YT YouTube2Text

YouTube2Text → Transcripts → Is Bitcoin Ready To Rip While Wall Street Breaks?

Is Bitcoin Ready To Rip While Wall Street Breaks? — Transcript

by Anthony Pompliano · 11,588 words · 1,694 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00You know what the hedge is? If I'm
0:01right, it's to go buy Bitcoin. So, if
0:05you believe that you're going to get
0:06disrupted by AI,
0:08then you start asking if I'm getting
0:10disrupted, who isn't going to get
0:12disrupted. It's really hard to come up
0:13with any company, which is why I focus
0:15my attention on the agentic compute
0:17side. I have total confidence over the
0:19next three years that as a basket of
0:21semiconductors of the optical companies
0:24of the transporta you know the
0:26industrial companies that make the data
0:28centers that over the next 3 years with
0:30humanoids coming and everything else
0:31that what's going on guys today we've
0:33got an excellent conversation with Jordy
0:35Visser in this he's going to talk about
0:36the death of hedge funds and why
0:38tokenization may be the thing that
0:39brings that right to their doorstep we
0:41talk about long-term capital management
0:43bitcoin tokenization what is going on
0:45with CPI why we don't care about it
0:47anymore how AI is now inevitable, what's
0:49going on in financial markets, what it
0:51means for your portfolio, and then of
0:53course we got to disagree on something.
0:54So, we're going to get into a very
0:56heated debate, heated for two gentlemen.
0:58And in that conversation, I think you're
1:00going to learn a lot about how AI is
1:01working, token prices, and how we see
1:03the world evolving over time, both for
1:05big companies, small companies, and how
1:06you can invest along all these themes.
1:09That and much more in today's
1:10conversation with Jordi Visser. All
1:12right, Jordy, you've got some very
1:14unique thoughts this week on long-term
1:16capital management and how we may be
1:17living through a very similar time right
1:19now. Can you give us context on how
1:20you're thinking about this moment in
1:22time?
1:23>> Yeah. And for those of you who are um my
1:26subscribers, you saw me do a very brief
1:30video on Thursday. And the reason I put
1:33it out is because so many people had
1:35reached out and I had a busy week. But
1:37because people had reached out and said,
1:38"What's going on? Is this something I
1:41should be worried about?
1:42>> The sell-off and all of the fears around
1:43the AI.
1:44>> Yeah, semiconductors are down seven days
1:46in a row. Coreweave, CDS, all this
1:48stuff. And of course, the echo chamber
1:51of bears gets louder and louder whenever
1:53there's a Oh, it's popping now. Finally.
1:56Um it to give you like a brief sports
1:59analogy,
2:00this is like you go into a game um and
2:03you've decided to bet on a football team
2:06that you know is is given 28 points.
2:09>> So there are four touchdown underdog and
2:12the you get the first touchdown and you
2:14start banging your chest. That is what
2:16bearish people on the AI bubble popping
2:18when the S&P is a couple percent off the
2:20all-time highs feels like to me. Like
2:22I've never met anyone who predicted the
2:24end of a bubble and I certainly didn't
2:26don't think anyone who's waiting every
2:28day for anything to happen. So this week
2:30brought out a lot of fear and I I want
2:33to share some thoughts on how LTCM fits
2:35into this because I think it's really
2:37important for people to understand
2:38leverage. So I'm a believer and this is
2:41the reason why I'm not in the hedge fund
2:43industry.
2:44>> I have no interest in being in a
2:47competition against artificial
2:48intelligence which everyone who's
2:49staying in the industry is competing.
2:51they're competing against AI agents. The
2:53fact that any retail person can set up a
2:54long short hedge fund instantaneously
2:56now back test things and do it. So, I
2:59didn't believe this was going to be an
3:00easy period.
3:02But there is no doubt in my mind with
3:04100% certainty, just like before the dot
3:08bubble. We knew before the.com bubble
3:11that the internet would change our
3:12lives. What we didn't know was how long
3:15it would take. So when the.com bubble
3:17happened, which was right after LTCM,
3:20which I'll get into, people have to
3:21understand every single thought on the
3:24internet ended up being right. The
3:26problem was the capital that was
3:29borrowed to build the pipes necessary
3:32for it took a long time to get there.
3:34Now, let's go back to LTCM. I was in
3:36Brazil.
3:38I got to see LTCM on the other side of
3:40their trades. And to give you the kinds
3:42of trades they did, Telos, which was one
3:46of the largest trading stocks in the
3:47world in 1997, was going through a
3:50privatization. So the company basically
3:53was going to be or not, they were going
3:55to be privatized. And what was going to
3:56end up happening was
3:59ordinary and preferred shares would
4:01converge to the same price. So, one of
4:04the trades that LTCM did was to
4:06basically short the preferred stocks
4:08trading at a premium to the ordinary
4:09shares and they'd converge as the
4:12privatization got the date became more
4:14certain. Well, before that date
4:17happened, the spread widened out. Now,
4:21why did it widen out? Why would it widen
4:23out? Well, there's a variety of reasons
4:25that go into why a spread would widen
4:27out. And that's what hedge funds are
4:28going through right now. And the reason
4:29I say with certainty
4:32AI and the compute needs for it are an
4:35inevitability. You can debate whether
4:37it's going to take 20 years and we'll go
4:38through some form of a.com bubble. You
4:40can you can have that debate. I everyone
4:43knows my views which are as long as
4:45demand is greater than supply. There's
4:48no chance in my mind of having an unwind
4:51that is very similar to the dot bubble
4:53because the people that are financing it
4:56are very different than the telecom
4:57companies. Uh, I made a joke on my video
5:00on Thursday when I compared it to the
5:03housing market. I'm like, "Okay, let's
5:04go through Cororeweave." Corweave
5:07doesn't have strippers funding
5:11their needs for capital. It's actually
5:13coming from Microsoft, from OpenAI, from
5:16go through Meta, go through the list of
5:18really big companies. And I made the
5:20joke about the stripper funding because
5:22that's what was going on with the
5:23housing bubble. Like they started
5:25defaulting on their payments.
5:27Microsoft's not defaulting. Meta is not
5:29defaulting. OpenAI is not defaulting. So
5:32for Coreweave to be an issue, the
5:34Gartner group said that they think by
5:362030
5:39when companies report earnings, they
5:41have they have visibility with orders
5:44sold out to 2030. This has been coming
5:47on repeatedly. So for the first time in
5:49the history of markets and certainly the
5:51first time in the history of my life,
5:53everyone is worried
5:56about whether we're going to hit
5:59revenues in 5 years.
6:00>> Mhm.
6:01>> That is not going to drive the
6:03inevitability of AI. And I find it
6:05almost impossible that we won't hit
6:06there because I listen to people like
6:08Jensen Yuang. I listen to people that
6:10have been in the technology field for a
6:11long time that realize that every day
6:14the orders keep coming in for compute,
6:16for compute, for compute. So everyone
6:17can take the nuance. I'm bearish
6:19anthropic. I'm bearish OpenAI. I'm
6:20bearish this. The problem people have to
6:23understand is the only leverage that I
6:25see in the system that is being knocked
6:27out is hedge funds.
6:30>> We recorded last week at the end of that
6:33day on Friday of last week one of the
6:36best hedge funds of this century. This
6:39century, not situational awareness, not
6:41some guy who just started managing
6:43money. Lone Pine Capital was up 40 plus
6:47percent in the first half of the year
6:50and as of the end of September
6:53and I want to say in a quarter that the
6:55S&P was up 2%. Never did the VIX go
6:58higher in any meaningful way never in an
7:02entire time was there a big move on a
7:04monthly basis and yet somehow they gave
7:05up the entire performance. And so people
7:07understand how that means. If you're up
7:0940 plus% and I don't know what their
7:10fees are, but let's assume you're
7:12charging 20% and let's leave the
7:13management fee out. That means by
7:15definition you were up more than 50% cuz
7:17you gave it all back and now your net
7:18return is is zero.
