.
- [04:09]
B
He will go to the game.
- [04:10]
A
He says he will.
- [04:11]
B
I know, but like, isn't someone going to tell him, look, you're going to get booed crazily. Like, you can't go.
- [04:15]
A
Apparently Dolan, like, threatened people if you boo or something. I don't know.
- [04:18]
B
I mean, I could see that, like, he would do that, but still, that's not going to stop people.
- [04:22]
A
No, I will, I will boo him if I see him.
- [04:24]
B
Yeah, I hope you do.
- [04:25]
A
I will. So. And then if I no longer next season ticket holder next year, we'll know why. But anyway, so, yeah, I'm incredibly excited and just, you know, it's a great time to be in New York. That's a good podcast.
- [04:40]
B
Yeah.
- [04:41]
A
Yeah.
- [04:41]
B
Goodbye, guys. That's all that everybody cares about this week. So do you want to talk? You have two major subjects.
- [04:49]
A
Yeah, well, so the biggest one by far is, you know, I noticed a couple times in different episodes when the notion of investing in an OpenAI stock once it goes public, my instinctual reaction time was like, that's idiotic. Why would you do that? And then they announced, or at least there was sort of a lot of public speculation about our September public listing. And so I decided to really dive deep.
- [05:14]
B
Okay, and that's unusual for you because you definitely don't consider yourself a market guy at all.
- [05:19]
A
I don't trade stocks.
- [05:20]
B
Right.
- [05:20]
A
I mean, I'm sure I own various indexes and whatnot, but, like, it's not. I don't trade stocks at all.
- [05:26]
B
How closely do you even follow it?
- [05:28]
A
You know, not that closely at all. But, you know, because it is a startup going public, I'm much more likely to pay attention to a startup that's going public than, you know, IBM or whatever. Right? So, like, and I just did a really, really deep dive into the numbers to figure out, like, should. And again, this is not investment advice, but, you know, from my perspective and analysis, is it. Is there any reason to invest in open open AI once they go public?
- [05:56]
B
Okay, I want to ask you two set of questions and then obviously you chose to do OpenAI, not SpaceX. SpaceX is going to be in June. Why'd you pick OpenAI?
- [06:05]
A
I think it's much more complicated. Okay, so look, on one hand I could write the same thing for SpaceX, and I don't know that I would have. I probably wouldn't have a different conclusion. And if anything, the multiple for SpaceX of, you know, 17 billion in revenue compared to $1.8 trillion valuation is even crazier than what OpenAI would be. But with that said, I just think that SpaceX in some ways is a more conventional company. Obviously what they do is radically interesting and different. But they make hardware, they sell it to the government, they launch it into space, they charge people subscription. In some ways it's a more straightforward business. The other reason why is Elon has this ability to defy reality. So Tesla is one of the only stocks I'm aware of where the fundamentals of the business and the share price don't ever align. And usually the market always corrects that eventually. In the case of Tesla, it never does, right?
- [07:06]
B
It hasn't yet.
- [07:06]
A
I mean, but it's been years, right? So he has this pixie dust that I think. I don't know that it'll potter SpaceX, but I don't know that it won't. And if it does, then all of the analysis you're about to hear just doesn't really apply. And so it's like, how do you analyze something in a way that, that doesn't adhere to the normal market realities in the first place? So that was why I picked OpenAI.
- [07:31]
B
And I want to frame this a little bit like you also, one of the major events of your life is Uber going public a transformational thing in your life. How does that affect your just perspective on this?
- [07:46]
A
Well, what did you learn from it?
- [07:49]
B
Because it was obviously a. I mean,
- [07:51]
A
the biggest thing I learned from the actual IPO was that I think it took too long and that would argue in favor of the timing for OpenAI and SpaceX. I don't mean to.
- [08:02]
B
They just waited too long.
- [08:02]
A
Waited too long. I don't mean took too long for me to have liquidity, right? I mean that I think they actually would have had a higher evaluation if they gone public sooner. I think there was a, a point of maximum hype, right? And I think they got weighted past that point. And then eventually, as a result, the end result wasn't quite. It was excellent, but it wasn't quite as good.
- [08:23]
B
Right.
- [08:24]
A
So that was kind of my takeaway was, you know, don't wait too long to go public. But also, and this is true in general market conditions have a huge impact when a company goes public. So there are some, if you take the open eyes and the Ubers of the world out of it and just take a normal, highly successful startup, right, to be able to legitimately do a public offering, you need at least half a billion dollars in revenue, but really much more like a billion dollars in revenue. It'd be a pretty small IPO at half a Billion. And, and you need to be, usually unless you are something like an OpenAI or a SpaceX, that's just considered in a different category, cash flow positive. Right. Or at least a very clear path to that. And for those kinds of companies, you know, there are some metrics like that, but there's also just the market conditions. Is this a good time for IPOs? Are stocks generally trading really well? What's inflation like? What's unemployment like? I mean, there are ultimate macroeconomic conditions that have some bearing as to the right time to list a company.
- [09:30]
B
Yep. All right, so let's, let's go into open air.
