YT YouTube2Text

YouTube2Text → Transcripts → Everyone Gave Up On Bitcoin At Exactly The Wrong Time

Everyone Gave Up On Bitcoin At Exactly The Wrong Time — Transcript

by Anthony Pompliano · 11,127 words · 1,559 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00And as we go forward, I'm going to say
0:01it again, the AI midcycle slowdown is
0:03basically the peak and the easy money of
0:05AI. And so if the money starts filtering
0:07back towards the other direction, I've
0:09laid out a pretty bullish case for
0:10Bitcoin, if you haven't got it, if we're
0:13overestimating the inflation side and
0:15productivity is coming. If we're
0:16overestimating the hawkishness of the
0:18Fed from what Worsh said, and at the
0:21same time the AI trade is not going to
0:23lose its earnings growth, it's just
0:24going to make it more difficult for
0:26institutions to stay long it the way
0:28they have in the past. Then all of a
0:29sudden the the lower volatility of
0:31Bitcoin will probably lead in if we get
0:34above the 200 day moving average with
0:35everything that I just said which is up
0:37up north of 70 at this point. I do think
0:40that that's going to be the beginning of
0:41the next phase of crypto and the next
0:43phase of AI. We'll see if it happens in
0:45the second half of the year. I think it
0:47will. What's going on guys? Today we've
0:49got a great conversation with Jordy
0:50Visser. It's 4th of July so he's here to
0:53explain to us what's going on in the AI
0:55trade. why he thinks that the midcycle
0:56slowdown is something you should pay
0:58attention to. He also describes why
1:00maybe some of the easy trades are gone
1:01and things are going to get a little bit
1:02harder from here. And then we talk about
1:04Bitcoin and why he's getting excited
1:06about Bitcoin once again. And then of
1:08course we go through a whole bunch of
1:09macroeconomic data points and he tries
1:11to unpack why exactly are some people
1:13ignoring these data points and how's he
1:15thinking about investing in his
1:17portfolio. All that and much more in
1:19today's conversation with Jordy Visser.
1:21All right, Jordy, happy 4th of July. a
1:23little Fourth of July edition here with
1:26the American flag hat. Uh let's start
1:28with Bitcoin. You know, a lot of people
1:29think Bitcoin is one of the most
1:30American assets you could have, but
1:32Bitcoin's not been doing very well, and
1:33it seems like people are starting to
1:35kind of give up a little bit online. Uh
1:37what's your take? Is that actually a
1:39good sign that you need people to give
1:40up to get to the bottom of a bare
1:42market?
1:43>> Well, first of all, I think give up is
1:45an understatement. Uh [laughter]
1:47I I it's very hard to find um people
1:50jumping into it. So, if you kind of took
1:53a poll of a hundred people that have
1:56never been in Bitcoin, a hundred of them
1:58will say, "I have no interest." If
1:59you've had people that have been
2:00involved, at this point, I would say at
2:03least 60 to 70% are questioning any
2:05involvement they've had. And so, you've
2:07got kind of hardcore people that are
2:09still in there. I I I'll just I'll give
2:11you, you know, my take in terms of this.
2:14This has been a um a bad year for
2:18for Bitcoin with a lot of headwinds. uh
2:22starting with software which we talked
2:24about a lot at the beginning of the year
2:25and how it got lumped in with everything
2:27that was built on code. Uh once the
2:30software stuff kind of bottomed you got
2:33into the what you said which is complete
2:36agreement a country like Korea which was
2:39part of the energy for Bitcoin moved
2:41everything into basically memory and
2:45we've seen massive leverage and margin
2:47debt and everything else. So, Bitcoin,
2:50which was always kind of the place you'd
2:52go to if you wanted to trade beta
2:54relative to tech, well, that ended and
2:56everyone now has beta in tech, where you
2:58can get 10 baggers uh in the span of a
3:00year. But then one other part that I
3:04don't think people um fully embraced as
3:07the last leg. And I talked about it
3:09briefly last week with this whole
3:10debasement unwide where it got lumped in
3:12with gold and silver. the rationale
3:15behind that which is starting to become
3:16more interesting to me and I'm seeing
3:18some technical signs finally uh with
3:20divergences which which I'll finish this
3:22off with but I I think that the basement
3:25trade um unwind there is a direct
3:29relationship here between something we
3:31talked about a lot and something I was
3:33concerned about for really the last five
3:36six weeks but then last week talked
3:38about that I'm not as concerned about
3:39anymore was this inflation uptick um
3:42particularly in core PC DCE core CPI
3:45with energy going higher. Well,
3:47obviously energy prices have collapsed.
3:49We've seen fertilizer. So, you're going
3:50to get a negative CPI print most likely
3:53for the June number. Already for July,
3:55we have a negative number. But I still
3:57hear a lot of people going out to macro
3:59clients talking about poor core PCE
4:02still being elevated. Then you get war.
4:04The second that we went from the Fed
4:07effective rate, which is about 3.6, six.
4:11We crossed over that for the December
4:13rate um of end of the year, meaning they
4:16had rate hikes built in. That was when
4:18Bitcoin had the last leg lower. I think
4:21we've hit the peak of that. I really do.
4:23Um you know, this is where the debate's
4:25going to come in. I am much more
4:27negative in my opinion on the jobs
4:30market getting better. And I think that
4:32is the difference. I listen to Darius
4:34Dale, who you've had on recently and
4:36who's a friend of the show and someone I
4:39respect, but I respect uh knowing what
4:42he's saying because I know his clients
4:43are responding to it cuz he's had such a
4:46good call on things and he was I just
4:49listened to him post something yesterday
4:50where he talked about wage inflation and
4:53I think that's where he and I disagree
4:55right now. Um I don't see the labor
4:57market as improving uh other than the
5:00fact that we've had more jobs. Some of
5:01this is related to the World Cup, but on
5:03the parts related to AI, there is no
5:05bounce whatsoever. Um the business side,
5:08the insurance, all of that stuff. And I
5:11think we people are underestimating the
5:13next part of AI, which is the
5:14application layer. The infrastructure
5:16side is going to slow down. So I'm just
5:18going to tell people technically we've
5:20got some nice divergences. Finally, we
5:22made new lows in Bitcoin, not just from
5:24the lows from a couple weeks ago, but
5:28also the lows from back in February,
5:31and it held in there. We've seen a lot
5:33of selling in the Bitcoin uh uh ETFs.
5:37So, I'm more interested in paying
5:39attention now. Like I've said, until it
5:41gets above the 200 day, I don't think
5:43it's worthwhile for people to to play on
5:45the other side. But, I do think the
5:46narrative is old. So now I've been able
5:48to find something where if the wage side
5:51doesn't improve
5:53uh meaning if we don't get a big change
5:55and we get some I think any weakness now
5:58that starts to show up from AI agents I
6:00think is going to be a a positive for
6:01Bitcoin.
6:03>> Now when we watch you know Bitcoin
6:04selling off like this it's coming at the
6:06exact same time that the Mag 7 have
6:08suffered to start the year and then on
6:10top of that we're now getting lots of
6:11debate in some of the AI names that
6:13people have been wondering you know how
6:15sustainable is it? Can this actually
6:17drive the returns or the ROI? Do you
6:20start to see even though there's
6:21divergence maybe from some of the things
6:23that used to be correlated? Are there
6:25new correlations that are starting to
6:26build or that you're watching where
6:27you're saying look maybe actually uh
6:29some of these assets are just rotating
6:31around in people's portfolio and they're
6:32looking at Bitcoin and you maybe some
6:34certain AI stocks or something as uh
6:37proxies for each other. Now,
6:38>> well, this is where the story gets very
6:40good for Bitcoin. Um, so about 4 weeks
6:43ago, I talked and I started doing videos
6:46on the AI midcycle slowdown. I don't
6:49think people really listened to me. I
6:51don't think they really cared. But these
6:53names have been hit pretty hard and
6:54they're going through a slowdown. The
6:57reason they're going through a slowdown,
6:59these headwinds are never going away.
