Limitless: An AI Podcast - The World’s Best AI Investor is 23 Years Old (Here's What He Owns)

Episode Date: November 18, 2025

23-year-old Leopold Aschenbrenner transformed $1 billion into $4 billion in less than a year, driven by his predictive essay on AI trends. We highlight his unique investment strategy, including a key bet on Intel and stakes in AI compute firms like CoreWeave. Join us for insights on whether his success is due to luck or strategic insight, offering valuable lessons for aspiring investors!------🌌 LIMITLESS HQ: LISTEN & FOLLOW HERE ⬇️https://limitless.bankless.com/https://x.com/LimitlessFT------TIMESTAMPS0:00 Leopold Aschenbrenner2:43 From Researcher to Fund Manager3:19 The Billion-Dollar Fundraising5:12 The Intel Trade6:08 Portfolio6:37 Investing in NeoClouds8:59 Core Scientific: A Risky Bet12:55 CoreWeave and Strategic Moves14:21 Analysis20:45 The Risk of Timing24:11 Conclusion------RESOURCESJosh: https://x.com/JoshjKaleEjaaz: https://x.com/cryptopunk7213------Not financial or tax advice. See our investment disclosures here:https://www.bankless.com/disclosures⁠

So you've heard of Warren Buffett, you've heard of Ray Dalio, but I bet you haven't heard of Leopold Ashenbrenner, a 23-year-old ex-open AI researcher who raised $1 billion and turned it into $4 billion in under a year. His secret? A 165-page essay which predicted the future of AI over the next decade. And he got a lot of these predictions correct so far. He predicted spending trillions of dollars on the AI Cappex bubble, Open Ares, checked that one off. He predicted GPUs becoming as scarce as gold, which is pretty much what we're saying. He predicted constraints on the energy grid, which is exactly what we're saying today. But he put his money where his mouth is.

He made some pretty crazy trades, one of the most famous ones being buying Intel before the US government took a 10% stake in it. And the recent 13F filings of his fund just got released last week. And it reveals his latest newest trade, which might make him billions more. Josh, have you heard of this Leopold guy? This Leopold guy is unbelievably impressive. It's like very few times that you come across someone and you see something that's like, oh my gosh, this guy's amazing. And then you see it over and over and over again.

And I kind of want to tell the story that I read earlier this morning when I was preparing for this, which starts in 2023. Because Leopold, he was fresh out of Columbia University. He shows up at Open AI in 2023. And he's the weird young guy on the Super Alignment team. Now, you might know some other names on the Super Alignment team. One of the most noteworthy ones is Ilius Setskever, one of the largest names now in Crypto, who was one of the co-founders of Open AI, they had a falling out.

And basically what happens with Leopold is in 2023, there's a problem at Open AI. And Leopold writes a security memo that he shares with the Open AI board saying, hey, we're building this really unbelievably important technology that should be private because we are at a high risk of espionage from China. And that wasn't received very well from the Open AI Board of Directors because less than one year later, he was fired from the company. What did he do? he left, he went to the drawing board, and wrote out a 165 page paper called situational awareness,

the decade ahead. And within that, he wrote about AGI in 2027, he wrote about the path from AGI to superintelligence, and he wrote about the trillion dollar compute cluster, and the power scramble that's going to happen in pursuit of that. Now, this memo went absolutely bonkers on the internet. Everybody read it, every top CEO in the world of AI and technology, they all read it, and they all said, who is this guy? I want to do. invest some money in him. So Leopold goes and raises $1 billion to start an AI fund. And that's where I'm going to pass up to you, EJS, because that's where the really fun story begins, is he went from super alignment to fund manager in the matter of, what, less than two years. And now this fund is blowing

everybody else out of the water. So the cool thing for me is that he predicted AGI in 2027, but not only that, he told you exactly how to invest in AI in order to receive the upside that his fund is getting. So that's kind of what we're going to go through now in this episode. Yeah, it's, it's, it's so impressive that he's probably lived multiple people's dreams before he's even hit the age of 25, which is just that you can say that you can say for most people. The other thing is he, I must emphasize this, he's had zero experience trading before he raised this fund. He just casually raised a billion dollars, which just kind of shows how convincing his thesis, situational awareness was. So I kind of want to dig into like him raising,