- [09:33]
A
Yeah. So, you know, I think the first thing to think about is just because this is a private company still today, that will then become a public company, you got to start off, I think, by looking at the distinction between what a venture investor looks at and what a retail investor should look at. So for us, especially as an early stage investor, first thing we're looking at is the founder, right. And do they have a vision that's worth believing in? And then if so, do they have the ability, the strength, the resilience, the charisma, the management skills, and so many other factors to actually turn that vision into reality? The second thing that we look at is the tam, the total addressable market. And what we're trying to figure out is, okay, what is the realistic total possible amount of revenue that a company could generate if it owned 100% of the market, which basically no company ever does, and then you back it out from there. And typically for us, at least, what we try to figure out is if they had 1% of the market, 2% of the market, can that get them into a multibillion dollar valuation? So if you have, let's call it $300 million in revenue, you're usually well north of $2 billion in valuation. It could be 3 or 4, depending on the sector that you're in. And then the third thing would be the team, the underlying tech. Are there other investors on the cap table that could bankroll future rounds? So, like for us on the investing side, we're writing three to $5 million checks, and then we do our pro rata. But that's pro rata is usually a fraction of that. But you sometimes often will need an investor that could put in 20 million for the next round, 40 million to the fund, whatever it is. And usually it's different VCs, but you want to know ideally that there's someone in the orb that could write that kind of check. If they wanted to and needed to. So that's typically what an early stage investor like us would look at. A later stage venture investor looks at all of the same factors I just mentioned. But then their economics are different because if you're us, we make our money on Kerry. So Kerry is for those of you who don't know, after we pay back the investors the amount that they put in, for every dollar going forward, we get 20 cents in profit, right? That's called carry. But if you are a multi, multi billion dollar fund, let's say you're a $10 billion fund and you're getting a 2% management fee, you know that's $200 million a year, right, that you're getting regardless of performance, year in, year out. Now you're paying all of your expenses out of that. But typically when you look at the hedge funding ventures, the private equity investors, some of the mega venture investors that are making, you know, nine, eight, mid eight, nine figures a year at least that's from management fees, that's not from carry. You might get lucky and hit something that makes you that kind of money. But typically speaking, so if you're them, your incentives are very different because you don't really care about carry. You care about assets under management AUM, because that's what the 2% fee is based on. And so your need is to deploy really large amounts of capital into deals so that you can invest that fund fully and then raise more money. Right? And so you don't necessarily mind investing in a valuation that's wildly inflated because you just need to get the money out the door. As long as, as various pension funds and sovereign wealth funds and endowments will just keep investing in you regardless, because you're a brand name. And their people working there just know that if they have an investment in Sequoia, they can't really get fired. Then the game is just really to collect more and more assets and deploy them kind of somewhat often, regardless of the underlying economics of the company itself. And so that's how venture investors look at it. But if you're a retail investor, what you at least should be looking at are the economic fundamentals of the company itself. So what's revenue? What's ebitda? What are the margins? Where are the opportunities for growth? What are the market risks? What kind of competition are you looking at? That's how your analysis should be based, not just based on potential. And because you're not getting the money, you know, if, if we lose money on our fund, we don't have to pay it back now, we don't get any carry in that scenario, but we don't have to pay it back. If you invest and you just throw money at something without thinking about it and you lose it, you don't, that's, it's your loss, that was your money that went out the door. Right. And so the market in my experience normally eventually correlates the fundamentals of a company with the share price, Tesla being kind of the exception. Right. And that's also why quarterly earning calls are important because they have to give you the fundamentals every three months. And I think they keep the market honest. And when you look at, I looked at IPOs between 2015 and 2025 of tech companies with valuations of $10 billion or more and kind of what I found confirmed what I suspected, which is after 12 months the median valuation was 32% below the IPO price and that was about two thirds of those companies. And then after 24 months it was also down 32% but now it's at 75% of the company. So oftentimes the first day is the best day and it gets worse from there. And that's one of the reasons that VCs like us have what's called a lock up. So for six months I typically can't sell my shares because otherw people like us know this and we would get out on day one. Right. And in order to prevent the stock from plunging because everybody would sell at the same moment, they make you wait six months. With the idea being that in those six months enough retail investors have come in that when the venture investors cash out, the share price shouldn't take too much of a hit. And so all of that then makes the sort of potential IPO from OpenAI. So the rumor is September. It might not be, but that's at least what's out there. Really interesting. So let me start with the obvious, which is in many ways OpenAI is an incredible company. They are at the forefront of the most transformational technology maybe ever. They have 800 million weekly active users. That's twice the population of the US plus they handle over 2.5 billion prompts a day. I don't watching their ads in the NBA playoffs are fantastic. I'm like totally inspired by them. But with all of that said, if you are a retail investor, so you're someone that doesn't have access to to the IPO itself, you haven't already invested on the venture side, you can't just look at this one based on revenues and multiples and margins and profits, because it's way too complicated for that and the numbers don't lend itself to that. So what it really means is you have to make a bunch of assumptions and a bunch of bets. And what I'm really trying to do in the substack is explain what those bets and assumptions you have to make