One of the most aggressive, successful,

some would say genius billionaire

investors in the market right now just

made a move that should make every

investor stop for a second. And yes, I'm

talking about Leopold Aeschbacher, an

ex-OpenAI researcher who founded

AI-focused hedge fund called Situational

Awareness that reportedly ramped up to

roughly 20 billion in assets. And the

craziest part, he's only 24 years old.

Because this is not just some random

bear on Twitter saying the video's over.

This is somebody who made his entire

name riding the AI wave. And now Leopold

Aeschbacher is suddenly going against

some of the biggest winners in the stock

market, some of which he bet on himself

that made him rich. So we're talking

about shorting exposure tied to Nvidia,

AMD, Broadcom, Micron, ASML, Intel, and

the broader semiconductor complex. This

is very strange. What is Leopold

Aeschbacher doing right now? And this is

not just some rumor or some vague Wall

Street chatter. This started getting

real attention because the firm's 13F

filing, which pointed to something like

$8 million to about $8.5 million

of bearish exposure with one of the

biggest puts tied to names like Nvidia,

AMD, and Broadcom. And seeing this, you

have to ask yourself, does he know

something that we don't know? Does

Leopold Aeschbacher suddenly think the

whole AI boom is done? As an ex-Wall

Street analyst myself, this made me very

curious. So that's exactly what I'm

going to be answering in this video. And

actually, don't really think this is the

most interesting part of all. The most

interesting part is where the money is

moving next and what he's doing with his

money instead. Because if you really

break down the thesis here, this looks a

lot less like an anti-AI call and a lot

more like a rotation call. A call that

says the easy money may already have

been made in obvious AI names and the

next bottleneck might be in power,

memory, data centers, and physical

infrastructure. Now, the reason this

matters is because a lot of people still

think AI investing is simple, and they

think that if you just buy Nvidia, or,

you know, even some of Leopold's

holdings himself, like Nebius, then it's

all going to be easy work, right? But,

that is not the case. Look, I get why

people think that, but Leopold's

position sizing right now, and his new

positions, is completely different from

what investors think. Those stocks have

absolutely exploded already. They have

been some of the cleanest winners of the

last couple of years. Some of these

stocks are on generational runs. Like,

in the last year alone, look how much

Micron Technology is up. Look how much

Nebius is up year-to-date. But, the

market is changing very quickly. And

once a trade becomes too obvious, too

crowded, too universally loved, you at

least have to ask where the next big

opportunity is somewhere else in the

stock market. Now, let me make this

super simple for you, and not gatekeep

any information. If you look at the

situation itself, the shock value is

obvious. The big headline is that

Leopold Ashchenbrenner appears to be

very aggressive on the short side

against a lot of AI semiconductor names.

So, what it's looking like to me is that

semiconductor names are overvalued. And

when you see a 13F from Situational

Awareness LP pointing to roughly $8.5

billion in bearish exposure, you cannot

simply brush that off, right? That is

absurd information that might move the

stock market. But, this is where I want

to slow everyone down, because I do not

think the smartest takeaway here is to

say, "Okay, let's sell everything. AI

stocks are going to be crashing

tomorrow." Because I think that smarter

takeaway is to ask, "What exactly is he

betting on against, and what exactly is

he still bullish on?" This is where the

thesis gets way more interesting,

because this is not just Leopold getting

bearish across the board and going

entirely into cash. It's not like he

woke up and said, "AI is a bubble, and

everything is going to be connected, and

everything is going to come crashing

down." What's actually going on here is

we're really looking at a split book. On

one side, pressure against obvious

winners is in semiconductors. On the

other side, continued interest in names

connected to data centers, power,

memory, and infrastructure is something

that he still likes. So, although

there's pressure one from one side, he's

rotating to another side. So, if you

actually zoom out, the message is not AI

is dead, AI is over, and there's no more

money to be made. The message instead is

that AI trade is inside this new change,

and he is making that pretty public, and

he's not doing that by choice. He's not

doing it out of the goodness of his

heart. We are seeing that live happening

right now by his 13F filing changes. And

before we go even further, one important

clarification. A 13F is a quarterly end

snapshot. Meaning that when I'm

referring to some of his positions that

I'm seeing within his, you know,

portfolio, I am looking back at some of

the positions that he has already

opened, and they're at the end of March

31st. Had disclosed put option exposure

at the end of the quarter. However,

based on my own research, I still

believe that Leopold Aschenbrenner has

some of these positions open. It does

not give us a real-time look at every

single trade that he has made since

then, but it does give us a very good

picture of what likely he is thinking

and the reasoning behind some of his

trades. Now, the nuances really matter

here. So, make sure to listen up. So,

look, the biggest bearish position in

his portfolio right now, the largest

one, is a put on Van Eck Semiconductor

ETF SMH, at just over $2 billion of

reported value. Right behind Nvidia,

which is at $1.5 billion in puts. And

then Oracle at about $1 billion in puts.