7:21That just says that something big is
7:24happening. It wasn't just situational
7:25awareness. It was some of the best hedge
7:27funds that have ever existed. And in my
7:30opinion, what is going on and will
7:32persist for the entire 5 years to the
7:35inevitable is what I watched with LTCM.
7:38You are making a bet as to how long your
7:41investors will stay with you.
7:42>> Mhm.
7:43>> You are making a bet on how long the
7:44prime brokers will give you leverage for
7:46whatever leverage you're using. We are
7:48watching leverage come out of the
7:49system. And the hedge fund industry, in
7:52my opinion, is going through a process
7:54which is going to weed out a lot of
7:56leverage in the system. And this process
7:58is going to continue and continue
8:00because the concentration of the winners
8:02in AI is so small and the losers are so
8:06big that this process will leave people
8:09needing to invest in certainty and with
8:11certainty you're making a bet on time
8:13and your ability to navigate that path.
8:15So I think this was a big week and when
8:17you listen to the bears going the
8:18bubbles unwinding the irony is the
8:20bubbles are in really smart people that
8:22have used leverage to make money for a
8:24long time. When you look at investment
8:27strategies, there is the Buffett, you
8:30know, era, buy great things, hold them
8:31forever. There are gold bugs,
8:33Bitcoiners, buy great asset, hold it
8:35forever. Um, hedge funds, they really,
8:39outside of the rents and maybe the
8:40citadels and market makers, most of the
8:42hedge funds have found, you know, kind
8:44of structural mispricing or structural
8:46arbitragees. They've used a lot of
8:48leverage and they've been able to
8:49generate significant returns. What I
8:51think I hear you saying is that that
8:53section of investing is now rapidly
8:57evolving on a day-to-day, week to week,
8:59month-to-month basis. And so, you can't
9:01sit here and say, "Hey, you know what?
9:03I'm going to buy the close and sell the
9:06open every day, and I'm going to do that
9:08for months or years, and I'm going to do
9:10it with lots of leverage, and so I'm
9:11going to clip my 100 basis points, you
9:13know, each day, and I'm going to end up
9:14with this great return." And because
9:16everything is evolving
9:18so much that that's really where now
9:20these people have to evolve with the
9:22market and you're just increasing the
9:23odds that they make mistakes. Is is that
9:25the way to think about it?
9:26>> So I'm going to I'm going to go a little
9:28bit further backwards and say this. If I
9:31asked you what has hurt private credit,
9:34if you really think about it, it was AI.
9:38So AI
9:39>> so the majority as we learned of the
9:42private credit investments
9:45They're from 21 22 and they're SAS
9:48companies. What did AI destroy was SAS
9:50companies and it didn't move them all to
9:53zero. But what it did was if you make
9:56investments in these companies and
9:57they're in competition with AI, it just
10:01eats at you. My thesis has always been
10:03that human beings are no longer the top
10:05of the intelligence food chain. If
10:08everything that was made was about LTCM
10:11when genius failed, we are now at the
10:13point where an AI agent can do
10:16everything that a hedge fund can do.
10:18Everything. There is not a single thing.
10:20And who would you rather lend to?
10:24Risk management that is instantaneous in
10:26seeing everything happening that can
10:27read every single thing from a company
10:29can understand am I in a crowded
10:31position or not by going through the 13F
10:33filings. Hey, I heard I mean there's a
10:35new hedge fund Vera not new but Vera Aai
10:39reportedly with 26 they were down 40ome%
10:41in July. So this is a new story today. I
10:44don't know anything about them. I don't
10:45know what the reality is in terms of
10:46where they are but they're going through
10:49a similar situation. And I'm sure just
10:51like with LTCM if you go back all of the
10:54positions that were unwound from LTCM
10:57made the people that took them on money.
10:58So Ken Griffin got the whole LEO like
11:01acknowledgement.
11:03I don't think people realize when LTCM
11:05went out, the Fed didn't bail them out.
11:07They gathered 14 banks, hedge funds,
11:10whatever that contributed, I think it
11:12was slightly less than $4 billion to
11:15take their stuff out. Now, I was part of
11:16the LTS tom closed down. When I went
11:18back to New York after Brazil, so I was
11:21on one side of LT Sam, I saw the other.
11:23I I don't want to pick on the hedge fund
11:25industry. What I want to pick is that
11:27illlquid assets have been impacted. VC's
11:30been impacted. Private credit's been
11:32impacted. Private equity's been
11:33impacted. Bonds have been impacted.
11:37Real estate's been impacted. Meaning
11:39remote work AI is going to bring digital
11:41agents. What are we going to do with all
11:42these seats? I just think that when you
11:45look at the returns you're getting
11:46relative to where interest rates are,
11:49you bogey is getting higher and higher
11:51and AI is pushing rates higher at the
11:53same time. So AI is crowding out all
11:56inefficient businesses, but it's also
11:58going to crowd out hedge funds who use
12:00leverage that depends on the volatility
12:03correlation of the market and the
12:05crowded nature of the positions. And I
12:07just think the industry is too big on a
12:09levered basis relative to the liquidity
12:10that exists. This will merge with
12:13tokenization because the final piece of
12:15this and I was at Morgan Stanley this
12:17week and and at the very end we talked
12:19about tokenization.
12:21If you're a hedge fund listening to this
12:23and you haven't thought about the way
12:24tokenization will change your life,
12:27forget your business, your life.
12:29Riskmanaging 24/7,
12:32365, it's it's an unthinkable process to
12:35me, but that's where we're headed to.
12:37The capital that will be immediately
12:40changing. It has huge implications for
12:42the volatility in the market has huge
12:44implications for the way you hedge your
12:45book and the and the the the ability for
12:47prime brokers to provide leverage. Now
12:50that may not be this year's business,
12:51but just like with AI, these things are
12:53moving so fast that over time your bogey
12:56becomes, hey, where's the cost of
12:58capital? That's why when you and I talk
12:59about for me personally,
13:01my cost of capital is related to
13:03Bitcoin. It's not related to this cuz
13:05that's the way I see this ending. And so
13:07for AI, if you really go through when
13:09the hedge fund industry entered this
13:11abyss, it started in June. And since
13:16June, let's say a exactly four months
13:20ago, go look at what Bitcoin has done.
13:23It's up close to 30%. Go look what
13:25Ethereum has done. It's up more than
13:2630%. The S&P flat. NASDAQ flat. Hedge
13:30funds, I mean, I just told you a lot of
13:33these funds are down 40 plus% and
13:35nothing has happened. This should be
13:37scaring people because the only way this
13:39ends is if this isn't the case. So,
13:41every time we get another one of these
13:42little episodes like this week, the
13:45coreweave thing is ridiculous. They have
13:48cash flow coming from big companies.
13:51This is not the same situation as
13:53telecom. Are they good equity
13:55investments? I don't know. Are they good
13:57bond investments? I don't know. But I
13:59know this. We won't know for sure
14:02whether they're going to be in trouble,
14:04according to Gartner, until 2028, 2029,
14:072030. That's not going to drive the
14:08market from there. Tokenization you
14:11think brings both liquidity and
14:13transparency and you're arguing that
14:15tokenization could be the death of the
14:17hedge funds.
14:18>> I think as time goes on it will put so
14:21much pressure on the way hedge funds and
14:22again there are the mega hedge funds
14:25which will sustain the longest over time
14:27and for no other reason they've just met
14:29most of them gone to like 5year lockups
14:31so they've locked capital in.
14:34I think this tokenization and this
14:36concept of of bringing in locked up
14:39capital. So let's go back to private
14:41credit. If you go into private credit
14:44and you look where the BDC's are
14:45trading, so real time marks on this and
14:47let's assume right now the BDC's are
14:49trading at 20% below where retail
14:52investors who got into private credit.
14:54So they're basically over the course of
14:56let's say the next four years, let's
14:58just use all this, they end up losing
15:0120% on their investments. Okay, it's not
15:05going to kill them.