and what it would take for them to be true. So if you assume that a reasonable return for a company with the risk profile of OpenAI, if you're a retail investor is 25% year over year, and I think that's totally a reasonable number, then these are the things that would have to be true. So the first thing to think about is if they go public on day one at a $1 trillion valuation, you don't have access to the stock on day one. Right. Those are the people who bought shares from the underwriters. And so a company like this probably has a first day pop of something like 30%. So that really means you're entering at a 1 trillion dollar valuation, not a trillion. Now, look, if it goes the other way, then all the math that I'm about to do adjusts accordingly down. But if you're going to earn 25% a year, year over year for three years, that means the valuation still has to go up to $2.54 billion by 2029. So that's basically a 2x from where it is today. So here are the companies that are currently worth 2.5 billion or more. Amazon, Alphabet, Apple, Nvidia, Microsoft. That's it. Those five companies, basically the five most successful companies in the world. You have to become one of them. They have an average of $400 billion in annual revenue and profits on average of over $100 billion. And let me give you Dell as an example. So on Friday, Dell announced earnings, and they were spectacular. Their servers and their infrastructure for AI have become an incredibly val product, far more than their, you know, PCs or whatever it is. Their market cap is $270 billion. That'd be basically 10% of where OpenAI has to get to. They expect to generate $165 billion in revenue this year. In 2025, OpenAI had $13 billion in revenue and lost $9 billion. Right. So 13 billion in revenue and it goes up for 26, but, you know, ultimately off of a $2.54 trillion valuation, $165 billion in revenue off a $270 billion valuation. Now, hardware is still very different than a platform, so they're not the same. But, you know, we're talking a tech company in the AI space, and that's effectively the baseline. So now we're talking about for open AI numbers that completely defy any version of reality. And the challenge is, if you have $100 billion or more in revenue, you don't trade at 54 times scale. So if you take the current run rate for OpenAI of $24 billion for 20, and then a $2.54 billion valuation, that means that you have to trade at 54 times your revenue. And typically, the kinds of companies that I just mentioned, the Apples, the Amazons, the Microsofts, the alphabets of the world, and I'll go through them, trade at much, much lower multiples. So Nvidia does trade at about 25, which, by the way, is still less than half of what we're talking about here. But they run a 50% margin. But, but normally speaking, the bigger the revenue base, the lower the multiple. So you can't say, oh, OpenAI doesn't need to get to a couple hundred billion dollars in revenue because they're going to trade at a 54x multiple. Because that's not how it works. So, you know, SaaS, companies that have really high multiples are typically working at 70 to 80% gross margins. And OpenAI is projecting through tearing through $665 billion in cash over the next four years. So we're not talking about an 80% margin. We're talking about massive losses. Right. You can't have the kind of capex that Open AI has, the kind of opex that OpenAI has, and get SaaS multiples. Now OpenAI's defenders will say, oh, well, there'll be a new agentic ecosystem model that will have a totally different type of multiple. But that's just like me saying that, oh, this podcast, Instead of having 10,000 listeners, will tomorrow have a billion? Like, sure, you know, so if you invested at a $1.3 trillion valuation, that means that revenue would have to more than triple from their projected 2026 revenue and to get 25% year over year. So that's just not to lose money. You got to reach $195 billion in the next three years at a 13 multiple, which is reasonable OpenAI's own wildly rosy projections, the number you would need to get that return would have to exceed OpenAI's projections by 56%. Right. So OpenAI is projecting $280 billion in 2030. Assuming that's true. And that's like, again, Us having a billion listeners or whatever it is, then the math would start to work. But think about it. Amazon, which is the company, when you say what's the company? That said, we don't care about margins, we just care about growth. We're just going to chase revenue and we'll figure it all out later. So in 2009 they had $25 billion revenues to about what OpenAI has today. You know, when they got to $280 billion in revenue, 10 years later, 2019, OpenAI said we're going to go from a run rate this year of $24 billion to 280 in four years. Right. And Amazon had a lot more business lines and they don't have anywhere near the kind of debt that OpenAI has. And you know what? Amazon trades at a 4x multiple. Now it's retail. The margins are a lot thinner, but still 4x compared to 54x. And then on top of that, OpenAI has $1.4 trillion in compute commitments. So it's not technically debt in the traditional sense, but it is contractually obligated spending. So it's pretty similar. They also spend, plan to spend, as I said, over 600 billion in compute operating expenses. And the two figures aren't additive, they're not interchangeable either. So you don't have to necessarily take that 1.4 trillion and stack it on top of the 2.54. But the debt holders have priority before the retail investors do. So it's not really 2.54, it might not be 3.94, but it's somewhere in between. Right. So if you did add the 1.4 trillion, now even if they got a 15x multiple, they've got to get to $260 billion in three years for you to make your return. And compare that to the 15x. Alphabet's got 12x, Microsoft 11, Apple 11 meta 9. And if you take the number of companies that have $400 billion in revenue, which is what it would take if Amazon were at Open Air, were at say a Microsoft or an Alphabet multiple, there's only five companies in the world that have that kind of revenue. If you put aside the state owned enterprises by China and Saudi, Walmart 64 years old, Amazon 32 years old, Apple 50 years old, Alphabet 28 years old, and UnitedHealth Group, which is sort of barely at the 400 billion number, 49 years old. Now we got a company that's a fraction of that. Having to achieve that kind of revenue in the face of massive losses with no real path to that whatsoever. So that's number five. Number six is competition. It's not like OpenAI's operating SpaceX. One of the reasons why at least you could sort of squint and maybe try to make a case for their valuation is in Blue Origin. But no one's really quite doing what they're doing. Right? And starlink is exceptional and rare. You have anthropic. So they have double already the revenue of OpenAI. They don't have anywhere near the $1.4 trillion in compute commitments. And the reputation is exponentially better than Open AI. Open AI.