a billion dollars, Josh, and then what made him like one of the most successful traders of this time? And then we're going to dig into like his latest trade. And I know we're excited about that. So after he raised a billion dollars, people were kind of like, well, what's he going to do? Like he's kind of predicting. So fast forward, sorry, rewind about nine months ago. He was claiming that like AI CapEx was going to reach trillions of dollars. And therefore, a lot of these compute generating companies, these are GPU companies like Nvidia, but also chip companies or competitors like Intel, and also things called Neo Clouds,

which are kind of like AI compute service providers. Think AWS, but they hyper-specialize in AI compute called Neo Clouds. He was investing in a lot of these companies, which people were like, this is kind of crazy, dude. CapEx is never going to get to this much. You're crazy. You're just a kid. You don't know what you're talking about.

Fast forward three months from then, Josh. And we've already talking about hundreds of billions of dollars raised. We're talking about Microsoft, Meta, Google, investing hundreds of billions of dollars, Elon investing 100 billions of dollars in their own data centers. And now, of course, we're still going to call weave on a previous episode, them benefiting so much from signing deals with the Microsofts and the open AIs of the world. And that brings me to the first trade that he absolutely nailed,

which was investing in Intel. So originally, his fund took a massive position in Intel. And back then, Intel was kind of considered as the dying competitor to Nvidia. and they kind of needed a saving. They needed some kind of a life-saving ramp for them. And that came in the form, Josh, of the US government, which decided to take a 10% stake in Intel and sent their shares skyrocketing,

but most importantly, it sent Leopold's funds, call options into the money massively, such that it brought it up 50%. So it went from a $1 billion fund to a $1.5 billion fund in under three months. just insane. AGI is coming in 2027. So how do you take that thesis of AGI coming in 27 and then turn it into investable positions? He's not really a stock picker who later discovered AI. He's kind of like an AGI timeline guy who asks like, okay, well, if my view

is right, where are the forced winners and the forced losers? Whereas typical investors, they're kind of like, well, what's cheap? They're looking at PE ratios. Leopold's like, well, what's misprice because the world itself is about to change categorically and the rest of the world does not agree with me. And that arbitrage is where we're really seeing these big improvements in his portfolio performance. And you could kind of map it directly to his portfolio. He had like a few bottlenecks. One of them was power.

One was compute hardware. And one was like secure well-governed labs, open AI being not one of them, which is part of the reason why he had a problem with them. And power kind of mets to bets like, maps to bets like utilities, generators, Bitcoin miners, things like that. And then we have compute hardware, which maps to chip makers, experiment makers and like upstream materials type stuff for batteries and hardware like that. And then we have the secure well-governed land labs.

which kind of translates to the elite AI labs like Anthropic and Google. And you kind of see this laid out through the portfolio. So could we get into the portfolio and kind of talk about more of these positions? We had Intel, which was a huge success. But he has like every single position is a success. So what else is there in the portfolio that we should talk about? He has these positions in these things called neoclouds. So you see two tickers on the tweet that I'm showing right now.

One is called Iron, which is up 6x year to date. This was like back in October 11th, right? So not too long ago. Then you have CallWeave that was up three and a half X year to date. Now, the best way to think about these two companies, Josh, is that they are, like I said earlier, AWS, but supercharged and specialized towards AI compute. So a company that would engage services with these types of companies would be Microsoft or a Google or a meta or an open AI who want to ramp up compute to train their frontier models, but maybe don't. want to invest all the money themselves and setting up the data centers and running it themselves, hiring the technicians doing all the wiring, they would rather just kind of pay a provider to do it all

for them. That is iron. That is core even. That's what their main kind of like unique offering is. And so he was like, he made the simple bet, which was simply like, okay, well, demand for compute and GPUs isn't going to die anytime soon. In fact, it's going to probably escalate pretty drastically. And they're not going to want to do it all themselves and take on all that risk. So they're going to invest in hyper-specialized providers. Now, back then, Josh, when he made this bet, that was seen as an insane thing to do. Today, when a call we've recently just signed a $19 billion partnership with Microsoft and they did the same thing with opening.