15:07It hurts in the fact that if they would
15:09have had that capital and been able to
15:11sell out at 20% below here and then put
15:16the money back in the market, well,
15:18that's double the impact. It's the
15:20trapped nature. When tokenization comes
15:23and people are like, "Okay, where can I
15:24get out of this?" Well, someone is
15:25willing to buy into private credit at
15:28some level. So, you have trapped capital
15:30that doesn't want to be there anymore.
15:32This is the same thing for endowments,
15:33same thing for a lot of as tokenization
15:35comes, it will free up the ability for
15:37there to be a price on things. And once
15:39there's a clearing price on things, then
15:40capital will go to where it should be.
15:42What it will also do is what would I
15:46rather own? Coreweave debt at 75 cents
15:49on the dollar or the private credit
15:52garbage that I own from the 2122
15:54vintages that I'm sitting in. I'd rather
15:56own coreweave right now for the next
15:58whatever years than this. So that
16:00capital would come in and core would go
16:02back up. Remember when stretch Michael
16:04Sailor's product traded down to 75?
16:07I heard the same the exact same people
16:10sending out expost on coreweave. The
16:13stretch things went all the way back up
16:15to par. So these arguments of people
16:18looking at short-term things and saying
16:19this is the end of it. This is the these
16:22are mainly hedge fund people or people
16:24that made most of their money on
16:25leverage. There are two ways like
16:29general ways that hedge funds end up
16:30making money. So Dr. Miller never talks
16:32about like leverage and I'm not sure
16:34what he's ever used in leverage. He
16:36talks about putting a position on being
16:38concentrated and that's the risk he's
16:39taking is concentration risk on a
16:42diversified hedge fund that's using
16:44leverage like an LTCM that's levered 30
16:46to1. They're betting on correlations and
16:48volatility staying in some kind of way
16:50that they don't have margin calls come
16:52in because of the amount of leverage.
16:53Well, that's the way most of the
16:54multistrats are set up. And so if we end
16:57up in a scenario biotech blew up in
17:00September, so everyone who runs a long
17:03short portfolio in biotech, and I say
17:04everyone, I don't like absolute
17:06statements. A lot of funds from what
17:07I've heard, uh, pods in particular, had
17:10trouble. Converts, converts are part of
17:13coreweave. They're also part of micro
17:15strategy. When you get these little
17:17things happening where people are
17:19running some kind of leverage where it's
17:20like, I'm long this convert, I'm short
17:22this convert. They're depending on these
17:24things not to widen out like LTCM. Well,
17:26this has been another widening out. So,
17:28I think for people that are trading and
17:30like I'm trying to give my subscribers
17:31thing of like if you need to make money
17:33on a position over the next month,
17:36you're in you're in trouble. Like take
17:38advantage of the time that you have. So
17:41maybe the most shocking thing to me in a
17:43positive way, all of the weakness in the
17:46stock market, all the weakness in uh the
17:48NEO clouds and you know the AI trade,
17:51Bitcoin has held in really strong. It's
17:55been trading pretty much between 82 and
17:56$85,000 and it almost feels like it's
18:00defying the market momentum where
18:03everything else was trading down and
18:04Bitcoin holds in. What do you read into
18:06that?
18:08So what I read in is
18:11as it gets harder and harder to own
18:13other things
18:15whether it's
18:18computer decided whatever it is human
18:20beings and computers look for good sharp
18:23ratios. Bitcoin sharp ratio right now
18:25over the last four months stands out and
18:28it stands out as the rest of the world
18:31is kind of going through this chaos. And
18:32and I do want to say this is not just
18:35the AI trade. Like I said, the NASDAQ's
18:38unchanged over four months. The S&P is
18:40unchanged over four months. The bond
18:42market is down. It's getting very hard
18:44to find investments just using a
18:46trailing period that do well. I believe
18:50it's because the size of the space is
18:52too small. Tokenization's a reality. Um
18:55like I said, I created my crypto
18:57ecosystem. You know, we talk about
18:59Bitcoin. I also can say that I watch
19:01Near every single day as kind of one of
19:04these AI agentic where I think it should
19:06do well. Ethereum the same thing. It's
19:08not just Bitcoin that's that's hung in
19:10there. It is the most quality parts of
19:12crypto. And I still believe that this is
19:15a reality of something for next year. Um
19:17if you ask me when the AI trade because
19:19if the AI trade is an inevitability,
19:22the time for it to rally
19:25is when people have just taken off the
19:26risk. And when do we learn this? First,
19:28with situational awareness. Right now,
19:31someone sent me something saying the AI
19:34long short trade on the Morgan Stanley
19:36Prime brokerage list is now in the
19:38seventh percentile of the last year.
19:40Seven. So, the risk has been completely
19:43taken off. Now, it's in the seventh
19:45percentile. It could still be high, but
19:47that is relative to the last year. For
19:50me, that means that people have bailed
19:52out. There are two three trigger points
19:54that I'm looking in terms of timing for
19:56the next wave. And this year, if you
19:58look at the S&P 500,
20:01it was very frustrating start to the
20:03year. First two and a half months and
20:04then all of a sudden we had a big fall
20:069%. And then out of nowhere with the
20:09expectations that oil was going to 200,
20:11the S&P started to rally purely from
20:13short covering and things like that. So
20:16there was really only a window of about
20:183 months this year where the S&P
20:20rallied. And during half of that time
20:22was the fear over
20:25oil. Now what drove the S&P higher? It
20:29was all the AI trade. So it was only a
20:31three-month window. And aside from that,
20:32that takes you from March to June.
20:35There's been no rally in the S&P the
20:37entire year. So your windows of making
20:40alpha have gotten smaller and smaller.
20:43And this is the way Nvidia's traded for
20:45the last 3 years. There's like they they
20:47make money on an alpha basis. It's like
20:5020% of the time. I think the entire
20:52market's going to be like that. So for
20:54people that are watching Bitcoin on a
20:56relative basis, it is starting to do
21:00what I thought it should do. And when we
21:02get into the end of the year, I still
21:04will say the biggest positive for me is
21:06that the amount of really smart people
21:09are starting to call up every day going,
21:11"Will you talk to me about tokenization?
21:13Will you like there?" If I write enough
21:15thoughtful pieces on how it affects M2
21:18and how it affects hedge funds, more
21:20people want to hear because they're
21:22interested in what if he's right. You
21:26know what the hedge is? If I'm right,
21:27it's to go buy Bitcoin. So if you
21:31believe that you're going to get
21:32disrupted by AI,
21:34then you start asking if I'm getting
21:36disrupted, who isn't going to get
21:37disrupted? It's really hard to come up
21:39with any company, which is why I focus
21:41my attention on the agentic compute
21:43side. I have total confidence over the
21:44next three years that as a basket of
21:47semiconductors of the optical companies
21:50of the transporta you know the
21:52industrial companies that make the data
21:53centers that over the next 3 years with
21:56humanoids coming and everything else
21:57that earnings will come to those
21:59companies and I have faith that relative
22:01to homebuilders and autos they're going
22:02to outperform which means those
22:04companies will go up. I just think if
22:06you're betting for it to happen at any
22:07month, you're gonna have a really hard
22:09time because hedge funds, I think, are
22:11playing the same game and they're
22:12crowded into the exact same trades.
22:14Today's episode is brought to you by
22:16Figure. If you're into Bitcoin, you
22:18already know selling's the last thing
22:20you want to do. But at the same time,
22:22life happens, man. Opportunities arise,
22:25bills come up, and markets move. That's
22:27where Figure comes in. Instead of
22:29selling your Bitcoin, Figure let you
22:30borrow against it so you can get cash
22:32without losing your position and may
22:34also avoid triggering a tax event. Right
22:37now, their loans are around 8.91%
22:39interest. That's 9.9% APR at a 50% loan
22:43to value. So, you keep your upside but
22:45still get liquidity. And I know the
22:47first question people ask is, "Yeah, but
22:50is it safe?" It's a good question.