- [23:59]
B
Do you think that's deserved?
- [24:01]
A
You know, I think they're smart, right? How much of it is real and how much of it is really good pr? I don't know. You know, there are people who know them. Dario would say that like it's legit. They genuinely do seem to care about the greater good. With that said, I don't know. But OpenAI between Sam Altman, there's that New Yorker. Do you read that profile? New Yorker basically had board members calling him a pathological liar. They said they're going to be nonprofit, and they flipped it. They're trying to take out Alex Boris, who's running for Congress here in New York, and they're spending millions and millions of dollars in ads against Boris. It has propelled his campaign. That's how much people hate OpenAI. Philanthropic, with more revenue, less debt, and a better reputation. You have Gemini, owned by Google, and then you have Grok, owned by Elon Musk, who's not only the richest person in the world, but he fucking hates OpenAI. Right? So you have to compete against all of those companies, too. And all of them have, you know, arguably much better fundamentals than you do, and then regulation. So AI itself is really unpopular. There was an Annenberg PEN poll from March 17% of Americans that said that AI will have a positive impact on the US over the next decade. 42% said it would be negative. Two thirds said that government's done too little to regulate AI. I'm sorry? Not just OpenAI. Another poll that says that the number of Americans who fear that AI could eventually threaten humanity is over 75%. And the reason why this is so relevant is it's not just like idle political speculation. For OpenAI to sell its products, it has to have data centers. Data centers require zoning and permitting approvals so they can't happen in a vacuum of regulation. They require the government to allow it. And when something is as wildly unpopular as AI, it is much, much harder to build data centers. If you can't build data centers, you can't sell your product.
- [26:00]
B
Well, I just, you know, you say wildly unpopular. Obviously true in this political context, but not in a consumer conte. The what explains those?
- [26:10]
A
Well, I think there's a lot of things where like we know something is bad for us. I think people now would generally say that social media is bad for us and most people still use it.
- [26:18]
B
Right.
- [26:18]
A
There's a lot of things, you know, there's a massive obesity Crisis in the US my guess is most people eating at McDonald's would say I shouldn't be here right now. But they are.
- [26:26]
B
But they don't hate McDonald's.
- [26:27]
A
No, but. And I don't know if they hate open. I look for polling kind of company based. I'm sure that they have it internally, but I couldn't find any sort of. They might not hit OpenAI, but they are just very, you know, in this case it's not even hate. They're frightened. And I think you could put that not just around the arguable risks of AI, but it's also just the pace of technological change. Moore's law says that effectively the rate of technology growth doubles every two years. Human evolution is incredibly slow. Species evolution is incredibly slow. Homo sapiens in our current form are somewhere between 60 and 100,000 years old. And that's like nothing. You know, you have sponges that are trillions of years old. Right. So it's nothing. And at the same time, our world is changing so fast. I mean, if you think about it, if you look at kind of population growth and technological growth from sort of that 60,000 years ago till about 250, 300 years ago, maybe 400, let's call it the Industrial Revolution started in, let's call it the late 1600s, I guess, guess early 1700s. Basically we're only at a billion people. Now we're at 8.2 billion people. And everything has basically skyrocketed in the last couple of hundred years. So even in the context of 60,000 years, a couple hundred years is nothing for a species that's relatively new to begin with. And so I think the reason why you see so much resistance to change, whether it's the antipathy towards technology by the left or MAGA on the right, is that people just feel like the world is moving too fast. They can't keep up with it. And when they become afraid, they become angry and they start to hate things. By the way, this is all before AI really has a massive impact on job displacement. Right? So it's had some already. But Goldman Sachs, which obviously wants to see AI do well because that's how it makes more money, is through underwriting AI IPOs, they're saying that somewhere between 2.5% of all workers in the US will be laid off, up to maybe 6 to 7%. And that's the most optimistic estimate out there. 41% of working Americans say they're worried about losing their job or having their hours cut because of AI. And here's the challenge. For OpenAI to have the kind of revenue it projects, it needs companies to lay off incredible amounts of people and replace them with OpenAI technology. But the more and more people who get laid off, the more and more regulations you're going to have against AI, which then means the underlying ability to even offer their product because they need things like data centers are going to become harder and harder to build. And so they're in this sort of catch 22. And look, if there's one theme of this podcast, it's that politicians respond to whatever impacts our next election, right? And when voters hate something this much, politicians do, too. And that means that the regulatory environment that OpenAI is seeing today is the most permissive it's ever going to get. And now imagine a Democrat winning in 2028, which the prediction markets would say is almost extremely likely to happen. They're probably gonna win amid a rise of anti AI sentiment. So you're gonna have, right now, the federal government really hasn't. Trump talks about AI, but haven't actually done anything, right? You're gonna start to have really tough federal regulations. You have a Democratic presidency, Senate, and Congress. You're gonna start to see really tough AI regulations at the federal level, on top of the state level, on top of the municipal level. So all of that, fundamentally, if you're a retail investor, has to be taken into account because it's going to crimp OpenAI's ability to sell its product. And the final thing is geopolitics, right? So look at what's happened to the entire global economy just because one critical choke point, the Strait of Hormuz, has been closed off right now. Instead, China invades Taiwan. 