By the way, all those companies I mentioned have actually gone ahead and done this. They've committed hundreds of billions of dollars over the next couple of years with these neoclouds. It seems like the most obvious bet ever. The other thing that I think about Josh, when I'm looking at this portfolio is, has he made, in your opinion, a pretty drastic bet here? Like, he's betting on the raw materials. He's betting on the supply chain, which is probably the obvious thing

if you think that AI is going to be the most demanding and transformational technology going forward. Am I crazy? No, I don't think you are. I think he was early and he saw it with clarity and he went all in.

And those are three very difficult things for the average person to do. We've been sitting here telegraphing the future every single week on the show. And still, we kind of lacked the foresight to actually go all in ourselves. Yeah. Like, a really interesting example

is this company core scientific because it's known as a risky Bitcoin mining company. And this is kind of a testament to how he thinks. Because from Leopold's vantage point, he doesn't see a risky Bitcoin mining company. He sees that they run massive power-hungry data centers, that they have power contracts, land contracts, interconnections between all of the two. And if AI soaks up all the demand and spare power, well, miners who are doing Bitcoin can pivot to become weirdly valuable in this world of AI. So he's just kind of seeing this holistically. He's like, okay, we need energy, we need infrastructure. Where does that exist? Where is it improperly

applied to and then how can I how can I profit off of that and so far is he down on any position I think every one of these is up significantly this is not by a small margin either over four billion four billion four billion four point one okay can we talk about that number two because that implies he has multiplied his money by over three X we talked about the intel one where word on the street he's made a billion dollars in a single day from the intel trade which is like which is what do you do man so what else what else is here okay so um I mean to put that into context Josh that makes him the number one hedge fund trader in the world right now. Because it's not like he's trading with a couple million dollars either, not even tens or

hundreds of millions. It is the billion market gap. And this is a 23-year-old on his first try ever doing this. First try ever trade it. It's freaking insane. And he's doing this off an essay, which is like even the best part. Like he had this envisioned out. Josh, you just mentioned a company that he's invested in called Core Scientific. I have a crazy tale to tell you about this company. Are you guys ready? Okay. Give me some more. Okay, okay. This is crazy law over here. Okay. So last week, we had an episode where we basically talk about Michael Barry. He's the guy from the big short that basically shorter the entire real estate market. In 2008, made a heck ton of money. And he came back last week and he was like, I'm shorting companies like Corweave because I don't believe there's enough GPU demand. And the company's stock prices are way inflated. Turns out he was massively wrong. And the day after he made that tweet, his fund shut down, Josh. But he had some kind of sensibility around.

his thesis, right? Because that week, Cor Weave's stock price dropped about 10%. Even though it was up 5.5x this year, it was down 10%. So he was like, this is the bubble popping, right? And if you looked into CoreWeave's earnings reports, which was also released that week, it was revealed that the reason why the stock dumped

was because they had a third-party provider that they depended on to deliver the compute that they were trying to scale up, Josh. So I think the numbers were, they promised 250 megawatts of power, but they could only end up delivering 150 megawatts for the rest of the year because this third-party service provider wasn't able to kind of convert its infrastructure into AI GPU supply.

Guess the name of that company of that third-party service. Oh, it's not core and scientific, is it? Oh, you bet. You bet. So now let me bring your attention to this tweet that we have up on here, Josh, because it revealed on Friday last week, situational awareness fund, which is Leopold's funds, recent 13F finance, which reveals the positions that he's taken,

a brand new trade. Josh, he is long, $563 million worth of CoreWeave stock. And if you notice just underneath that, almost $400 billion in a company called Core Scientific. So this is basically a massive, almost $1 billion bet on CoreWeave to be the leading NeoCloud provider. He's giving us all the signs and signals that this is the company, all the two companies, to back. This is fascinating too because I recently, I was watching the show Billions where it's about this guy, this big hedge fund guy,

and he runs a private fund and he sometimes does hostile takeovers of company in order to get his way with things. And that very much feels like that's what Leopold is doing. I think he owned 6.8% and now it's up to 9% of the company, which means that he likely gets some sort of seat at the board or some sort of say in what happens. And he could actually control the destiny of these companies that were previously, for Bitcoin mining rigs, moving them over to the AI world and working together with CoreReave. And it's this self-fulfilling prophecy. And it's funny because you cannot possibly be more in. He's hedging with the risk that is going to merge together into even more risk. But he has a hands-on view on how he's doing this and how he's operating this.