22:52Figure uses MPC custody, which basically
22:55means your Bitcoin stays in its own
22:57segregated wallet. It's not being lent
22:59out, pulled, or sitting on some
23:00exchange. They also have optional
23:02liquidation protection, which helps if
23:04the market gets volatile. And if you're
23:06on the other side, like you've got cash
23:08and you want yield, they've got
23:10something called Democratized Prime.
23:11It's earning up to 9% APY, paid hourly,
23:15and is backed by real world assets, not
23:17just token inflation. So whether you're
23:20trying to hold your Bitcoin long-term or
23:21actually put capital to work, it's a
23:23pretty interesting setup. If you want to
23:25check it out, I've got a link in the
23:27description that you can go to or you
23:29can go to figurearkets.co/p.
23:33When you play sports, when you play a
23:34board game, there's always at certain
23:37times you want to play offense, other
23:38times you want to play defense. And
23:40actually understanding when to flip back
23:42and forth between those is pretty
23:43important. Mhm.
23:44>> What I love about having conversations
23:45with you every week is you come from a
23:47very different world than I come from,
23:48but we both have arrived at the same
23:50thing. And for years, I've always
23:52thought about these complex financial
23:55trends or developments are very hard to
23:58communicate. The phrase that I keep
24:00using is AI is offense, Bitcoin is
24:03defense. And it feels like if you go
24:06back to that board game analogy, you
24:08want to use AI to go disrupt the market,
24:11disrupt a competitor, disrupt something.
24:14And so that offensive move of using this
24:16technology,
24:18immense amount of opportunity to do
24:20that.
24:21>> At the same time, if you play 100%
24:23offense, you expose yourself almost like
24:26boxing, right? If you don't cover your
24:27chin, you can get knocked out.
24:29>> And so Bitcoin then becomes defense. And
24:32it feels like if you can use AI
24:34offensively to go and get revenue, get
24:36profit, you bring that back, now you
24:38have to somehow protect it from being
24:39disrupted and that's where you store it
24:40in Bitcoin. Is that kind of a good
24:43summary uh summary of your worldview as
24:45to how this comes together?
24:47>> So, let me use a real life example. Um,
24:49and I do want to give
24:51>> sylvia.com
24:54plug early early in the podcast. Uh
24:59um Peter, where did Peter where did
25:01Peter Teal just move to?
25:04>> Well, I know the truth, but go ahead.
25:05You can go with the uh
25:06>> what did he tell people? He was What did
25:08he tell people? Or what's the news
25:09reporting that he moved to Argentina?
25:12>> Yeah, he's in Argentina. Uh you guys
25:14Google his address.
25:15>> Okay.
25:15>> Um let's just use the reported news
25:18story and the reasons why.
25:20>> Okay.
25:22>> Um
25:24AI is going to take down the world.
25:26Humanity is going to be destroyed. Okay.
25:29Um, did he move there for economic
25:32opportunities? Assuming he moved there.
25:33Do you move to Argentina for economic
25:35opportunities?
25:35>> Malay is doing a great job, but I don't
25:37think so.
25:37>> Do you move there for safety? Do you
25:39move there for um
25:43I lived in Brazil. One of the reasons
25:46that people moved to Brazil
25:48shortly after World War II
25:51is because of the fear of nuclear war
25:54and everything happening and where they
25:56would be the safest and where to hide
25:57and where to go through it. So, I do
26:00believe and I've always believed this.
26:01This is where the debasement crowd gets
26:03mad at me on one side. Um, the
26:05programmable side gets mad at me on the
26:06other. I believe in a deflationary world
26:10accelerating right now. So instead of
26:12disruption, I believe in deflation is
26:14just going to tear apart everything.
26:17Literally everything. Leverage the whole
26:19thing. You don't need the leverage. You
26:22don't need the returns that are able
26:25that you're able to get from these
26:26things in a world where AI is coming to
26:29just be deflationary. So Bitcoin to me
26:32is more about where agents make
26:34decisions going, hey, I don't want to
26:36take risk today. where do I move my
26:37money for a store of value that is
26:40creating good returns and it's safe and
26:43or safer than everything else type
26:45thing. So I do believe that's happening.
26:48I want to make sure I answer one thing
26:50that you said before that fits into
26:52this. So when will Bitcoin have its next
26:54rally? Well, I do believe that comes to
26:57when the AI trade starts to have let's
26:59say its next bounce.
27:02either right during or after the
27:04hyperscalers report because it'll
27:06coincide with two things. It's happening
27:07at the end of the month. So people start
27:09realizing when did situal aware
27:12awareness happen? When did the last
27:14bottom happen in AI? It was at the end
27:16of the month. The end of months are
27:18really important when hedge funds are in
27:19distress because they play short stacked
27:23until there's a new month. And if people
27:25don't realize this, as someone who was
27:27in the hedge fund industry for a long
27:29time, the stress on people in December
27:33versus January,
27:35>> different ball game,
27:35>> different ball, it's Oh. Oh my gosh.
27:37When you're in December and you've got a
27:39big stack of chips and you're like, I
27:41don't want to take any risks this last
27:42month. So the end of this month is a big
27:45one.
27:46>> You have the midterms. So right around
27:48there also kind of coincides. So if I
27:50had to guess, I'm going to take two
27:52times where the the next AI
27:54>> second week of November.
27:55>> No, no, no. I would say right it usually
27:57happens before the event. So this would
27:59happen when the hyperscalers report and
28:01I do think there's an opportunity for
28:02the hyperscalers to because the stocks
28:05are trading so well. This ROIC thing to
28:08me is so stupid. The anthropic and open
28:10AI things last two weeks are just so
28:12stupid. Everyone please go listen to
28:13Jensen Yuan with Brad Gersonner on the
28:15BG2 podcast. I will be covering it for
28:18at least five or six slides in this
28:20weekend's video. Just go listen to a guy
28:23who actually knows what's going on and
28:25how simple he makes the whole thing
28:27about where the revenues are going to
28:28come from. They're not all about OpenAI
28:31and Anthropic.
28:33The second time would be December and it
28:35would happen in the month of December
28:38because that is the time when a lot of
28:41people just say I'm done with the year
28:43and it gets away from them going into
28:45next year. Those are the kind of
28:46channels where if I only think the AI
28:49trade is going to work like three months
28:50of the year, which is what I think
28:52maximum. It could happen three months in
28:55a row or it could like the S&P did this
28:57year or it could happen one month it
29:00goes up dramatically then nowhere then
29:03another month. That has been the nature
29:04of the game. And so I do believe that
29:06those are the two times for the rest of
29:08the year. So go for Bitcoin.
29:10I expect it to have its next move up to
29:13the highs. But I'm going to say again
29:16that my belief is that next year because
29:19of all of the things that are happening
29:20on tokenization between now and the end
29:23of the year, real stuff is happening,
29:25more and more people are paying
29:27attention to crypto. And the more people
29:29that pay attention to it, the more
29:30people that get involved in it, the more
29:32the reality becomes that there's money
29:34flowing into the space. And if money
29:36flows into the space, Bitcoin is going
29:38to do well as well. Now, when we sit
29:41down for this, we try our hardest not to
29:43talk all week. Uh, we do a good job of
29:45that. Yeah. Although, uh, it's sad. We
29:47We try not to. Um, we make a little
29:50list. Here's all the topics we want to
29:51talk about. This week's list does not
29:54have the CPI
29:57report on it. CPI came in quote unquote
30:01hot. Everyone was freaking out. You nor
30:04I seem to care nor think it's a big
30:07deal.
30:08you're talking about deflation and it
30:11struck me as I was reviewing the list
30:13three minutes before I came in here uh
30:16that neither one of us were planning to
30:18talk about CPI even though most people
30:20in the traditional world would think
30:21that is the biggest story of the week.
30:22>> Y
30:23>> why do you not care?
30:26>> So right now um core CPI is whatever
30:302.6%.
30:33Whether it's 2.6% or 3.2% or 2.2% 2%
30:38it's not going to help you make money.
30:40It doesn't matter. So if I come in here
30:43and say that interest rates where the
30:45level is is not going to be the thing
30:47that makes you money,
30:49it's just the reality of the situation.