7:02So, let me rephrase that. These
7:04headwinds are never going away. So, I
7:07think people have to adjust their
7:08expectations. And this goes for Korea.
7:10This goes for Micron. This goes for
7:12everything. We talked about Micron
7:13having blowout numbers last week. Stock
7:15is down a decent amount since that
7:18report. Um that was the high. Um when
7:21you sell off on great news, that is not
7:23a good sign. And you have to dig a
7:25little deep. The reason these headwinds
7:27or these that we're we're at a point is
7:29the headwinds are real. The government
7:31shut down one of the models, the Korean
7:34models are catching up. We're at a stage
7:36where the governments are getting more
7:38involved. This is a headwind that people
7:40have to understand. If governments are
7:41getting more involved with companies,
7:43Sam Alman proposed a 5% stake to uh the
7:47to the uh the White House. I I I I think
7:51we're getting back into what Bitcoin was
7:53meant to represent, which is when the
7:55government gets too involved in the
7:58capital markets, in the fiat system, in
8:00the ability to make money in these
8:01things, which is what happens when you
8:03shut down the models, when you slow
8:05down. On the data center side, we are
8:07absolutely having problems getting data
8:09centers in. This is again about
8:11politics. You've got a lot of people
8:13spending a lot of money in these
8:14elections. On the social democrat side,
8:18they are doing everything they can. On
8:19the Republican side and on the tech
8:21side, they're doing everything they can.
8:22We are at a point now where I think
8:24people have to realize that AI is a
8:26disruptive force and the models from
8:28here only get better. So what I was
8:30talking about in the first quarter,
8:32we're not going to see that easy trade
8:33again. It's not coming. And so the
8:36benefits are going to be to the
8:37companies that are able to use the
8:39agentic side and get their margins up.
8:41And we're starting to see health care
8:42stocks and financial stocks start to
8:45outperform the market. And that's why
8:47we're seeing this rotation because tech
8:49has been the big winner. The
8:50infrastructure has been the big winner.
8:52I still think over the next 5 years
8:53you're going to get get great returns in
8:55these things because the build that is
8:56going to happen. It's just not going to
8:58be as easy and I think that's good for
8:59Bitcoin. What do you think was happening
9:02behind the scenes for Sam Alman and
9:04OpenAI to be floating 5% stake in the
9:07government, right? Or for the government
9:08like it feels like that's not out of the
9:10blue. And so is that like the US
9:12government actually wanting to have the
9:15quotequote people, the citizens
9:16participate in the AI boom. Is that a US
9:20China competition thing? Is this a
9:23regulatory capture, you know, approach
9:25from somebody like an open AI? What's
9:27driving that? I honestly I think um
9:31you've got Daario and Anthropic on one
9:33side and they've made their point very
9:35clear that their their dealings with the
9:38government have not been um good.
9:41They're not donors to it. They're
9:43involved with some of the opposing
9:44forces in terms of the politics. Greg
9:47Brockman has donated a lot of money to
9:50um the Trump campaign and to everything
9:52on the Republican sides from what I've
9:54read. Uh so you've got these kind of two
9:57diverging forces. The here here's the
9:59thing I'll say about all this for
10:00everyone. Again, these headwinds are
10:02real. Um Open AI obviously came under
10:05some pressure for the spending. We saw
10:08Meta announce that they're going to
10:09become
10:11a comput company after spending billions
10:13of dollars on people. And the silliest
10:15thing I've I I may have ever heard in
10:18terms of a way a company just pivots
10:20left and right. Meta, we're not meta.
10:22We're going to spend tons of money on
10:23this now. we're going to become a a data
10:25center comput because SpaceX did it. I
10:27think all of these things show the same
10:29thing. The headwinds are coming and
10:30whether they're from politics that are
10:32getting involved and you're trying to
10:33have an edge on this because it is
10:35important or on the other side your
10:37investors are getting involved and
10:39saying you're spending too much money. I
10:41just think people have to recognize we
10:43are at the very early innings of the AI
10:45buildout. We were at the very early
10:47innings of the benefits we're going to
10:48see from AI, but we are past the
10:51anything goes period, which was probably
10:53for only about 6 months. I know it feels
10:55longer, but 8 months ago, people thought
10:57it was absolutely a bubble and these
10:58were hallucinations. Now, we know the
11:00models are too good. And so, you've
11:02reached a point where I think the
11:03government has to be involved and the
11:04model companies are getting uh I I I
11:07think they're starting to realize they
11:08have to work with the government uh or
11:11be in trouble. So we obviously saw them
11:13with Mythos and uh Fable. The Chinese
11:16open source models have really started
11:18to drive uh drive adoption. You can go
11:20look online and people are talking about
11:22hey we stopped using some centralized
11:24American model for this open source
11:26model. It's significantly decreased our
11:28cost and we're not worried about sec uh
11:30security or safety because we're just
11:32self-hosting the models ourselves. Now
11:34another aspect of this is with the open
11:36source models. Well, what is the model
11:38telling you? and is it, you know,
11:40telling socialist ideas or communist
11:42type ideas to your users or how how
11:44exactly was it trained? What was the
11:45source material? All of that stuff I
11:47think people are trying to figure out
11:48and and don't yet quite know. But there
11:51is a very small yet maybe exciting
11:54aspect of American open-source models.
11:58So Nvidia started to play here. There's
11:59a couple of other players. Do you give
12:01any importance or credence to American
12:04open source or is this pretty much
12:05American closed source versus global
12:07open source?
12:09>> American open source is going to be
12:10there. I mean I I think I think Nvidia's
12:13model I believe it's Neatron is already
12:16high enough on the scale for it to be
12:18useful but also remember any of the
12:20frontier model companies could release
12:22one of their older models tomorrow and
12:24it would be good. Um, I think this is uh
12:28I think people are getting a little bit
12:30too dramatic on things like the token
12:32index. I think they're probably getting
12:34a little too dramatic on how much the
12:36open source is going to win. The main
12:38story is going to be that along this
12:39path, companies are going to have to
12:40figure a way to keep their cost down.
12:42There there's no way around it. It's
12:44just too expensive to use the most
12:45expensive models. But this was always
12:47going to happen. We're It's just another
12:50recognition that 6 months ago there was
12:53not much adoption. I listened to Alex
12:55Canra Canerit's interview Greg Brockman
12:58and I thought it was a great interview.
13:00I really highly recommend people listen
13:02to it. Greg is very smart but he talked
13:04about uh enterprise adoption and where
13:06it was 6 months ago and 6 months ago it
13:09was just get me anything I need to catch
13:11up. Well, that's why we've they've
13:13overspent. And now what their business
13:15is is find me ways to turn it into ROIC.
13:19Like if I'm going to invest this money
13:21in in getting Open AI, you got to find
13:23some way to make this make me money so I
13:26can get some profit margins. And he
13:27said, "This is accelerating and their
13:29business is exploding on this front."