So it implies that there's more people on the team that are helping with this. This is a lot of very difficult, complicated things to do. But it's all seemingly going according to plan thus far. Yeah, I mean, like I said, he's probably like the world's best trader right now at this point. He's just hitting win after win, which just doesn't happen, especially with capital that he's trading at this size. The way I kind of think about this, Josh, is he's gone levered long core weave because he's bought the riskiest asset that is either on or technically off Corwee's balance sheet, which is this provider, which is letting them down right now. So I'm guessing if he owns, what did you say, a 9% stake in Core Scientific? Correct. Right. He probably has a board seat on that company, right? So he's probably advising them in their ear day in, day out, how to construct their GPU supply such that, you know, they can meet the goals that Corweave has set for them. So that Corweeb's stock earnings next quarter is up only, stock pumps and all their partnerships with Open AI, Microsoft, Meta, Microsoft, let's play the violin, does extremely well, right? Because like it gets the kind of like domino effect. of stock appreciation for all these partnerships that they've already signed.

So this is an incredibly risky thing. Will it pay off? I don't know, but so far as track record kind of presumes, hey, well. So far, it is definitely paying off. And that leaves me with two questions that I want to get to to kind of conclude the second part of the show, which is, one, how do we participate? What does this portfolio look like? Where are the opportunities for us still available?

And two, is he just having the best luckiest year of his life? Or is this actually a durable strategy? you. So maybe we start with the first one, which is portfolio breakdown and how we want to think about how we can not necessarily emulate. This is not investment advice, but think about how he considers his bets and see which ones we think that are worth placing ourselves. Okay. I'm getting FOMO. That's like three extra turn in one year is crazy. I want in on this. I am also getting FOMO. So Josh, like not financial advice, but I'm showing you a tweet, which literally has the tick is carved out for the entire fun. Let me screenshot. Hold on.

Yeah. Please screenshot it. Okay. And for those of you are listening, you know, up to you, whether you want to do that as well. What's good about this is he's going, he's ranked these tickers on this tweet in terms of like how sized he is in this position. So the ticket that isn't in here that is still his biggest size is Intel. He's got a huge call position on there. And that's kind of like his kind of sit. That's his kind of steady state investment, right? And then he has core weave, which is now his new second largest position at almost $600 million. Then he has core scientific at, I think it was around $360 million. And then as you go down this kind of list, Iron, CIFR, Hut, right?

These are all much smaller positions, but much riskier bets. He doesn't know for sure if they are going to play out. That's kind of the way that I'm looking at it. And then do you have any, sorry, do you want anything to say? No, no, no. And then to answer your second question, Josh, you know, will this thing kind of play out? Well, let me introduce you to this tweet, which talks about the top five most contracted AI compute capacity in development.

Oh, those stickers look familiar? Yeah, yeah. Does those tickers look familiar to you, Josh? Oh, look, there's iron that is committed to 2.9 gigawatts of power. There's core weave at 2.9 gigawatts of power, and there's CIFR, 2.6 gigawatts of power, and Enbis, which the fund did have a position. I don't know whether it still has a position in. The point being, that is what?

3, 6, let's call it 9 gigawatts of power, Josh. That translates into hundreds of billions of dollars. who's spending all this money? It's cash flow positive mag seven companies, baby. It is meta. It is Google who all have like crazy balance sheets that can afford this spend. So it's not like it's even in a bubble where it's a leavened bet for those kinds of companies. This is just like agreements that they've made and signed that are inevitably going to be fulfilled.

