Loading summary
Transcript25 lines
- [00:00]
A
John, it's over. Bitcoin's dead.
- [00:02]
B
For the 285th time, we have to unfortunately attend Bitcoin's funeral. It's. It's as sad as it always is.
- [00:09]
A
The eulogies are getting better, though. They're getting better. It was fun. It's a nice college try. We kid. We kid. Obviously a lot of people out there wondering what's going on with Bitcoin. What is happening? We'll get to it, but I think we want to start on another side of the market. AI, I think, continuation of our conversation from last week. But first, before we get into any AI dividend funds for localities, we'll talk about the progression that Anthropic is really pointing at here in terms of what the models could do. And it's actually funny. I was at a lunch on Friday where we were talking about this and we naturally got on the conversation of humans thinking linearly, not being able to think in exponentials. And if what Anthropic said last week is true, which was in this When AI Builds Itself blog that they publish, we may be getting closer to some weird moment where maybe AGI, who knows, but basically Anthropic saying that, what is it, 80% of the code for new models is being written by Claude itself.
- [01:12]
B
Yep. Well, we say, you know, you said we're starting on the other side of the market and you know, we'll get to Bitcoin. But I think it's worth remembering as we talk about all these, that like, increasingly, I think, you know, not to go to Pepe Silvia on it, but I think it is all increasingly connected given how big AI is becoming on various dimensions. But yeah, I thought it was good to start here just because I think this maybe didn't get enough airplay in mainstream press, at least as far as I saw. This is the kind of thing that recursive self improvement, this has been the goal that a lot of us probably only really started paying meaningful attention to AI and progress in AI over the last few years as it's become mainstream and more involved in our lives. But this is the kind of thing that if you go back and look at what people in 5 square miles in San Francisco were talking about in the late 2000 and tens, early 2000 and twenties, this is the hurdle that you want to get to, if you're extremely, call it AGI pilled, and whether you're positive or negative on that future, if that's what you're looking for, you're looking for the moment where AI can be the best AI researcher. So you take humans out of the loop or minimize their involvement and AI analyzes itself to build the next one generational model analyzes itself to build the next generation of itself. Right. And obviously that can lead to some potentially frightening conclusions and some paperclip maximizer type futures that we could think about. But despite all that, I didn't really see this. Picked up on a lot of mainstream press headlines and to some extent they're talking their own book here. This is a marketing document from Anthropic. They're incentivized to tell you that they're just on the cusp of getting to this big milestone. But there's a good amount of data in there that points to the fact that this probably is happening and it seems, seems important. So this is going to the extent that this is not just marketing bs. It validates a lot of what we said last week about the pace of progress here, if you want to call it progress, potentially leading to some meaningful discontinuous changes in the economy in the next few years. And that'll have a lot of downstream implications that we have talked about and we will talk about.
- [03:24]
A
Yeah, and it's funny too. Not only did this, this blog post come out, and I agree, I don't think it was picked up by enough people. Was happy that it was recognized at lunch on Friday when I brought it up. Another you mentioned human in the Loop, taking them out for the model building themselves. But I'm not sure if you caught this, but you had Boris from Anthropic and Peter Steinberger from the Open Claw project basically coming out and saying the same thing, but on the prompting side, like they don't actually prompt agents anymore. They simply set up these loops of agents prompting each other and they basically set it off once. And the way their looping architecture works, the AI just builds everything for them. They don't even prompt anymore.
- [04:07]
B
Yeah, that's wild. It's only going to get weirder. And I think to be fair and not get too sensationalist, there's always a diffusion curve and the vast majority of people have never, largely never even heard of this stuff. There are enterprises everywhere that will be slow to adopt for reasons that we've talked about. They may get disrupted and they may ultimately be kind of dinosaurs that go extinct. But the call here is not that in 12 months every white collar worker will be without a job, but on the margin you could start to really see if this continues and accelerates even with the diffusion delays that you would expect. To see some meaningful impacts on at least segments of the economy, which maybe flows into the next slide nicely with long term longtime political allies Bernie Sanders and Donald Trump both arguing for the government or the public, as it were, to take a stake in leading AI labs so that the public at large can benefit from all of this crazy growth that we're seeing here.