13:32The reason I bring that up is the token
13:33usage is going to just continue to grow.
13:36He talked about the fact, and this is
13:37what makes the meta comment so
13:39ridiculous. And for people that follow
13:41these things and get involved, I mean,
13:43it is honestly insane. I I spend my life
13:45listening to this stuff. There is not
13:47enough compute. So, if these guys say
13:49they have excess compute when they've
13:51done tons of major deals this year to
13:54secure more capacity, this honestly is
13:57just a way for them to get some cash
13:59flow in the door and make the investors
14:01think that they could always do this to
14:03put some floor on the stock. There's no
14:05doubt in my mind, but that doesn't mean
14:07they should see multiple expansions. So
14:08I think the model situation with the
14:10open source US will have open source
14:13models. The risk is that foreign company
14:16countries who've been shut off from
14:18fable and been shut off from others.
14:20[clears throat] Will they move to
14:20Chinese models which are way better than
14:23the US models at this point. So there's
14:25a gap in time and there's a gap in
14:26capabilities right now.
14:28>> If you were Mark Zuckerberg, what would
14:30you do? Would you continue to invest
14:32capex for your own internal consumption
14:34in the compute needs? Would you move to
14:36more of like this like neocloud type uh
14:38approach? Would you split and use some
14:41internal and rent some for the cash
14:42flow? How would you play it?
14:45>> So I I I don't even and anyone I think
14:48who who gives suggestions on this,
14:53they have a lot to lose. So you know, I
14:56think if you want to use something, why
14:58don't why doesn't Mark Zuckerberg go put
15:0120% of his money into Bitcoin? Because
15:03he has no need to. He has more money
15:05than he'll ever need, so he doesn't need
15:06to try to make money. The reason I bring
15:08that up in this story is with AI and
15:11with AGI sitting right around the
15:12corner, I I don't know the right answer
15:14for any of these companies. So, I
15:16understand why they're flailing all over
15:18the place. I mean, Microsoft is
15:19flailing. Oracle had a big rally. It
15:22dumped all the way back to lows again.
15:24We're in a point where the uncertainty
15:26going forward is extreme. And I think
15:28that is the part that we've recognized
15:30is that with all of these headwinds, the
15:33same uncertainty we saw in the first
15:34part of the year where software
15:36companies got hit, we're now seeing it
15:37with the infrastructure names because
15:39people are questioning whether the
15:40buildout will happen, what what will go
15:42on. And I think with Meta, it's just
15:45there's no answer to the question that
15:47you gave. And I think the investors are
15:48like, we don't know. And whatever Mark
15:50says, he says one thing, we're going to
15:52go out and spend billions of dollars on
15:54talent to play catchup. Well, that
15:55clearly didn't work. Maybe they have
15:58something behind the scenes, but I think
15:59when they do things like the data center
16:01sign, it's a bad sign to me that you're
16:03just kind of throwing stuff against the
16:05wall. The one thing we've learned with
16:07AI, and we certainly learned it with
16:09OpenAI, if you make any kind of a
16:12breakthrough, you do have the chance to
16:14catch up. So, there's a lot more talk,
16:16and I say this every week that we get on
16:18here, I use Codeex more than Claude at
16:21this point. I absolutely do. And the
16:23reason is because I'm not throttled as
16:25much. So I talked to a lot of people
16:27that's not the same case. But the more
16:29that enterprises are using it, they
16:31don't have as much compute. And I think
16:32that's where OpenAI has made the
16:34catch-up is they overspend already.
16:36Anthropic is trying to play catchup, but
16:38OpenAI has enough compute right now.
16:40Mark doesn't have anything ready right
16:41now. He doesn't there's no models and
16:44the data centers aren't finished yet. So
16:47all right [music] folks, I'm going to
16:48take a little break from this episode
16:49right here to tell you something about
16:51Figure Markets. Now I think Figure is
16:53super interesting. I've been an investor
16:54in this business for [music] a very long
16:55time. From the private market to now
16:57they've gone public. They built a very
16:59cool business. And I think what's most
17:00interesting here [music] is they have
17:02two sides to a marketplace. If you've
17:04got Bitcoin and you want liquidity, but
17:06you don't want to sell the Bitcoin, you
17:07want to keep the upside and just take a
17:09loan out against it, [music] Figure Now
17:11offers these cryptoback loans. They
17:13charge about 8.91% interest. That's 9.9%
17:16[music]
17:16APR. And you can borrow up to 50% of the
17:20Bitcoin that you give over to them. And
17:21so when you do that, you're unlocking
17:23capital, but you're not triggering taxes
17:25or giving up the Bitcoin exposure. Now,
17:27on the other side of this marketplace is
17:29somebody who's got cash and they want to
17:31get [music] yield. And on that case, you
17:33can actually go and you can use what's
17:34called the democratized prime product
17:37and it's allowing you to earn up to 9%
17:39APY. [music]
17:40What's cool about is they pay you
17:41hourly. It's backed by real world assets
17:44and there's no yield gains or this like
17:46token inflation that is susceptible in
17:47the crypto industry. So, if [music] you
17:50have Bitcoin and you want to borrow
17:52against it, go check out Figure. If
17:54you've got cash and you want to earn
17:55yield up to 9% APY, [music]
17:57you should also go check out Figure.
17:59They've been a great partner to me. I'm
18:00a very happy investor in the business.
18:02Go check them out at figurearkets.co/p
18:05[music]
18:06or click on the link in the description.
18:07Again, that's figurearkets.co/p.
18:11We've seen a lot of technological
18:13breakthroughs recently. Um, some of it's
18:15on the power generation side. Um we just
18:17saw um a nuclear uh company partner in
18:21some form or fashion with Nvidia. Um
18:24Etched just uh kind of came out of
18:26Stealth and they've got a whole bunch of
18:27innovation um around what they're
18:29building. How much of that innovation is
18:32going to hit the market in the next I
18:33don't know 12 24 months and will have a
18:35material impact here versus it took
18:38Nvidia 30 plus years to really become a
18:40household name and drive you know the
18:42value that they ultimately were able to.
18:44Is this around the the new tech,
18:46especially the hardware and chips? Is
18:48this stuff short-term or do you look at
18:49it much more long-term?
18:51>> It's long term. Um, there's no way to
18:54solve the compute problem from
18:57I mean, I can talk to a gazillion people
19:00and listen to as many podcasts and I try
19:03to cherrypick the ones with smart
19:05people. The OM podcast with Gavin Baker
19:07the this past weekend was great. There's
19:10just we're short compute there. I don't
19:12care who you talk to. There's no
19:13solution to that problem because right
19:16now to say that we don't have enough
19:17compute, we don't know what the demand
19:19is and again I I try to use these
19:22numbers but this is the reality. I keep
19:25hearing from everyone that demand is
19:27insatiable and the amount of demand that
19:29we have. So the only way I could say it
19:31is if you have one McDonald's restaurant
19:34and it says a billion served. Okay,
19:37great. Over how many years did that take
19:39to get to a billion served? If there was
19:40a line of a billion people outside of
19:42any McDonald's right now, how long would
19:44it take them to go through it? You can't
19:46just snap your finger with this stuff
19:48and get it. So, everyone can come up
19:50with, oh, there'll be more efficiencies.
19:51Oh, we're only we're we can't be in the
19:53third inning. Oh, blah blah.