30:51I've actually become numb to all of
30:53these things because they're so
30:55inconsequential to the reality of the AI
30:58trade. And that's the thing I just want
31:00people to understand. If only 20% it's
31:03really 24%
31:05of the stocks in the S&P 1500 have
31:08created alpha since Chachi PT a time
31:11when inflation was actually up at close
31:13to double digits. This has become
31:16insanity and I think the Fed makes it
31:19insane. I think history makes it insane.
31:22It really does not matter whether
31:24inflation is 2.3% or 3.2%. The only time
31:29in history of markets that something
31:31really bad happens
31:34is when we lose lots of jobs and the
31:37country goes into a recession because
31:39credit starts to unwind. That is not
31:43going to happen. So I have complete
31:46certainty that that will not happen. Can
31:49the stock market be down 10% over a year
31:52because the CPI surprises and goes up to
31:553.5%. Yes, it can. Is that the end of
31:58the world? No. I'm going to highlight
32:00this weekend that from today, if you
32:02look back over the last 10 years, the
32:04S&P has given you about 17% a year. If
32:07you look back over the last 20 years,
32:09it's greater than 10%. Last 30 years,
32:10greater than 10%. Last 40 years, last 50
32:12years, last 60 years, last 70 years,
32:14last 80 years, last 90 years, greater
32:16than 10%. Do you know how the majority
32:18of people get wealthy?
32:21Through compounding of money in
32:23investments. the S&P 500 that is 10, 20,
32:2730, 40, 50, 60, 70, 80, 90 double digits
32:30in total return for the entire time
32:32period. Why are you sitting here worried
32:35about people telling you that a bubble
32:37is going to unwind? Why? Now, if you're
32:4090, I get it. If you've got money in the
32:42market and you're 90, you may not live
32:45long enough to get the benefits of
32:46staying in the market. If the market
32:48goes down, as I told my kids, I'm like,
32:50that's where you stop spending money and
32:52you put more money into the compounding
32:53machine. You buckle down and you go,
32:55"Okay, I have a job." Everyone has a job
32:58in this country who wants a job.
32:59Everyone. You may not have the job you
33:01want, but you have a job. That's why the
33:03jobless claims are at all-time low. And
33:05when I say all-time low, so people know
33:07that number that comes out every
33:08Thursday, the weekly jobless claims,
33:11that is not adjusted for population.
33:13So, it being at 200,000 and sitting at
33:16200,000, that is the amount of people
33:18that have had to go claim insurance for
33:22being unemployed,
33:23that is the lowest number basically in
33:26history. And if you adjust it for the
33:28population, it's lower than it was when
33:31the population was half as much. It's
33:33lower than all of these other times. So,
33:35there's no job loss. And my grandmother
33:37taught me very well, like unless people
33:39are losing their jobs, they're still
33:42spending. and they might be spending
33:44more than they have and they're taking
33:45out credit cards. But when the real
33:46problems come up, it's when people start
33:48paying stop paying their debt because
33:51they're out of work. And if they're out
33:53of work, we start having housing
33:55bubbles. We start having the telecom
33:57companies because the telecom companies
33:58built stuff. They didn't have any
34:00revenues coming in the door.
34:02I know the revenues are coming into
34:04companies like Coreweave. I know the
34:06revenues are coming in because Anthropic
34:08is the fastest growing company in the
34:11history of the world. And what do people
34:13talk about? I think it may have slowed
34:15down for a month. I mean, are you
34:18worried
34:19>> about what?
34:20>> Oh, I This is exactly what I told
34:22everyone was going to happen. Today's
34:24episode is brought to you by
34:26investorhealth.com.
34:27You've built the portfolio. You've made
34:29the long-term bets. You still have to be
34:31around though to compound. Your body is
34:34the balance sheet. And most people never
34:35invest it in a way that they invest in
34:37everything else. Investor Health is a
34:40product.
34:41Investor Health is a company that we've
34:43built that is a clinician prescribed
34:46protocol for weight, metabolism, energy,
34:48and longevity. GLP-1s, NAD+, B12, all
34:53reviewed and prescribed by board
34:55certified clinicians, not picked off a
34:57menu. The process is a short online
35:00intake form, a clinician review, then
35:02medication compounded and shipped
35:04straight to your door. Most patients are
35:06approved and shipped within 48 to 72
35:09hours. No office visits, no referrals,
35:11and no waiting rooms. Ongoing check-ins
35:14at 30, 60, and 90 days, and a 247
35:17clinician support line. Health coaching
35:19is included as well. Plans start at $149
35:22a month. There are HSA and FSA eligible,
35:26and it's available in all 50 states,
35:28plus Washington DC. There's no
35:30contracts, and you can cancel anytime.
35:32Go to investorhealth.com
35:34and complete your intake form today.
35:37Investorhealth.com. Go long on life.
35:40Investoralth is a telealth platform, not
35:42a medical provider. Medical services are
35:44provided by affiliated licensed
35:46clinicians. Compounded medications are
35:48not FDA approved. Individual results may
35:51vary and treatment requires evaluation
35:53by a licensed provider. Terms and
35:55conditions apply. See
35:57investorhealth.com.
36:00Anthropic's growth rate is going to
36:01continue to decelerate and we can
36:02disagree on it because uh people are
36:05realizing that there is they they have
36:08the worst position in business possible.
36:11Ready?
36:12>> Okay.
36:13>> They have increasing competition from
36:16wellunded good entrepreneurial uh
36:19competitors.
36:20>> Disagree. Go ahead.
36:21>> Hold on. So it used to be open AI and
36:25anthropic.
36:26Now it is open eye, Anthropic, Grock,
36:29Gemini, Muse, etc. Everyone is now
36:33getting in the game. And then
36:34>> agree, keep going.
36:36>> And then that is just the major labs.
36:39But then you have all of these companies
36:40that are starting to come out and say,
36:42"Wait a second here." Coinbase comes out
36:44and says, "Hey, we're using open source
36:46models and we're training our own stuff.
36:47We don't need those guys anymore. We can
36:48do it better ourselves." We come out, we
36:50say that. Harvey comes out and says it.
36:51And so you have pressure from basically
36:54people saying, "Hey, I'm going to
36:55fine-tune open source." If you look at
36:57open router, you see massive
37:00acceleration in all the open source
37:01stuff. And so a company like Anthropic,
37:04which used to benefit from an oligopoly,
37:07that oligopoly doesn't exist anymore. It
37:10is now becoming decentralized and people
37:12having lots of options.
37:14>> Disagree.
37:14>> And okay. And
37:17what ends up occurring at the same time
37:19that you're getting competition death by
37:21thousand cuts from open source uh owning
37:23your own intelligence and the other
37:24models is token costs are falling off a
37:27cliff because there's massive pressure
37:29there as well. So when you see grock uh
37:33openai etc. So what you basically do is
37:35the equation that I use is number of
37:38tokens processed times price per token
37:41equals revenue. The number of tokens
37:44processed on a percentage basis going
37:45through anthropic on a per customer
37:47basis, my guess is is falling.
37:51Ramps AI index shows this. The rise of
37:54open source shows this. The increase of
37:57competition from major labs shows this.
37:58Then your price per token is falling as
38:00well. So they will continue to grow. I'm
38:02not here to say that they are going to
38:04stop growing, but the growth rate is
38:06going to go from, oh my god, this is the
38:08greatest thing ever to it will go from
38:10asymmetric to it slowly become more and
38:13more linear because not anything that
38:15they're doing. I actually think they're
38:17continuing to innovate. It's just the
38:19market pressures end up being good for
38:21us as consumers, but like that's a very
38:24different business. And we saw that
38:25their valuation people thought they were
38:27going to come out at two trillion. Now
38:29it looks like one one and a half
38:30trillion. It's a 25% decrease in
38:33potential valuation, $500 billion that
38:36they've lost in six weeks, eight weeks.
38:40>> So,
38:40>> you disagree. Go ahead. Have the floor.