90% of the world's semiconductors, advanced semiconductors, come out of Taiwan. And while the US Is trying to build capacity, we're way behind best case scenarios. They're at like 17 to 20% of global chip production in the early 2000 and 30s. If China invades in the next year or two or three, it's going to be really hard to get chips made. If you can't get chips, Nvidia can't get their chips made. OpenAI doesn't have the computer it needs to operate, in which case there are no sales. And now layer in the unpredictability of Trump. A never ending Russia, Ukraine, war, the safety risk of AI itself, worsening climate change, this proliferation of nuclear weapons, falling cost of assembling bioweapons, and a world just generally drifting away from cooperation and towards zero sum nationalism. You put all that together and like the odds of none of that going wrong is like incredibly low. So if you're a retail investor, you have to assume so many things go right. You have to assume that OpenAI has massive unprecedented revenue growth of at least 10x in a couple of years, which is faster than any company ever. You have to assume that a new agentic ecosystem category with new multiples will come into existence, which defies how the rest of the market works. You have to buck the trend of major new public companies trading below their IPO price in the first few years. You have to bet that they can service a billion foreign capex, 600 billion in cash burn and still reach profitability. You have to assume that they can do all this in the face of unrelenting well funded competition that's either better funded, better liked or incredibly powerful. You have to believe that public sentiment around AI will flip on its head rather than only getting worse. Even though the job loss necessary for OpenAI to hit its revenue targets has to happen and which means that public sentiment will only grow worse. You have to regulation at every level of government, every part of the world will subside rather than increase. You have to believe that all the data centers needed to reach the revenue targets can be built as quickly as envisioned. You have to believe that all the chips needed for compute will continue to be readily available and that nothing else terrible happens geopolitically. So look, maybe somehow OpenAI hits all 10 of these things and they do go up, but even if they do all of those things and they reach that $2.54 billion valuation, or really more because of the CapEx, you're making 25% year over year, but the risk you're taking is losing all of your money. And once it goes public, you can buy the stock at literally any time. So why would anybody rationally jump in on day 1 or day 10 or day 100 and buy open AI shares on the open market when you can just wait to see how all 10 of these different risk profiles materialize and then decide. So it just.
- [32:55]
B
To me, I mean, the reason people would do that is because it goes up 10x and like, you know, you missed your shot, right?
- [33:02]
A
So to go up 10x, that means it would have to be a, somewhere between a 10 and $13 trillion company. That's 5x any valuation pretty much of any company in the world right now, 3 or 5x. Let me put it to you, that makes no sense.
- [33:14]
B
Let me put you exactly like I did on the, on the Knicks, right? I'm going to give you $10,000 on the, on the, on the opening day price, you could go short or long.
- [33:24]
A
Short it.
- [33:25]
B
You take your $10,000 and you short it.
- [33:26]
A
Yep. Okay, maybe I'll hedge a thousand.
- [33:29]
B
Someone comes to you and says it, I'm with the future, I'm just going to invest, you know, do it.
- [33:32]
A
Yeah.
- [33:33]
B
Do you say you're an idiot or you just like, that's just one view. And you know, look, there are people,
- [33:38]
A
you know, usually I'm a techno optimist, right. My entire job is based on, well,
- [33:42]
B
the weird way your analysis is techno optimistic. It's not, it's not, you know, it's not like, oh, the future is going to be scary and apocalyptic.
- [33:50]
A
No, not at all. But I would send them my substack and say read this before you invest. And then if they want to, I don't give a shit. By the way, I don't short anything, so I don't care how OpenAI does. I'm just thinking about this, is that if I were a retail investor, I think that what happens is people, retail investors, they read blogs, they get excited, they get sold by people who frequently have a financial incentive in the first place. Pump a stock and then they throw money in and it's dumb money. And I just think that if you were to be an early retail investor in OpenAI, that's dumb money.
- [34:29]
B
I guess one of the big things, one of the criticisms of the SpaceX sort of IPO is that the float is really small. So it's like 5% of the company, I think is going to be sold and there's obviously going to be tremendous demand for a relatively small number of shares. And I guess that's the kind of thing that a make a big.