19:56Unless everyone I'm listening to is
19:58completely wrong. We are so early in the
20:01stage. In fact, in fact, Brockman said
20:03right now 10 million users are using
20:07Agentic systems. 10 million. We have
20:11billions of people that will be using
20:13Agentic systems. Billions. We are barely
20:16in this no matter what. So, everyone
20:19just needs to woe down because once it
20:21gets easy, I mean, this is the reason
20:23why you can't have Siri on your phone do
20:25everything that you need. At some point,
20:26you will get that. We are still ways
20:29away. So, all these energy solutions,
20:30they're just making sure that we don't
20:32stop progress, but they're not going to
20:34make us be able to satisfy everyone's uh
20:36demand needs. If you look in your
20:38crystal ball, this midcycle slowdown,
20:42how bad is it? How long does it last?
20:44How does maybe the rest of this year
20:45play out in your eyes?
20:47>> I think we've done, you know, I I last
20:50week on the video I just showed that
20:52there are three different types of
20:54consolidations, and that's what we're
20:55in. We're in a consolidation of a bull
20:57market with inside the AI thematic
20:59trade. um they can be short and sharp,
21:02meaning they could have a a scary 20%
21:0630% correction in the names in like a
21:08day or two. During really strong bull
21:11markets where the market earnings are
21:13going up so much, that's less likely.
21:15They could have a very tight, quiet
21:17consolidation. Well, when it when
21:20something like Micron goes from 60 to
21:221,200 in the span of 15 months, that's
21:25also not likely for it to stay around
21:261,200 for the next 3 months and allow
21:28everyone to kind of stay in their trade
21:30and then just ride it higher. I think
21:32the the third one that I highlighted is
21:34the one we're in. I'm going to go
21:36through it again this week. And that's
21:37where you have more volatility where you
21:38just kind of go up and down in a much
21:40wider range. So, let's take Micron as
21:42just an example here. Um the 50-day
21:45moving average as of today is somewhere
21:47around 840. Um it made a high of above
21:511,200. So let's just say 1240 to 840. I
21:56mean that's a big range. We're talking
21:58somewhere around 50% of the price from
22:01where it was at 840. I think we'll
22:03eventually get to the 50-day moving
22:05average on most of these names. It'll
22:06either happen through a combination of
22:08falling and then bouncing back or by
22:10time continuing to go forward to let's
22:13say the next earnings period in July and
22:16then gradually will go higher. The one
22:17thing I I feel very strongly about is
22:19that the earnings for these companies
22:21are still going to be great. That's the
22:22underlying cushion that they have.
22:24That's why I feel very strongly to say
22:26this is about multiple compression. This
22:28is about making sure people are um not
22:32able to ride a trend for free. there has
22:34to be some volatility to shake people
22:36out. And so that's the way I think it'll
22:37play out for the next 3 months. Uh but I
22:40think by the time we get in the fourth
22:41quarter of this year, these stocks will
22:42be making new all-time highs for sure.
22:44>> Now, we've seen the S&P in the first
22:46half of the year up 10%, NASDAQ come up
22:4820%. What's your take? We just continue
22:51to go up and to the right throughout the
22:52rest of the year, and AI is a big driver
22:54of that, or will we get more divergence
22:56and and maybe some of these indexes
22:57actually will perform as well even if
22:59the AI stocks do? Yeah, I'm kind of torn
23:02between this. Um I I think it really
23:06comes down to the hypers scalers.
23:09They've been the biggest drag on the
23:10market. Um the S&P finished the first
23:12half of the year up about 7 and a half
23:15uh 8% I believe. No, I think it was
23:17closer to 10 actually by the time we got
23:19those two big days in. I think the week
23:21before it was around 7 and 1/2. But then
23:23we've had a pullback so far this week.
23:26We've underperformed Korea. We've
23:28underperformed Japan. we had even just
23:30barely performed with Europe. Um, I'm
23:33going to guess that because I expect a
23:35broadening out, we're already starting
23:36to see it in the healthcare stocks and
23:39the insurance stocks and the bank
23:40stocks. I think that's what's going to
23:42happen is I think you're going to see a
23:44broadening out into other places. The
23:46hyperscalers will probably be able to
23:48hang in the lows here even though I just
23:51don't see them having much upside. Uh,
23:53and as long as that's the case, I would
23:54expect another, you know, 10% by the end
23:57of the year in the S&P. I don't see
23:59anything on the horizon other than these
24:01headwinds becoming very political. I do
24:04think the midterms are probably going to
24:06be a negative for AI in general,
24:09particularly after the recent elections
24:11where Mandani uh uh candidates were
24:15doing so well. I think there's an anti-
24:18AI situation. And I don't think it's
24:20going to matter. And I think once we get
24:21to that period, it we should see
24:24inflation have come down. And if that's
24:26the case, then we don't have to worry
24:27about the Fed. And if we don't have to
24:28worry about the Fed and we start to see
24:30the dollar at least be on the weaker
24:31side again, I think we can be in a
24:33better position for all assets uh for
24:35the final four uh two months of the
24:36year.
24:38>> What are you looking for in the second
24:39half of the year? Are there specific
24:40milestones that you're either excited
24:42about or worried about?
24:44>> Uh well, it's really it still comes down
24:46to the earnings. I mean the Wall Street
24:48Journal had a good article um just
24:50highlighting how unique the first
24:52quarter was. So people have lowered
24:55let's say their belief in earnings and
24:58that's because there's been two factors
25:00that have juiced earnings which will
25:02continue to juice them on one side for
25:04at least uh you know this year but then
25:07the other one will run out after the
25:09second quarter. So just so we go through
25:12them. The first one is the reality that
25:15you get to um spread your expenses uh
25:18for the hyperscalers over the course of
25:20the next five years. Uh so the money
25:22you're spending uh you only have to hit
25:25against your earnings uh a little bit
25:27over the next 5 years gives you time to
25:29get your ROIC. But for the Nvidas and
25:32the people receiving the money, you get
25:34obviously a huge boost because their
25:36revenues jump up and that comes in. So
25:38there's a little bit of an accounting
25:40mismatch on this uh which in my mind
25:43because I believe that the profit
25:44margins will remain robust and will get
25:46better uh but slowly it you'll get the
25:50ROIC it'll keep earnings at a higher
25:52level but the margins probably won't
25:54expand as much as we saw there. The
25:56second thing is that a lot of these
25:58companies because they were owners in
26:00anthropic open AAI and SpaceX and they
26:02have to take marks on them well they get
26:04a big boost in their uh in their uh
26:07earnings. It's not in their operating
26:08earnings, but it's definitely in their
26:10earnings. That'll run out, too. I don't
26:12think those companies now that they've
26:13all been marked higher are going to go
26:15that much. But you could see Anthropic
26:16when it goes public. It could trade $3
26:18million$3 trillion. There's no reason
26:20why it can't. Even if it didn't stay up
26:22there after we saw what we saw with
26:24SpaceX. So regardless, I think that
26:26means that the benefit or what I'm
26:29looking for milestones is that we
26:30actually see a broadening out. If you go
26:33read what Travelers said in their
26:34earnings report and you go through it,
26:36it's an insurance company. It's boring.
26:38It doesn't really matter. They're really
26:41starting to talk about AI in a bigger
26:43way. I've talked about Eli Liy on this.
26:45Those are two healthcare example or the
26:47two uh companies just in general. One in
26:49the insurance side, one on the
26:50healthcare side. I think people need to
26:52start realizing that the benefits from
26:54AI really come with the agentic side.