38:42Professor,
38:43>> so Professor Jordan,
38:44>> we we've only had, I think,
38:48so let's separate when we have different
38:50views on things that are related to the
38:53next month or two or things like that
38:55and longer term views.
38:57One of the ones that we disagreed on,
38:59which to me is the exact same situation,
39:01is when you thought the budget would be
39:03balanced. Okay.
39:05>> Um,
39:06>> guy goes back to the one time that he
39:08got me.
39:08>> No, no, no, no. Not the one time, but
39:10when you're when you're taking a
39:12long-term view of something, you're
39:14making a long-term statement based on
39:16one month trend of something. And where
39:18we agree is there is no way. So,
39:21Anthropic has 10 times at the end of
39:24this year, they will be around 10 times
39:26where they were a year ago on the ARR.
39:28No question about it.
39:30>> They're on that pace still. Even if they
39:32do what they did from Q2 to Q3, they
39:34will be close to 100 billion of ARR at
39:36the end of this year
39:37>> allegedly.
39:38>> Okay. So, the beauty of that is they've
39:4210 times. If you ask me where I think
39:43they'll be next year, I don't know,
39:45three times, three and a half times. So
39:47we agree on the fact that it won't grow
39:49at the same pace that it has of course
39:51bigger company everything
39:52>> and and where I disagree and this is the
39:55most important thing. We haven't talked
39:57about this so I'm going to I'm going to
39:59I'm going to get to it in the end by
40:01comparing Sylvia to JP Morgan. Okay, in
40:04terms of what they're going to use.
40:06Okay. You're you're saying everyone's
40:08going to I'm telling you, I've worked at
40:09a bank. When you're spending 16 billion
40:12a year or whatever they're spending now
40:15in their technology, you are not putting
40:18open source as your big thing. That is
40:20not happening. No, no, no, no. It's not
40:23even I this one I won't even like let
40:25you go too far.
40:26>> If you're going to tell me exclusively.
40:29>> No, no, no. Ju, just so and maybe back
40:32up for a second.
40:33>> Forget all the competitions for a
40:34second. I can just tell you from my
40:35experience when we started with Sylvia
40:39using the uh models and I've gotten a
40:41lot of questions actually maybe it's
40:42help for me to understand it explain
40:43this we start out chatgbt wrapper so
40:46basically build a bunch of software you
40:47hook up your accounts when you hook up
40:48your accounts you can talk to AI and it
40:50feeds in the context window your query
40:52how to get my taxes down along with the
40:54context of your personal finances and it
40:56goes asset by asset and says okay this
40:58is a tax loss harvest here bonus
41:00depreciation here etc so it's better
41:02than Google it's better than just a
41:03regular LLM because you get the context
41:05plus that.
41:06>> We went from zero to $550,000
41:09per month of compute cost in like 5
41:12months or something. Like it was
41:14explosive.
41:16>> That was like an alarm going off saying,
41:18"Oh my god, every time I tweet and tell
41:21people to go use Sylvia, money is coming
41:23out of my pocket because they're using
41:25more compute."
41:26>> Now, the first thing we did was we
41:28didn't even touch the model stuff. It
41:29was just like architecturally. If you
41:31refresh a page and it hits the model,
41:33well, like when we refresh the page, we
41:34shouldn't hit the model, right? So,
41:36there's a bunch of like people were just
41:37trying to figure out how do you build
41:39tech uh architectural technology today
41:41with this whole model compute thing.
41:43Once we got all this this low hanging
41:45fruit stuff, then we looked at the
41:46model. I was not convinced to switch
41:49from the frontier models to open source
41:51until I realized that it was going to be
41:54lower latency and better performance.
41:56>> The cost is incredible. If we process
41:59500 million tokens on Anthropic versus
42:01our own hardware and software, we pay
42:0397% less. That alone should be a big
42:06incentive for people to go do it. But
42:08the thing that got me convinced was,
42:10wait a minute, you're telling me that
42:11the answers will be more accurate.
42:12You're telling me that they'll be
42:13quicker. It'll be a better performing
42:15experience for the user. Oh, and by the
42:18way, it's 97% cheaper. Of course, we're
42:20going to go this way.
42:20>> Mhm. So, what I think ends up being a
42:23huge headwind, which again people
42:26they're still trying to like kind of
42:27feel in the dark on this is I find it
42:30very hard to believe that JP Morgan is
42:31going to take all of their information
42:33and they are going to send it to
42:35Anthropic and they're going to say,
42:37"Please don't look. Please don't use
42:39this for training. Please don't do any
42:40of this stuff." and instead they have
42:44the option to go and take open source
42:46software and then they can train it on
42:48their own internal data architecture
42:51etc.
42:54There are still regulatory concerns,
42:55legal concerns, data privacy, security,
42:57like there's all this kind of stuff. But
42:59I actually think that this experience is
43:01more secure, more private, more
43:03regulatory friendly than we're going to
43:06use this open, you know, we're going to
43:08use the anthropic, open AI, etc. So to
43:10me, it's not just the cost. It is
43:13controlling. Think of it as sovereignty.
43:16JP Morgan wants sovereignty over their
43:18AI. They do not want to be reliant on
43:21Daario.
43:23Do you agree or disagree on that? Uh
43:25>> see the problem is your your your
43:28thought is coming from your experience
43:29at Sylvia.
43:30>> Yes.
43:31>> Okay.
43:32>> Maybe I'm blind.
43:33>> Well, here's the thing. You would die
43:35for the valuation of Sylvia to be equal
43:38to the tech budget of JP Morgan.
43:40>> Of course. Okay.
43:42>> JP Morgan
43:43>> a matter of time, Jordy.
43:44>> Exactly. So, but this gets into the
43:47difference between the two economies.
43:49And this is the part that I want to say.
43:50JP Morgan doesn't have time to wait. You
43:53mentioned Open Router. Who just bought
43:55Open Router?
43:56>> Stripe.
43:57>> What is Stripe's valuation? $200 billion
44:00about like 160 is the last one, but the
44:03people telling me who are thinking of
44:04investing is closer to 200. What's City
44:06Bank's market cap? 200 billion. How many
44:09people does Stripe have? 10,000. How
44:11many people does City Bank have?
44:13200,000.
44:15So Stripe bought Open Router. Why?
44:17They're not a They're not a routing
44:19company.
44:21JP Morgan's net interest income, did you
44:24see the number on their most recent
44:26earnings report?
44:27>> Tens of billions.
44:28>> It was 107
44:31billion.
44:33Okay, I want to put that number 107
44:35billion. So JP Morgan is going to be in
44:39a speed race to not lose that money
44:41because what does Stripe offer you? the
44:43ability of not you put your money in,
44:46you're getting paid yield as opposed to
44:48I put my money into JP Morgan, I get
44:50zero yield.
44:51>> They're borrowing money at zero and
44:53they're lending it out at four, five,
44:55six, seven and they're making a huge
44:56spread.
44:58There's two parts of a bank. There's the
45:01let me put it in the bank and keep it
45:03there. And then there's the okay, the
45:05bank is making money borrowing and
45:07spread. So I believe what Torson lock
45:09that AI agents will move that money out.
45:11I've said it publicly, I'll say it
45:13again. If you go from 2022 and you look
45:15at money market f fund growth, it's been
45:1870% growth since 2022, bank deposits
45:22have grown by 56%.
45:25The the the taking money out of the
45:27banks is already happening. So, the
45:28spread has gotten wider and they've been
45:31able to lend it out. The reason I bring
45:33that up is when you talk about the
45:35intelligence side, the only thing that
45:37matters to JP Morgan over the next 3
45:39years is the agent workflow.
45:43They need to basically
45:46deal with people cuz their spreads are
45:48going to be going down. So the game for
45:49profit margins is how does AI in the
45:52Fortune 500 companies the City Bank the
45:54only way city bank can compete with
45:55Stripe who has 190,000 less people
46:00they need to get rid of people and still
46:02do the business and keep their spreads.