- [34:47]
A
I mean that's the thing that can help when you have that high of a valuation. 5% is still a huge amount of capital. Right. We typically speaking, companies go public for two reasons. One, to give liquidity to the pre IPO investors and two, to generate capital that you could then use to run the company and build the company and everything else. But the difference is SpaceX has already figured out how to build their rocket ships or satellites, everything without public money. You know, without, you know, the market. OpenAI's commitments are so much bigger that even the float is not going to be enough to pay for it.
- [35:25]
B
I love that out. I think it was great. I like that you come down with a firm conclusion too. And I think the analysis is excellent.
- [35:32]
A
Again, not investment advice.
- [35:34]
B
No, no, it's an exercise. It's a way to think.
- [35:36]
A
Yes.
- [35:36]
B
Let's talk about your walk. Was it Sunday?
- [35:38]
A
Saturday.
- [35:39]
B
Saturday. Why'd you do it? Explain the whole thing.
- [35:42]
A
So my friend Kenny Kwaku and I have been just for like a year now.
- [35:46]
B
Now where do you know Kenny from?
- [35:48]
A
School. His son K.J. was in Lyle's class and then they moved there for a couple years. Yeah, kid's school. I'm sorry. Oh, from, from our. Yeah, the school. And then they moved up to Westchester. But Kenny and I stayed friends. So we've been friends ever since. And I don't know, somehow this idea one day just came up of like, hey, wouldn't it be fun to walk the entire length of Broadway? And we kind of, you know, it was one of the. Talked about it. Talked about it.
- [36:12]
B
Not the entire length. Well, just the New York City length
- [36:14]
A
does keep going all the way up like Dobbs Ferry, apparently, yes.
- [36:17]
B
Wait, does it go to Albany?
- [36:19]
A
I don't think, I don't know, maybe. I'm not sure. But anyway. Well, we decided was originally the plan was to do Manhattan. So from the very tip in Inwood down to the Battery. And then because Kenny grew up right by like Van Cortlandt park, he was like, why don't. Can we start 2 miles before that at 2 42nd in Broadway so we can kind of start my own neighbor. I was like, yeah, sure. So I met him up there. Saturday morning weather was kind of perfect. It was like 60 was drizzling when we started, but kind of almost the minute we started walking it cleared up. And so it wasn't hot, wasn't cold, you know, for walking weather. It was perfect. So we started up at 242, saw a couple landmarks. The building Kenny grew up in, he pointed out the 50th Precinct. And that's where the phrase 5o comes from.
- [37:09]
B
Oh really? Yeah.
- [37:10]
A
That's interesting.
- [37:12]
B
Wait, is that, that's not Fort Apache though.
- [37:14]
A
No.
- [37:14]
B
Is there a precinct that's associated with Fort Apache?
- [37:17]
A
I don't know. Yeah, probably, but I'm not sure. Then, you know, we kind of saw that part of the Bronx crossed over the bridge into Manhattan, kind of hit a really cool coffee shop in like the 190s. And it was a good. It was interesting because it was. It was a heavily black crowd, but also a big white hipster crowd in what is becoming sort of a hipster or become a hipster neighborhood. Then kind of walked down into Washington Heights. That's always pretty interesting. And then kind of went pathway. Borico College pointed out Columbia Business School, where I used to teach. Hit Columbia, had lunch over there. Then kind of upper, kind of lower Harlem, upper West side, Columbus Circle. Then Times Square was a fucking madhouse. Obviously, like worse than usual. I don't know. It was just crazy. You know, I avoid Times Square.
- [38:15]
B
It's terrible. It's truly awful to spend two minutes walking through it.
- [38:19]
A
It was terrible. I mean, it was amazing in the sense that like I got really good pictures. Cause it is visually spectacular.
- [38:24]
B
Humanly chaotic, but.
- [38:25]
A
Yeah. And then. And then you unfortunately get a few blocks later into Heralds Court, which is also terrible. And someone's even worse because is Times Square is at least visually stimulating. Herald Square is ugly and packed. Right.
- [38:37]
B
Why shouldn't Herald Square be a bigger.
- [38:39]
A
You know, they've tried, like, they have like nice pedestrian areas and stuff. But why anybody would choose to sit there, to me is insane.
- [38:44]
B
Yeah.
- [38:45]
A
I don't know. It's just ugly.
- [38:46]
B
Yeah.
- [38:48]
A
And then you kind of nomad Madison Square Park. And then Union Square was crazy too, because of the green market. So that was a mess.
- [38:55]
B
A mess.
- [38:56]
A
I mean, it was just like, I don't like walking, you know, I'm a super fast walker. So when everyone's fucking slow, it drives me crazy.
- [39:04]
B
Did Kenny keep up with you the whole time? Is he a fast walker?
- [39:06]
A
Not as fast as me. He was sweating a little bit, but he kept up. Village kind of went right past your house into Soho. We actually did pivot off Broadway briefly. Cause when we hit Bleecker, we went out to Crosby. Cause we were gonna stop at my place to use the bathroom, have some water. And Broadway.