26:57And we only had the realization of the
27:00AI agent side in the first 3 months of
27:02this year, the first 6 months. As Greg
27:04Brockman said, he's seeing massive
27:07adoption and now Open AI is getting
27:09heavily involved in showing them how
27:10they can have real impact. He said the
27:13second half of the year is going to be a
27:14big gamecher uh in terms of the AI
27:17agents and particularly looking a year
27:18backwards. I think people need to start
27:20to build that into their uh their
27:21estimations and I I think that's what's
27:23going to happen. So I think there's a
27:24bit underneath the market and there's no
27:26AI bubble. It's the productivity boom
27:28that's going to start to show up.
27:30>> What I find really interesting is take
27:32um claude artifacts, you know, as an
27:34example or or something like that where
27:36people are now starting to take these
27:38agents and put them into their company
27:40workflows. There's a lot of debate as to
27:42are you actually giving the information
27:43to the models and they're just going to
27:44go compete with you. At the same time, I
27:47think there's a lot of people saying,
27:47"Hey, this is going to make us more
27:49productive." And so how do you just see
27:51this like agentic component? If you were
27:54running a company, you had, you know,
27:55thousand employees and your employees
27:57were saying, "Hey, this is going to make
27:58me more effective." You were seeing the
28:00revenue growth. What would your
28:01evaluation be as to whether you let the
28:04major LLM into, you know, the kind of
28:06hen house, right? Do you do you do it or
28:08do you worry about it? How do you look
28:10at that?
28:12I mean I think they've gone through this
28:14in a way that aside from the risks that
28:17come from uh the cloud as not being as
28:20safe. You they'll be using I mean
28:23they'll be using for example AWS and AWS
28:26will have their own security around it
28:27and that's one of the reasons why when
28:29people go well Amazon told on meth well
28:32no they want to make sure their job is
28:34security. Um, if a company like
28:37Coca-Cola or Eli Liy or whatever is
28:39using their cloud and they believe that
28:42Mythos is something that is an issue,
28:45they want to make sure that everything
28:46is tied up. The one of the key things
28:48that came out in the insurance side was
28:51how Mythos is actually a big positive.
28:53And I hadn't thought about this. Um, we
28:55hear about the vulnerabilities, but if
28:57companies like Travelers, which sells
28:59insurance, and part of the thing they
29:01want to check is, do you have risk
29:03inside your um your company, well, now
29:06they can have Mythos go check
29:08everything. So, it's not just does the
29:10company paying for it, the insurance
29:11companies and the consultant side, the
29:13Marsh Mlons, all of that stuff, they can
29:16now get involved with, well, how do you
29:17deal with this from a policy
29:19perspective? I I just don't think people
29:21have fully thought. And the reason is
29:23because it moves so fast. It was a year
29:25ago that people were still bothering me
29:26about hallucinations. You barely hear
29:28hallucinations now because we're talking
29:30about the models with 140 plus IQ. So I
29:33just think as we go through this
29:35companies will feel more secure because
29:36they are going to have the Amazons and
29:38everything uh as a defense. That being
29:41said, the amount of times being
29:43mentioned about cyber risk for smaller
29:47businesses, I do think there's going to
29:49be uh a market shock within the next
29:52year where we might come in in a day and
29:55the market could be down 10% globally
29:58because there was a hacking into a ma a
30:00major bank or a major place and whether
30:03it was IP that was stolen or whether it
30:05was um you know private information
30:07that's going to be released. I just
30:09think we're going to have that and
30:11people are just going to have to um deal
30:13with it. But I think those types of
30:14things continue to make it where if
30:17you're running a portfolio where you you
30:19can't handle volatility. I wouldn't have
30:22too much money in the market um because
30:24I think it is going to be volatile. I
30:26think you're going to get paid over
30:27time, but I do think that there's going
30:29to be a lot of shock waves and that the
30:30easy part of AI is over and the model
30:32capabilities have just gotten too good.
30:35There's been a lot of talk of for many
30:37of these cyber security reasons etc the
30:39US government coming in regulating
30:41approving not only the models getting
30:43released but who they get released to
30:44who the customers are you know there's
30:46many levels to the regulation um at the
30:49same time somebody like a bill gurly
30:51many others are saying wait a second
30:52this means that China or you know
30:55whatever model uh whether it's open
30:57source or another geography they're
30:58going to get this massive advantage
31:00because we're essentially shooting
31:01ourselves in the foot which side are you
31:02on on the debate
31:05I I there's no way that the Chinese open
31:08source models aren't going to um hurt
31:11it. It it it has to. Um
31:14[snorts] I I I've kind of thought about
31:16this from the way that I think about
31:18everything over the next 5 years. So
31:21humanoids are coming and part of the
31:24reason that that gets brought up in the
31:26conversation is you can't have humanoids
31:28without having
31:31super intelligence. Well, we're going to
31:32have super intelligence in 5 years and
31:34we're going to have humanoids in 5 years
31:37and that combination is going to change
31:39the the equation again. So, as much as I
31:42want to believe that opensource is going
31:45to hurt or help, I I just kind of look
31:50at it as the total revenues of the S&P
31:53500, a fairly high percentage of them
31:55come from overseas. Uh I forget where
31:58the number is now, but let's assume it's
31:5940%. It's somewhere around there. That's
32:02a lot of revenues, meaning the US is
32:04selling into all of these countries and
32:06almost everything that we're selling is
32:07services and software and things along
32:09those lines. Now, the hyperscalers have
32:10a huge part of it. So, do I think
32:13open-source uh and and LLMs are going to
32:16be critical towards
32:18maybe keeping that? I I don't see any of
32:21the other countries as being in a
32:23position where they can honestly use the
32:25Chinese open source models in the way
32:26that they want. They're going to need
32:28the US for something. And if we learned
32:30a lesson last year, everyone's
32:32interconnected already. Um, to say that,
32:35okay, you have our open source, we're
32:38shutting it down. The only two countries
32:40in the world that seem to be at a
32:42stalemate of how to deal with things is
32:44China and the US. Uh, the US was
32:47throwing tariffs on this, threatening
32:49invasions, all kind of stuff. And we
32:51ended up just kind of getting through
32:53the tariffs and then the Supreme Court
32:55said, "No, there are no tariffs." So, I
32:57think this is going to be a very similar
32:59thing. I really do because I don't
33:00believe that there's a time where AI
33:02gets so good that everything just ends.
33:04I'm going to continue to lean that the
33:06frontier models are being built to solve
33:10the world's greatest challenges. That
33:12that is basically the Manhattan project
33:14for every single great problem you want.
33:17If you go listen to that Greg Brockman
33:19interview and you and you fast forward
33:20all the way to the final 10 minutes, he
33:22starts talking about cancer. He starts
33:25talking about people being able to solve
33:27this and what as has already happened
33:29and he said this will be the standard.
33:31So if you haven't listened to me or
33:32don't believe me in terms of like the
33:34ability to have to not die from disease
33:39he doesn't do it from the Daario way. He
33:41does it from the it's already happening
33:43way. And so I don't think open source
33:45should be the you know the place that
33:47people debate. I think for Bill Gurley
33:50he's a VC guy. I think you're going to
33:52start to see capitalists and the
33:54capitalist system, and this is the
33:56reason why I got involved in Bitcoin in
33:58the first place. I think you're going to
34:00see major people start to question what
34:03AI means for the capitalist system and
34:07for their money. I really do believe it.