46:04It is a very complicated game to do that
46:06with agents on the one side that are
46:08killing you because they are eating at
46:10your net interest income but at the same
46:12point you have to find a way to replace
46:14the people's workflows. So, you're
46:16dealing it from a position of no
46:18employees.
46:20You want the user
46:22to have a better experience and you want
46:24the company to find a cheaper way to get
46:26that experience to be good. I completely
46:27agree with that whole thing. I have no
46:29problem with the fact that that'll go.
46:31But in a Fortune 500 company, so let's
46:33separate the consumers and let's
46:34separate what I consider Sylvia as being
46:37part of. I consider this the tokenized
46:39digital economy.
46:41>> The digital economy, the Amazon verse
46:43retail malls, the Google vers
46:45newspapers, all of that rise has led to
46:47the most profits. So the part that I
46:50agree with is the winners of the next 10
46:51years are the people that only use open
46:53source. So we agree on that. But the
46:55enterprises will pay anthropic and open
46:58AI because of the advance the smarter
47:01intelligence and the the companies that
47:03have the most money they should be
47:05spending the most money on their AI and
47:07not taking a risk. Jensen Wang talks
47:09about all the time and he did on BG2
47:11that people don't want to take the risk
47:14of buying these chips that are cheaper
47:15when I can deliver them an entire system
47:18for 100 billion that will work because I
47:20have decades of of doing this or at
47:21least the last 5 years. The same thing
47:23will happen with anthropic and open AI
47:25at the higher levels of things, but
47:27gradually each company will do what
47:29you're saying, which is okay, I got to
47:31find a better way because I can't have
47:32these costs keep going up because my
47:34income is going at a faster pace. So, I
47:37do view this in a very similar way to
47:39you that the open source will win. But
47:40if you listen to Jensen Yuang, he said
47:42it perfectly. He goes, I think it's a
47:44mistake to view anthropic and open AI
47:46and focus on their revenues. They will
47:48be very successful of what they're
47:49doing. They have the best models. They
47:51have the highest intelligence. They will
47:53dominate research and science. No
47:54question about it.
47:55>> Agreed.
47:56>> They're not letting anyone see their
47:57models anymore, which means distillation
47:59goes down, and we're already seeing that
48:00in the China data. So, all the open
48:02source people that think they have a
48:03smart model, that's great, but not for
48:04math, not for science, not for
48:06rebuilding things there. I think all of
48:08that's going to start to take shape. And
48:10we're in the very early innings. But I
48:11think it's a mistake for people to
48:13doubt. The revenues will come in. I
48:16think they're going to come in from so
48:17many different sources. And anthropic
48:19and open AI may never meet the valuation
48:22of the company of what people thought.
48:24If we sat here in 5 years and anthropic
48:27and open a open AI combined were $4.5
48:31trillion. It was a waste of money to
48:33talk about the whole thing both from are
48:35they going to make it or go extinct or
48:37is it the greatest investment
48:38opportunity? We already missed the
48:39greatest investment opportunity.
48:41>> I I I actually think um they will
48:43continue to grow and and people are like
48:45are they going to be a great company?
48:46They're already great companies, right?
48:47So like this is not me like on
48:49them in any any form or fashion. What I
48:51think is uh and maybe to be explicitly
48:53clear the per company or per customer
48:58token consumption is going to keep going
49:00down over time because what they
49:04open and anthropic are the starter pack
49:07for a company to get involved in AI. You
49:10go and you adopt them first because
49:12>> you don't know nothing, right? You you
49:14need, hey, what is the thing to to get
49:16this going? I'm going to build software
49:17and then I need to plug in AI. They have
49:19intelligence. As companies move up the
49:21curve of I need to now customize it,
49:24improve it, optimize it, whatever. Then
49:26you start to move off of the general
49:28purpose models into the specialized
49:29models, the fetine tuning, the open
49:30source, like all that kind of stuff. And
49:32so for us, it's super easy to do that in
49:35a matter of 6 months. For JP Morgan or
49:37even large law firms or these big
49:39companies, maybe it takes 6 to 10 years,
49:42but the end result is they're going to
49:44end up wanting to own as much of that as
49:45possible. So it's sovereign, it's
49:47cheaper, it's better, whatever. I think
49:49that there is a window of time where
49:52the other aspect we haven't talked about
49:54is like Anthrop's got a sales team,
49:56right? Open source doesn't have a sales
49:57team. And so if you're the guy sitting
50:00at JP Morgan and you get to make the
50:01decision as to what you're going to do,
50:03well, I don't know, maybe the dude
50:04taking me to the Knicks game is going to
50:06get some business, right? And so like
50:08there is a whole other element of this
50:10that isn't just like what is the
50:11long-term best thing for the company.
50:13There's human relationships and all of
50:15this. So, I'm not arguing that they're
50:18zeros or they're not bad investments or
50:19any of that stuff. I'm just arguing that
50:22if I fast forward 15 years from now,
50:25most companies I think are going to want
50:26to own their full stack. Big companies
50:28are going to be slower. Small companies
50:29are already doing it. And so, the trend
50:31to me is like that there is way more
50:34headwind here than just like uh have we
50:37ever talked on this podcast about the
50:38fat protocol thesis?
50:40>> No.
50:40>> No. All right. So, let me go back for a
50:42second. Most of my view of AI comes from
50:45my experience on crypto.
50:47>> And if you look at all the people who
50:48are saying AI is going to kill us, they
50:50were telling us that the government was
50:51going to shut down crypto, right? So
50:52like there's a lot of parallels between
50:53new technology, whatever. And I was uh
50:56only now fortunate enough to have lived
50:57through one of these like zero to one
50:59moments in technology and now I'm seeing
51:00it. I'm like, oh, this must be like
51:02every technology.
51:04But in crypto, if you go back 12 years
51:06ago or so, there was this very famous
51:08fat protocol thesis. And the whole idea
51:10was all of the value acrru to the
51:13protocol. So the apps, the
51:14infrastructure, like all of that,
51:16there's going to be no value there. It
51:17all goes to Bitcoin, Ethereum, you know,
51:20Salana, whatever. Ended up not being
51:22true because obviously there's lots of
51:23value at the app layer, the
51:24infrastructure layer, and the protocol
51:26layer. AI is very similar in that no one
51:28has called it this yet but there's
51:30basically this like fat model thesis
51:32that people have had for a lot for a
51:34couple years which is like all of the
51:35value is going to acrue just to open AAI
51:38or anthropic
51:39>> I think now you know UI many others are
51:41realizing hey no no there's going to be
51:42apps there's going to be infrastructure
51:44there's going to be uh also value at the
51:46model layer and so the fat protocol
51:48thesis was wrong I think the fat model
51:50thesis is wrong as well but it doesn't
51:52mean that it's a zero it just means that
51:54it doesn't take 100% maybe it takes 40%
51:56of the value. There's still value in
51:58these other layers as well. And so
52:00companies have to choose where do they
52:02play in the stack. And like I'm going to
52:04put dollars out. Do I put it to the
52:06model companies? Do I put it to the
52:07infrastructure? Do I put it to, you
52:10know, open source? Where does my money
52:11go in that stack? It's not all going to
52:14go to the model companies.
52:15>> And I So you brought up the point that I
52:17think everyone listening should should
52:19be paying attention to, which is we're
52:21getting caught too much. And I think
52:23this is where our original conversation
52:25went on this, which is if you focus on
52:28the economics of something,
52:31you're just you're wasting your time.
52:32Meaning to think that anthropic and open
52:35AI are losing everything to open source
52:37is just stupid. Like it it doesn't make
52:39any sense. Do I think that the pie of
52:42TAM of intelligence is so big that
52:45everyone has a chance not to win but to
52:49survive? Yes. I think that's the key
52:51thing is that when Jensen talks about
52:54it, he literally says there's $90
52:56trillion going into
52:59compute and AI. He's been right about
53:02almost everything. And yet he still
53:05doubted. And that was the the part about
53:06BG2. If you go listen to Zuckerberg and
53:09Alexander Wang talk with on Cheeky Pint,
53:12it's the same thing. Go listen to what
53:14he said the benefits are for Meta over
53:17the Neoclouds. He didn't say the
53:19Neoclads weren't necessary. He's heavily
53:22involved with Coree. So is Nvidia.