And by the way, if you looked at Oracle,

here's the chart. Just in the past week

alone, Oracle has been on a very huge

decline. So, for me, I believe that

Leopold is personally profiting right

now. That is my personal belief. I still

think he's in the Oracle trade.

Broadcom, he has 1 billion of short.

AMD, 969 million dollars of short.

Micron at 584 million. TSM, about 500

million. And ASML, about 494

million dollars. Then, he has smaller

put positions in Intel and Corning. So,

it's very clear Leopold Aschenbrenner is

making a very strong bet in the market

right now. If you don't pay attention,

you might be making a very big mistake

because this guy, I'm not saying he has

info that we don't have, but honestly,

this story kind of smells like a like

kind of like a Nancy Pelosi here because

this guy is a little too good. At age

24, this guy has literally built his

fund. You can see how his fund has grown

since Q4 2024 and then what happened in

just 15 months. So, what he's doing

right now is a broad bet against a huge

part of the current AI hardware and

semiconductor trades. And that is what

makes the filing so interesting because

the long side tells a very different

story. His biggest long common stock

position is still Bloom Energy at

roughly 879 million dollars plus another

rough 55 million dollars in call

options. He also owns about 724 million

dollars of SanDisk plus roughly 389

million dollars and call options. So,

Leopold Aschenbrenner is not only long

stock in SanDisk, he's long call

options. We're seeing a very interesting

bet because obviously call options is a

massive bet. If it goes down, your call

options lose all their value. So, he's

making a very clear message right now in

the market on his own beliefs. Now, he

also has 556 million dollars of Core

Weave and he has 141 million dollars in

call options in Core Weave. And then,

you get to IREN at roughly 400 million

dollars. Core Scientific, Applied

Digital, and Ride Platforms at about 142

million dollars. And then, And his

smaller position, CleanSpark at $100

million.

This is not somebody saying that AI is

over. It's actually pretty much the

opposite. From my observation, he thinks

that semiconductor trade is overcrowded

and not worth it, and that a lot of

investors are going to be heading for

the exits and making a new rotation. So,

this is somebody that's saying the

market may be paying too much for the

design layer and not enough for the

deployment layer. And when you compare

this to the previous quarter, the shift

gets even clearer. The big semiconductor

book is new, while several

infrastructure names were either

maintained or increased. Coreweave moved

higher, Applied Digital moved higher,

IREN moved higher, Riot moved higher,

and CleanSpark jumped sharply. So,

Leopold Aschenbrenner seems to have, you

know, the wisdom, the foresight, the

technical analysis, or just the industry

experience that has been telling him

that some of his plays, including

SanDisk, that is higher as well, has

been very successful for situational

awareness. He's leaning away from the

crowded chip winners and leaning in to

power, memory, Neocloud in data center

capacity. Now, let's talk about why

those names make sense if that is a

thesis. So, Coreweave and CoreScientific

are basically expressions of compute

deployment. Applied Digital, Riot,

CleanSpark, IREN, Bitdeer, and Bitfarms

are tied in different ways to data

center capacity, power access, or

infrastructure that can be repurposed

for AI workloads. Bloom Energy is the

obvious power angle, which he has liked

for a long time now. And the stock has

run so much. It's like $350 per share.

So, this stock kind of seems obvious to

me, but he's still holding it. SanDisk

is the memory angle, so obviously all

this data has to be stored somewhere, so

the memory angle is very important. And

I'm pretty bullish on SanDisk myself

personally. Now, these picks is very

different. It's a different map

completely of the AI trades that average

retail investors has on top of their

head. This is completely the opposite of

most likely what many investors are

doing, and that is where it gets really

interesting as an opportunity for

investors looking at what to do today.