34:09Um, and this is the reason why I
34:11continue to focus my attention on
34:13outside the system. So to sum it up, if
34:15Bill Gurley uh and and people like Bill
34:18Gurley are saying, "Well, Chinese China
34:20it open source will always win." I kind
34:23of agree with it, but not in the way
34:25that that means China wins. It means
34:27that small businesses win, that
34:29entrepreneurs win, and that pl the
34:31places that have the most money don't
34:33win. And that's kind of the way that I
34:35see it happening is that there's a
34:37distribution of economics that happens.
34:41And I that's the reason why I think
34:42people are going to look for a place to
34:43be outside the system and protect their
34:45net worth. And I don't think it's the
34:47government taxes. I think it's AI.
34:49>> You were banging the drum on Micron,
34:51memory, etc. back, you know, sub$100
34:54Micron stock. Um, now it seems like
34:56everyone is talking about this. There's
34:58everything from inference to, you know,
35:00podcast and and um we recently saw the
35:03president tweet about Micron. Uh we have
35:06seen um Gavin Baker, you know, kind of
35:09beat the drum as well. you wrote a paper
35:11this week on the importance of memory
35:13and and kind of this belief that even
35:15though everyone knows this is important
35:17and it's a bottleneck, it is not going
35:19to be solved seems to be the consensus.
35:21Can you just elaborate on that?
35:23>> Yeah, I I just wanted to make sure
35:24people realize that um memory is
35:26[clears throat] the most important
35:27thing. And I used a book that I don't
35:30think we've talked about and I don't
35:31mention too many books cuz I don't I
35:33don't read um really
35:35>> You read a book? You read a book? I I've
35:37read I've read a decent amount of books
35:39in my in my lifetime. Uh but not not
35:42recently. But Moonwalking with Einstein
35:44is a fantastic book. And I wanted to
35:48bring it up because I was always f I get
35:51fascinated by little nuggets. But
35:53knowing that Greek philosophers were
35:55able to memorize speeches that would
35:57take four hours, no teleprompter, and
36:00they could do the whole speech. And they
36:02did it through this um well I'll use the
36:05the technique the way it's called in
36:07moonwalking with dimes time which is the
36:08memory palace which is somewhat similar
36:11to the way we all remember songs that we
36:13haven't heard since we were 10 and all
36:14of a sudden it comes on the radio and we
36:16know every single word but you and I
36:18could ask each other what we spoke about
36:20on this we'll have no idea tomorrow. Uh,
36:22so how does a song stay in your brain?
36:26And memory is critical for the IQ to
36:29continue to go up. Like you have to have
36:31more memory because at some point you're
36:33going back and forth. You're connecting
36:34dots. And the only way your IQ is going
36:36to go up is the ability to retrieve
36:38things quickly and go through this. So I
36:40wanted to number one write a paper that
36:42just highlighted to people that watching
36:44the way Micron and SKH Highex and
36:46Samsung trade is very very important to
36:49what's happening in AI.
36:51it's going to be a shortage for a long
36:53long time. Um, we're just tipping the
36:55scales. Number two, I wanted to make
36:58sure that as people go through this,
36:59they realize we are entering the age
37:02where this IQ and this memory is getting
37:04us to a point where magic is going to
37:06start happening. And that was the point
37:07of kind of connecting it back to
37:09moonwalking with Einstein. So uh I just
37:11think that uh for people that are trying
37:13to understand AI and understand what's
37:16going on and realize that memory is not
37:18a bubble. Memory is an insatiable need
37:21and eventually we'll get more
37:22efficiencies uh in algorithms. We'll
37:25eventually build more memory. My point
37:28is I think from this point on people
37:30should expect normal type memory returns
37:32of maybe 30 to 40% a year and not 30 to
37:3640% a week. Uh, and that's why we're
37:39starting to see the girrations. That's
37:40all it was.
37:42>> Today's episode is brought to you by
37:43Mogul. $4 trillion. That's how much real
37:46estate Deote says moves onto the
37:48blockchain over the next 10 years.
37:49That's good news for the rest of us
37:51because this year the VIX, Wall Street's
37:53fear gauge, it [music] ripped past 30.
37:55The S&P went red and stocks repric every
37:58time a politician tweets. Real estate
38:00[music] doesn't do that. Tokenization is
38:01the key to giving everyone access to
38:03real estate. Here's what most [music]
38:04people don't realize. You can invest in
38:06tokenized residential real estate right
38:08now thanks to Mogul. Mogul's founders
38:10come from Goldman Sachs where they
38:11deploy [music] billions into real estate
38:12for institutional clients. Now they
38:14handpicked the properties and handled
38:16the management for you. Historically
38:17[music] an 18% average annual IRRa
38:20that's 3x better than their competitors.
38:22You get targeted [music] yields over 8%
38:25monthly rent payout, tax benefits from
38:27depreciation, and zero 3 a.m. tenant
38:30calls. Over 100 properties [music] have
38:31already launched on Google. More than
38:3330,000 people have already signed up,
38:35too. If you want in plus get a special
38:37offer, head over to mogul.club/p.
38:41Again, that is mogul.com/p.
38:44You can see important disclosures in the
38:46description and [music] head over to
38:48mogul.com/p.
38:50Today's episode is brought to you by
38:52Blowin. If you're actively trading
38:53crypto, then you already know the
38:55platform matters. [music] Execution
38:56speed matters, liquidity matters, and
38:58reliability during volatile markets,
39:00that definitely matters. That's why more
39:02traders are starting to use Blowfit.
39:03They built [music] a trading platform
39:04focused on fast execution, deep
39:06liquidity, advanced futures products,
39:08and a really, really clean user
39:09experience. They don't over complicate
39:11things. You can trade hundreds of spot
39:13and perpetual pairs. [music] They
39:14integrate directly with Trading View,
39:16and they've built a platform for both
39:17active traders and people who are just
39:19getting started. Security is a major
39:20focus. Obviously, assets are custody
39:22with Fireblocks. [music] They provide
39:23proof of reserves, and they've built a
39:25strong reputation for platform stability
39:27during those high volatility
39:28environments. And to celebrate the
39:30partnership with us, Blofin [music]
39:31is now giving away $100,000. That's
39:34right, Blofin is going to give away
39:35$100,000 in deposit and trading rewards.
39:38So, if you're looking for a better place
39:39to trade crypto, go check them out today
39:41using our link in the description.
39:43Again, that's [music] Blowin, the
39:44trading platform built for serious
39:46crypto traders. Go sign up. Pay
39:48attention [music] cuz you're giving away
39:49$100,000 in deposit and trading rewards.
39:53Now given that it is July 4th um it's
39:56the 250th birthday of the country I
39:58think people are very um excited and and
40:01there's a lot of patriotism but also
40:03there's another side of this which is
40:05like the future of AI is going to have a
40:08material impact on financial markets on
40:10people's individual portfolios and on
40:13what the economy and society looks like
40:15you know over the next 10 to 20 years um
40:18we talk a lot about the software side
40:21and I think rightfully so but you
40:22mentioned earlier uh kind of the
40:24physical AI, the robotic side and I
40:26think that that is probably going to be
40:28one of the areas where the average
40:30American first starts to see this stuff.
40:34What do you envision over the next 10 or
40:3620 years? Is this like uh you know
40:38technological utopia where we're all in
40:40self-driving cars and humanoids are
40:42walking our dogs and you know every
40:44factory is completely automated or is
40:46there some like messy middle between
40:49that vision and where we are today that
40:51you think is much more likely and then
40:52like what the ramifications are for
40:53investors as they're trying to navigate
40:56you know what are the next trends that
40:57maybe aren't as popular today but will
40:59become popular over the next 10 years.