53:25Nvidia supplies the GPUs. I mean, Jensen
53:28came out and said their cost of capital
53:29is actually getting lower. And if you go
53:31read on how their structure is set up,
53:33there's a lot of truth to that. So, I
53:36just think there's a separation for
53:37people from the economics of a business
53:39and the companies.
53:41Anthropic was a $50 billion valued
53:44company 18 months ago that is now valued
53:47at 2 trillion. Whether it's 2.4
53:50trillion, 1.9 trillion is irrelevant.
53:52The one thing I know for sure is to bet
53:55that you're going to get the same kind
53:56of return on it over the course of the
53:58next 3 years is silly. Like that's
54:00stupid.
54:01>> We're we're running out of time, but I
54:02have to ask you, did you see Elon's demo
54:06of the of his new cars?
54:08>> What are your thoughts?
54:10Um,
54:12you know, he I I've avoided talking
54:15about SpaceX. Uh, and I'm I'm I'm
54:18driving this into SpaceX for a reason
54:20cuz Elon has also said that um he sees
54:24SpaceX as being bigger than the global
54:26economy.
54:29I want to put that in perspective of how
54:31my brain finally got to the place that
54:33he could be right. Okay.
54:34>> Okay.
54:35>> Now, do I think that's going to happen?
54:37No. But do I could I come to the terms
54:40of what my vision is of the future and
54:42how Elon could be right? Here's the way
54:45it could be right. There are $900
54:46trillion on the asset. I've been
54:48thinking about the equity market as the
54:50total sum. So he was basically saying
54:52I'd be all public equity and I went well
54:54that's not possible. But what he's
54:56really saying is a world that I believe
54:58in which is the 900 trillion which is
55:01predominantly illlquid assets
55:04actually moves through tokenization into
55:06liquid assets. It's no longer trapped in
55:08this part. And that way you could have
55:10one company worth 50 trillion. I have no
55:13problem with it because Bitcoin to me
55:16same argument like could it be 50
55:17trillion? Of course it could.
55:20That is why when I saw the flying car
55:22thing and I saw some funny memes that
55:24came out from people saying, you know,
55:26talking about different things, I just
55:27went, I believe this. I saw a bald eagle
55:30three times outside my apartment in
55:32Brooklyn this week. So, if I can see
55:34bald eagles nesting outside of my uh my
55:37balcony, I think I can see flying cars
55:39everywhere.
55:40>> All right, I got some alpha for
55:41everybody. This is if you listen to the
55:42end of the podcast is where the good
55:43stuff is. Ready? I tweeted uh on
55:46Thursday morning, "As a shareholder, I
55:49would fully support a Tesla SpaceX
55:50merger. It is clear that Elon Musk is
55:53the best entrepreneur of our generation.
55:54Give him as many resources as possible
55:56and let him do his thing." Explicitly
55:59saying that I support it. Do you want to
56:01know what the alpha is? Who is at the
56:04top there who liked it?
56:08For those who can see, Mr. Elon Musk,
56:12which I think is kind of the subtle way
56:14if you've noticed, you know, he likes
56:15that tweet. He put SpaceX, X, Tesla,
56:20like he's probably going to merge them.
56:22I don't know if he does it this year.
56:23The rumor is that he could possibly do
56:24it before the end of the year, which
56:25would be pretty, you know, epic. Now,
56:27the reason I bring that up is maybe one
56:30of the alpha generating ideas from this
56:32podcast is it is likely that SpaceX
56:35would acquire
56:37Tesla and would have to do it at a
56:39premium to the current Tesla valuation.
56:41>> And if that's the case, then I don't
56:43know 15, 20, 30%, you know, kind of
56:46upside from there.
56:47>> And so my guess is that we are going to
56:49see Tesla's stock start to kind of grind
56:51up here because there are going to be a
56:53bunch of hedge fund guys who start
56:54speculating on, hey, is he actually
56:55going to do this or not? Yeah.
56:57>> And then on top of that, by the way,
56:58he's got a flying car that, you
56:59know, hops over traffic. And so you get
57:02these kind of like there's like
57:03fundamentals of the technology, the
57:05product, plus just pure speculation on,
57:07hey, is this guy going to merge these
57:08two? And if he does it at a premium,
57:11Tesla stock could run here into the end
57:12of the year.
57:13>> It could. And I think the Starlink um
57:16situation is really important. I think
57:17Elon has here's where I think he's
57:20become very smart in this thing. At the
57:23very beginning, we talked about LTCM.
57:25I want to make sure people understand
57:27that um at some point and this is what
57:30we've seen this year. The amount of
57:32capital needed for the AI trade from the
57:35public markets is enormous. This is
57:38another reason why tokenization has to
57:40happen faster. Um you need to free up
57:42capital from other places to where the
57:44economics can be distributed to places
57:46that people want. The the problem we're
57:47running into with the capital side is
57:50it's not that these trades are bad. that
57:52if Broadcom comes to the market for 100
57:54billion and Google comes to market for
57:55100 billion well then someone is buying
57:57that money they have a portfolio of
57:59investments you do I do I have a 100% of
58:02my investments in Bitcoin no and nor
58:04should anyone have all of their
58:06investments in one thing you
58:07>> scared you scared
58:08>> the the problem with AI and this is the
58:12issue if you keep coming with all of
58:15this what's the difference between a
58:17coreweave bond and a Broadcom bond and
58:20an Nvidia stock and they're all part of
58:23a very concentrated portfolio. So the
58:26amount of capital coming to the market,
58:29not the size of it, but the
58:31concentration of it is why the CDS blows
58:33out. It's why there's so much
58:35correlation with the names that are not
58:37winners of AI. And I think this is a
58:39very unique situation in time. And so
58:41what needs to happen is the capital
58:44needs to basically be digested
58:47and open AAI is coming to the market.
58:51Anthropics coming to the market. These
58:52are big things. SpaceX came fast. Like
58:55if you really go down, it starts to make
58:57sense. So what's happening to the
58:58market? Why are hedge funds blowing up?
58:59They're blowing up because they
59:01underestimated the impact that all this
59:03capital would mean to people who can't
59:05handle buying any more of it. They'll
59:07eventually buy more of it, but you know
59:09when they'll buy more of it? when they
59:10start having profits in their portfolio
59:12because then that delevers that position
59:14purely through the profits that have
59:16come through the gains and the other
59:17positions. That's what needs to happen
59:19for the next year. And that's why when
59:20Paul Tudtor Jones goes out and I don't
59:22know if he actually said this is some of
59:24the stuff, but he did say that it's
59:25going to be a negative liquidity year in
59:272027 because you're going to have so
59:28many companies coming to the market.
59:30Although I don't agree with him that the
59:31market will necessarily be down or that
59:34it's worth talking about in that manner.
59:36Uh I do believe that what is happening
59:39is there's an enormous amount of capital
59:40that is needed for this buildout and
59:42it's growing bigger and bigger every
59:44year. It won't take the market down
59:46because the economics behind it are
59:47supported, but it does mean that as an
59:50investor, you have to be patient with
59:51the time side and that goes back to the
59:53leverage. If you're leveraged and you
59:55need the AI positions in your portfolio
59:57to work for the next year, you might be
59:59in trouble.
1:00:01That's it folks. Make sure you go to
1:00:02Jordi Visser on YouTube. Subscribe to
1:00:04his channel there. If you want to use AI
1:00:06to better manage your finances, go check
1:00:08out silia silvia.com. It's all we got.
1:00:11See you guys next time.

About this transcript

This page contains the full transcript of Is Bitcoin Ready To Rip While Wall Street Breaks? by Anthony Pompliano, generated from the public captions YouTube serves with the video. The transcript has 11,588 words across 1,694 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.