Crazy. And this is what we talk about with that, that layer stack where there's like the core infrastructure, then chips, then there's energy and then there's the app layer and then there's the like APIs. It's, it's like he's, I guess you could think of this as like, we have the conglomerates, we have Google, we have Meadow, we have those big ones that you described. They're bringing in tons of cash because they just have very durable businesses that are super cash flow positive, make tons of profit.

They're sending their profits to the companies that can do the hard work for them, that could build the infrastructure, that could create these AI labs that they just don't want to do. One noteworthy one actually, as I'm describing this is Oracle, because I believe Oracle is partnered with OpenAI to build a lot of their AI infrastructure. 300 billion. The biggest. So do you have any idea why something like that would be missing?

Because that seemingly feels obvious to add to this position. Or maybe it's just too big to fit into this category. Because it seems like he has megabets and then some mini bets. An Oracle could perhaps be a weird middleman that doesn't quite fit the mold. Honestly, it's a good question. So my head immediately goes to two things. One, I think it's more of a middleman play than the safe bet that he's already kind of taken with conviction, which was Intel. And that was probably a much better bet that he would take over and over again because he invested pre the government taking 10% right.

That government state kind of gives it any kind of investor safety that like, the government's got it. It's not a bailout per se, but it's more of like a, I think this stock probably isn't going down and they're going to sign a bunch of future partnerships, right? Like with the government's help, Intel signed a massive partnership. I think it was like $100 billion worth of GPUs with Nvidia and like all these kind of like alliances which you wouldn't expect. So I think he's kind of fine with the Oracle side of things. What's interesting to me, Josh, and I'm curious what your take is because you and I have gone back and forth on this. Listen, you and I both know GPUs are super essential, right?

But we talked about Satya Nadella from Microsoft last week, basically saying that a lot of his GPUs that he's already bought hundreds of millions of dollars worth are collecting dust. What was the reason? Didn't have any energy. Didn't have any grid supply for that. I don't see any tickers necessarily that are directly on the energy supply train here. He's gone kind of like for the AI compute provider, which they kind of have involvement in setting up the data compute and the grid supply, but like nothing directly there.

Like is that something that he's missing? Is that something that you and I maybe have alpha on? I don't know. Maybe. But I wonder if it really exists because when I think of these companies, I'm not sure how many are actually in the business of generating power. Energy. Like when I think of that, I know Siemens is a company.

It's a, it's an European company, actually. It's not even traded on the New York Stock Exchange. They specialize in making these gas turbines that a lot of, gask turbines that a lot of these power centers are using. But their backlog is for like years and years of time. So I wonder if there's just genuinely a shortage of large companies that are doing it. And it's more of a distributed effort in terms of who is actually able to get power to these data centers. So that could be why.

That's my best guess. I'm not totally sure. I mean, he knows who the biggest consumers are going to be, regardless of who builds the energy, right? So maybe the way he's thinking about it is it's too risky, to your point, Josh, to bet on who the energy suppliers are going to be, let them figure that out. But I know once they have ramped up that energy and boy will they, because the demand is overflowing at this point, it's going to flow directly into these companies, these neoclads,

because that's who all the tens of billions and hundreds of billions of dollar partnerships are being signed by the hyperscalers. But what a great masterclass in conviction. And like really having the foresight to see where the puck is going, going, all in in creating this essay, getting a billion dollars fundraise, deploying it, and then actually hitting. Now, that leaves me with, I guess, this final point that I want to ask you, EJez, because, like, he's not really betting on AI. Well, he is, but he's not betting entirely anti-A-I. He's betting that his timing of AI, which is this 2027-ish AGI is right. And if that's

off by a couple years, by a decade, you very quickly go from genius to cautionary tale. And so far it's working, and so far there have been no gluts. But with everybody's screaming bubble with everyone screaming, you know, there's a lot of uncertainty in the market. The prices of lots of stocks are going down. Is this a stroke of luck or is there real genius? I suspect it exists somewhere in between, but I'm curious what you think about that. Leopold is the smartest investor in AI. I think you and I probably might agree with that right now. Turning one bill to four bill is pretty nuts. The smartest man in AI could theoretically be Sam Altman. And Sam Altman recently released his roadmap.

for Open AI. Do you remember when he said he would achieve AGI, like a core research scientist? So AI is curing cancer. Yes. It was... It was... It was a decade?