41:01No, this is going to explode over the
41:03next 5 years. And anyone that doesn't
41:05believe that is um is in the same hole
41:09they were with AI hallucinations. Um so
41:13let's just first put it into terms with
41:16the Chinese.
41:18What do people think is a bigger risk?
41:20Right now the Chinese models are only um
41:226 months behind our models.
41:26And yes, most of this is through copying
41:29them through distillation and then doing
41:31some changes. [clears throat] But the
41:32reality is their models are as good.
41:36Their robotic situation is far far far
41:39advanced over us. Now, not necessarily
41:41the capabilities, but the production
41:43numbers, the ability for society to
41:46embrace them and want them and use them.
41:48Um, China just in the last couple months
41:52announced that um, and I forget which
41:54company is, they have 140 different
41:56humanoid robotics companies there and
42:00they're now releasing uh, one for less
42:02than $5,000.
42:04So again, how are we going to win a war
42:07if the Chinese are embracing robotics
42:10and humanoids and we're not? So for
42:12anyone out there that doesn't believe
42:13this also won't be a military situation,
42:16believe Elon Musk when he says there
42:17will be millions of them uh certainly
42:20within the next decade. But the concept
42:22of them building themselves is already
42:24happening. And that doesn't mean again
42:26them going out and putting the parts
42:28together. It means them doing what
42:30humans do right now to create them. And
42:31that's already being done in Tesla and
42:33and and other shops that are going
42:35through this. So the scaling out of
42:37humanoids is going to happen. But I want
42:38to bring this back to it's already
42:41happening in society. We already have a
42:43car that has done flying flying car
42:45that's done trips from the west side of
42:49Manhattan to JFK. Uh that's already
42:52happening. There's drones already
42:54working in Florida going in front of the
42:57police to go to situations. We're going
43:00to have ambulances like all of these
43:03things. is the first place it's going to
43:04happen is in places where people won't
43:06fight it because it's an emergency
43:08situation for crime where someone's
43:10getting hurt for an ambulance where
43:12someone's getting hurt. Full
43:13self-driving will happen. The humanites
43:16inside our home will take a lot longer
43:17to go because those have to be perfect
43:19to go through. But again, this gets back
43:21into being able to solve something on a
43:24computer textually is nothing compared
43:26to the compute necessary for a humanoid
43:28to be making good decisions. For anyone
43:30who's been in one of the older Teslas,
43:32which I have, and has used the full
43:35self-driving, I would never let my hands
43:38off it for very long in Maine. Um, these
43:41roads are not on the map perfectly.
43:42Every there's things there, the bikes
43:44are all over the place. Like, I I just
43:46wouldn't do it. Um, and the newer
43:48models, I'm sure it's much much better.
43:50And as I get another car and I get the
43:52new models, I'll go through it. But I
43:54think for people not to realize that
43:56this is coming, it's going to be a big
43:57part of our lives. And it's why they
43:58should never fade the hardware
44:00situation. The hardware buildout, the
44:02parts is the place to be. Memory is part
44:05of that. The infrastructure part is part
44:07part of it. This is going to be a $90
44:10trillion buildout. And I think you want
44:12to be focused on a bull market in stocks
44:14because that buildout is going to have
44:16the money that's going to be spent to do
44:18those things.
44:19One of the aspects um that has really uh
44:22I think gotten for been forgotten in the
44:26AI conversation is the impact of the
44:29Fed. And we've been doing all of this
44:31with you know somewhat elevated interest
44:33rates. And so if wars does come in and
44:36start cutting rates and driving all
44:39kinds of money into the market that
44:42should provide cheaper capital to these
44:45companies to then go and use that to
44:47build out some of this stuff. And like
44:48there's a flywheel there. The question
44:50just is like what would the inflationary
44:53aspect be? You know, what would the the
44:55pain or the trade-off? Oil prices have
44:57come down. What's just your general
44:59sense right now on like war inflation
45:02and then the ability for macroeconomics
45:04to help or hurt these AI companies?
45:08>> Well, first of all, I think people
45:12completely overreacted last week to to
45:15Worsh's first statement. Um, and and
45:18I'll I'll just say I I like looking at
45:21the movements and things and I think the
45:22way the media and especially people in
45:25the macro world talked about it like oh
45:27it was hawkish. You just didn't get a
45:30big enough move in anything to say
45:32anything happened. There wasn't like
45:34some massive move where it was a reset
45:36where he said this is going to happen.
45:37In fact, he said I want less guidance
45:40from the Fed. I don't want to be coming
45:42out there and trying to stop volatility.
45:45Uh, I think it's important for us to get
45:47back to monetary policy being the way
45:49that we're moving things because that is
45:51the only way that we're going to be able
45:52to get
45:54a change in what's happened. So, I think
45:56one of the messages that he said is that
45:58he would love to reduce the balance
46:00sheet. Now, I personally don't see how
46:03that's going to be possible. Um, and he
46:06did admit in an interview this week uh
46:08on a panel where he said it took us 18
46:11years to get into this balance sheet.
46:13we're not going to get out of it in 18
46:14weeks with the implication being that
46:16this is going to be a slowmoving thing.
46:18And he said we are not going to shock
46:20the market. We will give them plenty of
46:22time to know that this is going to
46:23happen. But at the same point, he wants
46:26rates to be lower. And the reason he
46:28wants rates to be lower is because he
46:29knows that they have an impact on
46:31housing. They have an impact on people
46:33that maybe don't own assets and still
46:35need to borrow money. And so all of
46:38these things fit into a Fed that
46:40probably will be much different than the
46:43past and I think the market has to
46:44adjust to it. But the most important
46:46thing that he said both last week and
46:49also this week which I want people to
46:51understand I very seldom meet any macro
46:54person any macro person who understands
46:56artificial intelligence. And when I say
46:59understand it I mean uses it
47:02every day for hours a day. And the
47:04reason that's important, if I say this a
47:07hundred times, I'll end up saying it a
47:09million times in my lifetime for every
47:12hundred times I say it in a day. If
47:15you're not using the model, if you're
47:17not understanding how powerful it is,
47:20how can you possibly understand the
47:22impact it's going to have on
47:23productivity? It is productivity that is
47:25the most important change that is going
47:28to go on and it has already started and
47:30it's showing up and I don't even think
47:31GDP is properly measured. So, we're at a
47:34point where what Kevin Moore says is, "I
47:36don't know what a AI is going to do."
47:38But in this weekly uh this interview,
47:40which anyone can get the transcript from
47:42YouTube, and I'll go through it this
47:45weekend, he literally said, "We are at
47:47the beginning of an exponential change
47:49in AI." He's admitting I don't know what
47:53that means in productivity, but I do
47:55know that the history of Gen Greenspan
47:57when one of these technologies come is
47:59not to overreact to the inflation data
48:02today. That is the message that I heard.