- [39:25]
B
Is that allowed under formal rules?
- [39:26]
A
Nah, you know, I'm a commissioner. I'll allow it. You know, Broadway in SoHo is horrible. Right. So avoided that especially on Saturday. And then the last stretch was just a mile and a half. So kind of, you know, cross canal through kind of past City Hall. And then you got into kind of World Trade Center, Battery Park. We ended at the Staten Island Ferry.
- [39:50]
B
Was there anything you saw that, like, was, like, sort of stuck in your mind as, like, being different than, you know, you obviously live in the city.
- [39:57]
A
Not really different, but, you know, the. The part that I enjoyed the most were the first hundred blocks at 242 to 142, let's call it.
- [40:03]
B
Yeah.
- [40:03]
A
Simply because I'm not up there that much.
- [40:05]
B
Yeah.
- [40:05]
A
Right. So everything from kind of like 130 on down, like. Yeah.
- [40:10]
B
Pretty familiar.
- [40:10]
A
Yeah. You know, anywhere from, like, my. Where my office is, where my apartment is to just, you know, the normal Manhattan that I tend to live in. So the other part, which is what I really miss about working at the Parks Department. So you really would cover the whole city.
- [40:23]
B
Yeah.
- [40:23]
A
Or just working in city government campaigns. You do. When you kind of aren't in that world, you kind of settle more into. I would say I'm more aggressive or about using the entirety of the city and going places and all the different boroughs and all that than probably a lot of Manhattan residents. But nonetheless, there are still big chunks of the city that I never get to. And so that was sort of, for me, the best part.
- [40:47]
B
How much better is the walk now than when, say, you first moved to New York as a professional person? So if you'd done that walk in the 90s,
- [40:58]
A
you know, in. When I finished college In December of 94, City was getting safer by then, but I would say, you know, you would have been a little more tentative on that kind of first part of the walk probably than today, where you wouldn't. No one thought twice about it.
- [41:11]
B
Right.
- [41:13]
A
But then once. But, no, not. I don't think it's meaningfully different. I think the difference is, in some ways, it's more built up because there's just been, you know, growth in a lot of ways. You know, place like Washington Heights is sort of, I think, kind of the same. Like, they're different commerce, but just. It's always been super vibrant. Right. But also, in some ways, because of the Internet, there's a lot more empty storefronts. So a lot of small businesses probably haven't fared as well.
- [41:42]
B
Yeah, I mean, that's the one thing. Like Broadway on the Upper west side, like, in the 70s, it's amazing how bad it is. Like, I mean, those. That was, like, such a. Such a great stretch when I was a kid. So many interesting stores and things, and now it's like, a lot of vacancies, and like, they just sort of suck
- [41:58]
A
the life out of It. Yeah, it just feels. What's funny? I've never been an Upper west side fan.
- [42:03]
B
Yeah, me neither.
- [42:04]
A
And in part, like, I don't like the whole vibe in general, but in part, yeah, I find it to be depressing and kind of lifeless. Like, if you're on Central Park West, I get it. Maybe if you're all the way over on the other side by Riverside Park, I get it. But in between, to me, it's like, I don't know, I would want to be here. And I just generally, to me, like, the reason to live in New York is for the energy of New York, and there's just more of it downtown than anywhere else. I mean, you live downtown, I live downtown. And, like, I just don't. I mean, live wherever you want. I don't give a shit. But, like, to me, why would I live here, put up with all of the negatives of living here, and then live in an area that's kind of lifeless?
- [42:43]
B
Yeah, yeah, I don't like it much either.
- [42:44]
A
And in fact, at least between the Upper east side and Upper west, at least the Upper east side's kind of nice. The Upper west side's, like, not even that nice, and it's kind of lifeless.
- [42:51]
B
There are two quick questions I have, and then we'll get your recommendation. Do you want to do a ticker
- [42:54]
A
tick tape thing or no? Yeah, so this is just a little fun thing. So I wrote a quick substack last week about if the Knicks were to win what a Mandani style, which you
- [43:04]
B
have, you have picked them to win. Just to remind Everybody would look Nixon 6, says Bradley.
- [43:10]
A
So these are 10 things that I think would be specific to a Mandani style tipper tape parade. So the parade would be in Bushwick rather than the traditional Canyon of Heroes in Lower Manhattan, because that, of course, abuts Wall street, and that would be like the ninth circle of hell to Mondani. And the bus fare to the parade will be free as long as Albany pays for it. To the worst players get to speak first. Normally it'd be like Brunson and Kat, but, you know, to the left. Fairness is the number one thing. And so, like all the Scrubs. So Tre Jemisin and Pacom Dotier and Kevin McCuller get the mic and then a few of the Nets get to speak, because that's only fair, right?
- [43:45]
B
What about the liberty?