Cuz most people still think that AI

investing means you just buy GPU

winners, and then you're done. What the

filing is really arguing is that the

harder problem now may not be designing

the chip. That is already kind of

yesterday's issues that are already

pretty much solved. It may actually

today be powering those chips, housing

them, cooling it, and actually getting

it online fast enough. And that is a

serious point. If compute demand keeps

rising, then the next bottleneck may

belong to whoever controls electricity,

memory, data center slots, and physical

capacity. And that is the real analysis.

So, the actual question now is where the

next dollar of AI capex is going. Now, I

also want to be careful here, because

this is where most people overreact. A

filing like this does not prove that

Nvidia's broken. It does not prove that

Broadcom is done for or semi suddenly

stopped working. What it proves is that

at the end of the first quarter,

Situational Awareness LP had a very

large put book positioned for a lot of

these stocks going down against those

names while still owning a lot of

AI-adjacent infrastructure on the long

side. So, this is actually what a hedge

fund is for. Hedge funds, the name

hedge, was originally done for

protection. That's where options also

came into place. A hedge or buying a put

option was for hedge funds back in the

day to protect investors from downside.

That's exactly what Leopold

Aschenbrenner is doing. He's making

large bets on what he thinks is going to

go down, what he thinks is going to go

up. And this is much more than a precise

statement. And once you say it that way,

the debate becomes very clear. He has

put options because he thinks that

semiconductors are overpriced, and then

he has call options because he has a

view that the infrastructure layer in

the AI cooling capacity, electricity, is

going to be in high demand. So, the real

question is whether semis are actually

crowded, where their power and

deployment is becoming the next

bottleneck and whether the market may

end up paying more for physical scarcity

than for more turns of excitement in the

obvious chip names that have already

ran. So, that is the conversation that I

personally care about because that's

going to give me insight and my discord

community on what we could be doing next

that could be the next money-making

opportunity within the stock market. So,

my own read is that the most interesting

part of the filing is not actually the

bearishness. It's actually the

selectivity. Because if he were simply

calling the top of AI, the book would

look very different. Instead, the book

says something more nuanced. It says

that you can still believe AI spending

is real while thinking that the best

risk reward may be shifting away from

names that already is owned by everyone

else. That is very different from saying

Nvidia is doomed and honestly, that

distinction matters a lot. Nvidia can

remain a phenomenal company and still be

a harder stock to buy after huge move if

expectations are already extreme. And at

the same time, power, memory, and

infrastructure names can still have room

if the market has not fully priced how

important they have become in the next

stage of the build-out. This is why I

would not treat this as an all or

nothing call. I would instead treat this

as a watch list call. If I were doing

the work from here, I would split the

spaces into buckets. First, the obvious

one is chip leaders. Second is memory.

Third is power. Fourth is neo cloud and

data center capacity. That one I'm

personally most bullish on myself. And

then, I would ask whether expectations

already look perfect and where the

bottleneck is getting worse faster than

the market is appreciating. This is

exactly how a retail investor should use

this information. Use it as a smarter

question. Where is AI spending actually

flowing next? Because that is where the

next winners probably come from. Now,

here is my personal pushback on

Leopold's bearish side. I still think

betting too aggressively against Nvidia

is very dangerous. This is the leader in

all of AI. They're supplying the whole

entire AI boom. And betting against

Nvidia to me doesn't really make sense

because that's just timing. And I mean,

unless Leopold has information on

exactly when the stock is going to go

down, this is not something that I would

look at as a retail investor. Nvidia

still has real software lock-in, real

customer demand, and real ecosystem

advantage. So, if his view is that

Nvidia is simply over, I would actually

completely disagree with that. But, if

his view is that the semis can remain

great businesses while the better

incremental upside move to power,

memory, and deployment, well, I can

agree with that. I do think that is a

strong argument. The right takeaway is

to notice where a very smart, very

aggressive investor seems to be thinking

is the next move within the stock

market. In this case, the answer looks

pretty clear. Power, memory, deployment,

and physical AI infrastructure. So, if

you're a retail investor, I would leave

this video thinking about two questions.

First, am I too concentrated in obvious

AI winners just because they worked in

the first phase of the AI build-out?

Second, if the bottleneck really is

moving from chip design to real-world

deployment, which names are actually

going to benefit the most from this

shift? That is the part of the video

that matters most for you as a retail

investor. The long book tells you that

AI stocks are still going to explode. Do

you believe that Leopold Aschenbrenner's

bearish call on semis is overblown? Let

me know in the comment section below.

Thanks for watching. Make sure to like

and subscribe. It's totally free, and

I'll see you in the next video.