48:05Now, that does not just go with oil. And
48:07I think this is the mistake people are
48:08making. He's also talking about core
48:10inflation. if he believes that AI is
48:13going to take down core inflation. He
48:15talks about the job market and he talks
48:17about it in a way that it may hurt jobs
48:20today, which I agree with more than it
48:24will in a couple years because the
48:26displacement takes time. People lose
48:28their job and then you got to go find
48:29another job to get hired. But if you're
48:31fired as an accountant and all the
48:33accounting jobs are leaving, what do you
48:36go become? You can't become an
48:38electrician quickly. Could you do it in
48:39a year? Yeah. could you become a nurse
48:41in a year? So, there's a time factor and
48:43I think what he was saying is he expects
48:45the labor market to remain on the weaker
48:48side. He expects the inflation data to
48:50lag. What will actually happen because
48:52of AI? And I think that's what people
48:54have to read the part on is what is his
48:56view on AI and productivity. And then
48:58finally, he wants these task forces to
49:00go question the way inflation is
49:03measured. And he's someone that has
49:04talked about median inflation. core
49:07inflation is going higher, but median
49:10inflation has not gone up. And the
49:11reason is there's a lot of stuff
49:13happening in AI like memory that is
49:15having kind of a one-time impact that
49:17will eventually come back down. And will
49:20the insurance companies, if they're
49:22starting to benefit, will the health
49:23care companies, if they're starting to
49:25benefit, specifically on the insurance
49:27side, be able to lower the cost of
49:30insurance, which we all know has gone up
49:31exponentially.
49:33I just think there's a lot of changes
49:35coming and he wants to be more
49:36forward-looking and not sit there and
49:38take every tick as being the way to do
49:41it everything. So, I'm going to say that
49:42the Fed is going to remain on hold
49:44despite everyone building all this stuff
49:46in. And I actually have liked reading
49:48what Wars has said. I think he's going
49:50to be much a much uh necessary person in
49:53an age of AI.
49:55>> One of the things I've been thinking a
49:56lot about is we talk a lot about
49:58individual companies etc. But the
50:00NASDAQ, let's just look at numbers over
50:02the last uh couple of time periods. So
50:04year to date, it's up about 16% as of
50:07today. Over the last 5 years, it's up
50:09somewhere in the ballpark of uh about
50:11100% or so. Um if you look at, you know,
50:14last month, it's down about 5%. And so
50:18you can go and kind of pick these
50:19different time frames. But if I said to
50:21someone, hey, you can deliver 15 20%
50:24return every year, year after year.
50:28Why do I go invest in a private venture
50:30capital fund who may or may not put up
50:32that money, you know, that those types
50:34of returns like I lock my capital up and
50:36be illquid? Why go pick individual names
50:38if I can deliver returns that are going
50:40to beat most hedge funds, you know, for
50:42example? How are you thinking about
50:44index exposures versus uh actually going
50:47and buying individual names or how do
50:49you think maybe people who are watching
50:50or listening to this should think about
50:51it?
50:52Well, again, one of one of the themes
50:54that I've tried to help with my
50:56subscribers on um has been the situation
50:59that all these including the NASDAQ,
51:01it's it's a modified cap weight. Um so,
51:05it's not 100% cap weight. They they do
51:07have limitations. Uh but I do think
51:10Nvidia is still a very large portion of
51:11it at this point. So, it's been pulled
51:13back by Nvidia, but it has a lot of
51:15names that have benefited from AI. So on
51:18the S&P 500 again as of today you're
51:21probably up about eight little bit
51:24percent uh maybe a little bit less eight
51:26and change is my guess which is a little
51:29surprising given the fact that earnings
51:31are up as much as they are but I think
51:32in the case of the S&P it is a cap
51:34weighted index and you suffer with cap
51:36waitings and the fact that you're
51:38heavily weighted towards the winners of
51:40the prior 15 years and you're
51:41underweight the losers. Well, if the
51:43semiconductors are winning and they were
51:45not the winners and the hyperscalers are
51:47losing, then you're suffering in that
51:48side. Um, I still believe stocks are
51:51going to outperform bonds uh for at
51:53least the next couple years until AGI
51:56starts to become a little bit more
51:58scary. Uh, I still believe that
52:00commodities are going to be a a a big uh
52:03necessary including silver. Uh, and I
52:06still believe that Bitcoin is going to
52:07be the best performer out of all of them
52:09once we get to the point that the
52:10negatives of AI start to become
52:13apparent. And I think they're becoming
52:14apparent right now. If the best you can
52:17get from the S&P 500 is just a guess or
52:20then or Q's is 15% a year, well, that's
52:25the world that Bitcoin thrives in. If
52:27you can get a 100 to a,000% which is the
52:30world we've been living in for retail
52:33which is the energy of Bitcoin well then
52:36Bitcoin's going to suffer. I think that
52:38world's going away and more importantly
52:40for people that are invested in private
52:42credit in private debt which wealthy
52:44people are in real estate I think
52:46gradually over time especially when the
52:48S&P starts to run into some some bigger
52:50resistance
52:52the volatility now is in the equity
52:55market. Um, I've shown this repeatedly.
52:58Remember when Bitcoin was such an
53:00extreme volatile position? Well, now
53:01when Bitcoin has a big down day, it's
53:03down one and a half percent. So, it's
53:05moving on a 30 V. Not even a 30 V using
53:08that. I'm telling you right now, the
53:10tech momentum index I the last 60 days
53:14is an 85 V. That's higher than any time
53:17over the last 25 years, including
53:19the.com bubble. Now, a momentum index
53:22like that is long short. So that has the
53:24winners of tech over the losers of tech.
53:27Well, that's the way most hedge funds
53:29run their book. When you have an 80
53:31volatility, that prevents you from
53:33having much of a position in that cuz
53:34you'd be hitting your draw down limits
53:36very quickly if you had that trade on.
53:38This is what's happening in the equity
53:40market. And this is not something that
53:41people should minimize. In fact, I've
53:43showed that over the last year, it's
53:45been a megaphone, meaning we're making
53:47higher and higher and higher volatility
53:50bands as we move forward. And that's
53:53going to make it more difficult for
53:54people to be long stocks. The sharp
53:56ratios are going down. And as we go
53:58forward, I'm going to say it again, the
54:00AI midcycle slowdown is basically the
54:01peak and the easy money of AI. And so if
54:04the money starts filtering back towards
54:06the other direction, I've laid out a
54:07pretty bullish case for Bitcoin. If you
54:09haven't got it, if we're overestimating
54:11the inflation side and productivity is
54:13coming, if we're overestimating the
54:15hawkishness of the Fed from what Worsh
54:17said, and at the same time, the AI trade
54:20is not going to lose its earnings
54:21growth. It's just going to make it more
54:23difficult for institutions to stay long
54:25it the way they have in the past. Then
54:27all of a sudden the the lower volatility
54:29of Bitcoin will probably lead in. If we
54:31get above the 200 day moving average
54:33with everything that I just said, which
54:34is up up north of 70 at this point, I do
54:38think that that's going to be the
54:39beginning of the next phase of crypto
54:41and the next phase of AI. We'll see if
54:42it happens in the second half of the
54:44year. I think it will.
54:46>> I love it. That's a great place for us
54:47to end. Uh anyone who has not yet,
54:50please go and subscribe to Jord's uh
54:52YouTube uh channel. He does a fantastic
54:54job there. And uh also go check out 22V.
54:57Uh he writes a bunch of research um that
54:59I uh I read every single time he
55:01publishes. So uh go check it out. And
55:03Jordy, I hope you enjoy a great Fourth
55:05of July in Maine. And uh we'll do this
55:08again next weekend.
55:09>> Same to you, bud. I'll see you next
55:10week. Same to you, Matt.

About this transcript

This page contains the full transcript of Everyone Gave Up On Bitcoin At Exactly The Wrong Time by Anthony Pompliano, generated from the public captions YouTube serves with the video. The transcript has 11,127 words across 1,559 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.