Oh, wait. Oh, wait. Maybe I don't know. No, no. Okay. What's a decade? No, no. So, okay.

So I might be cheating a little bit here. But he said that we would reach expert research, scientific grade, AI, by the end of 2020. He said we're going to get the AI research intern by the end of next year, 2026. So technically, if we had an AI that was producing, you know, potential cures for cancer or kills for like things that kind of like cured trillions of dollars worth of ailments or whatever that might be, I would say that's somewhat on the verge of AJ. And I can imagine a model that powerful, what could it be from the consumer side?

That would be chat GPT 5.1 on steroids. So if I had to guess with all the bets that are paid off so far with Leopold, and I'm guessing he's updating his thinking over and over again. He's seeing kind of decades in advance at this point. He might be right. I don't know whether, like, he might be right. What do you think? Do you think I'm crazy?

I think it probably doesn't matter to an extent because the gains will be so large in the meantime. Like he's collected that zero to one on all these companies. Like he bought a Bitcoin mining company that nobody cared about and now it is going to be an AGI mining company. And that like delta between the like nobody cares about this to, oh my gosh, this is so valuable, is very high. And just collecting that alone means that the timing probably means a little bit less.

And I guess the takeaway for myself and everyone listening is that you don't really have to agree with his AGI timeline exactly. But there's something to learn from the way that he starts with a worldview and then looks for these choke points and then asks, where is the market still pricing the old world instead of the new one? And I think that's kind of how I'm going to leave this episode with is where is the world not seeing a future that I believe. Like, where is there misalignment and where is there an advantage there?

And Leopold with this fund has clearly shown that there is a huge arbitrage opportunity for those with the eye and the will to actually seek these out and pursue them. Really well said. I want to end with this term that Nick Carter used to describe Leopold, which is he is not an AI duma. He's also not an AI extremist. he's more of a third in-between thing, a realist, but a really smart one. And he describes him as a pragmatic, cautious optimist or optimism as he's describing the behavior.

And that, I think, is Leopold's secret source. It's not the essay. It's not kind of his general approach to trading or his investment style. It is his pragmatic, cautious optimism. I think that is it for the rest of the episode. that, like I am just, I can't get over the fact that he's 23 years old, Josh. For the listeners of this show, Limitless is the 23-year-old Leopold Ashon Brenner before we turn our billion dollars into four billion dollars.

Get in early. We know that over 80 percent, and this is a crazy statistic, guys, 80 percent of you listening aren't subscribers on YouTube or don't follow us on Spotify or Apple or wherever you listen to us. It would mean the world to us. It helps us out so much. Give us a five-star rating if you think. This is five stars. We do. And we will see you on the next one.

We have a bunch of episodes coming out. I got to add a spoiler too. Please. I got to let people know because our backgrounds are different this time and people might have some questions. And the reality of this is that I'm actually looking at EJAS in person right now. I can see him through the window. We are staying in the same place.

We are in, what's up over there? We are in Argentina. What I'm not realizing EJAS, what I failed to compute when I came here is that it's springtime in Argentina, which means there's pollen everywhere. And my allergies are insane. So while it feels so nice being outside in shorts and a t-shirt, I cannot breathe right now. Josh has been an absolute trooper, guys. He's doing this for the love, love of the game.

But it's great. So once we get the camera situation sorted, I'm hopeful that later in the week we'll have an actual IRL shoot set up, which should be really fun. So that's something to look forward to. But like you just said, do not forget to like, share and subscribe. Share us with your friends who are interested in this. Share us with your friends who you want to get rich. Or maybe, maybe, maybe just maybe. do your own analytics and then share it in the description of this video in the comment

sections that we can take your ideas and we can start placing some bets because he's been doing well and I think we all deserve to participate in a little bit of that upside so thank you all for listening for watching wherever you watch this and we will see you guys in the next one hopefully hopefully with Gemini 3.0 fingers crossed hell yeah see see you guys see you guys