- [43:47]
A
Yeah, probably throwing some liberty. But they're good, though, right? Aren't they good? So I'm not sure they'd be allowed. They're too good. Celebrity rogue gets moved up to the 400. Chalamet's on its own. Fend for yourself. All Knicks merch must be manufactured by bearded craftsmen in Brooklyn. Prices are frozen forevermore for artisanal beer and MSG security is replaced by violence interrupters. Number four, players on the team that win the NBA championship. You do get about $900,000 bonus. Mondami will announce a 90% tax on those winnings as well as a new wealth tax for all season ticket holders. Five, the players. Instead of riding down the Canyon of heroes or Bushwick, this case on a float or convertible. Everyone has to ride a bicycle. Ken Griffin, though, will get his very own Schwinn as an attempt to mend fences. And then Maidani will use the ceremony to unveil a fully funded protected bike lane connecting Williamsburg to absolutely nowhere that anybody needs to go. Number six. The Knicks will become public property owned by the people of New York City, except Staten Island. Privately owned teams is a bachelor of capitalism, so Jim Dolan's out. Number seven. The Knicks locker room is turned to a publicly owned grocery store. And you know, construction shouldn't run more than like 50 million bucks with taxpayers, so the Knicks have to share stalls with visiting lockers. But they do get free childcare during the games. Number eight. After the parade, Mondamba will lead a protest outside NBA headquarters on fifth Avenue just because. Number nine, rather than letting the team choose the best players they're offering for next year, will reflect the full diversity of New York. Nominated by a DSA working group and voted on after a long and divisive caucus meeting followed by an amusing Instagram post by the mayor. And this is definitely the best news Tyler Kolak has heard all week. And 10th. And finally, anyone who hops a ter style to enter the Garden will now be allowed a free seat and a chance to win $75,000 by hitting a half court shot, which they can instead take from the top of the ark.
- [45:37]
B
Very nice, Bradley. Let's just do your recommendation.
- [45:39]
A
What do you got? I got a couple. I know you don't like this, but
- [45:44]
B
I'm sorry, I want you to mention one thing just before I do the recommendation. Just because the party for the for the Gotham Book Prize was last week and it was really great and, and
- [45:52]
A
I don't know if you saw this, but did you see that Daily News yesterday ran McNally a two parade spread of McNally speech? No, but I'm glad that you mentioned that.
- [45:58]
B
I didn't see it.
- [45:59]
A
Yeah, they did it was excellent. Yeah.
- [46:00]
B
Okay. Well, that gives me a chance because I missed a speech, but everybody was talking about how great it was.
- [46:04]
A
Weren't you there?
- [46:04]
B
No, I did come, but I just got there late and just given the speech.
- [46:07]
A
Oh, I'll turn to you. Yeah, he gave a great speech. Yeah. So Gotham Book Prize went, as mentioned before, to Keith McNally for his memoir, I Regret Almost Everything. I thought it was a great book. He donated the $50,000 prize to the city, the mayor's fund, and to the Museum of the City of New York.
- [46:22]
B
Oh, my God, that's great.
- [46:23]
A
Yeah. Incredibly nice of him. He gave a wonderful speech that you could find on the Daily News, and
- [46:28]
B
it was a wonderful event up at the city. Great event in the city of New York.
- [46:30]
A
City of New York. This is perfect. They did a great job. It was far and away the best Gotham Buck Prize event we've ever had. And it does feel like now that we've hit year six, Megan pointed out to me, the books that get nominated now, they publicize it. Like, it's on the book, it's on the website. I looked at some books, some writers like Wikipedia page. Gotham Book Prize nominee wasn't even a winner. It's kind of becoming a thing a little bit, which is cool. That's great. Yeah, it's nice.
- [46:57]
B
Awesome. All right.
- [46:58]
A
Recommendation three. Sorry, Three. Yeah, I know. Sorry.
- [47:02]
B
Go ahead.
- [47:03]
A
One, a novel called A Violent. I think you would like a lot called A violent masterpiece by a guy named Jordan Harper. And it's just totally in that tradition of LA crime. Very modern, not noir at all. Very Internet, you know, really kind of almost the depravity of society, but I thought an excellent book. Two, and this was a movie that a lot of people have seen or did incredibly well, but I happened to just see it, which was. Did you see Project Hail Mary?
- [47:25]
B
No, but I want to.
- [47:26]
A
It's fantastic.
- [47:27]
B
I know. I keep meaning to go see it.
- [47:28]
A
It literally was the kind of movie where, like, I laughed, I cried.
- [47:30]
B
Like, I laughed, I cried.
- [47:31]
A
It was great. And then there's a really stupid show on Hulu called Deli Boys about these, like, drug dealers, bodega in Philly. And the season two just came out and I. I just tore through it.
- [47:41]
B
You love it.
- [47:41]
A
So anyway, those my recommendations.
- [47:43]
B
Thanks, brother. Thank you.
- [47:46]
A
Firewall is recorded at my bookstore, PNT Netware, located at 180 Orchard street on the lower east side of Manhattan. We'd love to hear from you with questions, feedbacks, or idea for a guest. Just email me at Bradley Firewall Media or find me on LinkedIn. And to keep up with what's on my mind in my latest writing, please follow my new substack@bradleytus.substack.com thanks again for listening.