- [05:08]
A
Yeah, I mean they were listening to us last week, must have been, and they said, hey, we need to do this. I've got some thoughts on this is Tom Mazziero from nelsphere3d post murder. I mean he's been beating the drum about this for about a month now. And behind the scenes, I mean I've worked with Tom going back to Great American mining and in the context of bitcoin mining, we had this idea almost a decade ago, which is when we were running into similar problems with bitcoin mining operations and the questions that neighbors would have about what we're doing and complaints about the noise. The idea of replicating the Alaskan Permanent Fund, which I mentioned last week, is. But bringing it to either bitcoin mining centers or now, I think AI data centers is something that, that is obviously becoming part of the conversation. But then you get into like, okay, if we're going to have the conversation, I think implementation details are important and both the Bernie Sanders and Trump's approach I would not agree with. I think we should get more local like these. If you're going to do something like this, a dividend, a compute, permanent fund, whatever you want to call it, it should be between the individual companies and then their individual operations within specific counties and specific states. And you get it. Hyperlocal. This is the idea of the US government buying a 50% stake or any material stake, and us expecting for the revenues derived from those data centers to actually be used and allocated appropriately by the federal government is simply like a nonstar. We know the government can't allocate capital efficiently. At least the federal government.
- [06:48]
B
Yeah, I mean, agree with all that. I think all of this is so preliminary still that it's tough to give a ton of weight to the exact implementation. But I just think I will say the horseshoe theory here is really interesting. Right. You're seeing all sides starting to seed the idea that something like this would be trackable and feasible. And I think the interesting implication of that too is if something happens at the federal level, it kind of goes back to and reminds me of what Sarah Fryer, OpenAI CFO said last year, briefly mentioning the idea of a federal backstop for the OpenAI build out. And if the federal government indeed did take some kind of stake in these companies that are becoming, it's clear, based on the Department of War's stance and the State Department stance, key national security assets, you start to really move into the direction of the big labs, the Frontier labs, and maybe some of the key infrastructure players effectively becoming too big to fail. What does that then mean for the state of public finances and actually making that happen in practice and providing that backstop with federal debt load where it already is, I think as it relates to how things connect to Bitcoin, that's something to keep in mind, something to think about for the next few years. If these become truly too big to fail, what's required to drive the acceleration that's necessary to keep them ahead of open source labs and open source models? And what would be potentially necessary if and when there's a meaningful hiccup or slowdown in the rollout or the adoption to the extent that ever happened, or if they could make the economics work on frontier models? Not saying that's my base case, but it introduces a lot of spiraling public obligations on both sides, whether you're bullish or bearish to keep this going.
- [08:38]
A
Yeah, to your last point there, I think another underappreciated theme of the last week, maybe two weeks, is the recognition that the open source models are maybe not exactly where the frontier models are at any given point in time, but they're close enough. And when you consider the fact that a lot of these hyperscalers running closed source frontier models are beginning to actually charge people for the compute, companies are having to really weigh the opportunity cost of using a frontier model versus a open source model that's 90% cheaper. So I think what we're seeing too is on the implementation of actually using these tools, companies are getting smarter about how much they use the expensive frontier models versus siphoning off subtasks and sub agents to open source models that are 90% cheaper. And so then you get to the question like, oh, is the US going to back these hyperscalers? And then the demand for their compute wanes because the open source models are so much cheaper.
- [09:42]
B
Yeah, I mean, I think it's a legitimate open question. If you get to the point of if we hit the recursive self improvement, fast takeoff moment, then perhaps the lead that Frontier Labs have over the open source models might expand and meaningfully increase. So maybe it's first person to that milestone kind of wins. I think you could have that argument. I think some people make that argument. I don't think either of us are smart enough to totally predict that. But it is absolutely a reasonable question. How does compute stratify long term? And is there a long term business model that can sustain the amount of capital that's already been raised by the Frontier Labs? I think it's still tbd. So we'll see.
- [10:23]
A
Yeah, we will see. And as it pertains to the closed source frontier models, Anthropic, OpenAI, Gemini, others in the space, Google, they're juxtaposed to the open source models which are predominantly being produced out of China. And a lot of what we discussed over the last six months is everything from this AI race to what's happening in the Strait of Hermes being a proxy war for the US versus China and this great power struggle in the 21st century. And you have some headlines here out of basically describing some moves that China's been making recently.
- [10:58]
B
Yeah, I think it was an interesting week last week. I don't necessarily have a grand narrative on top of all of these, but I do think it's connected to your point. The AI race, that's the open source versus closed source race and Chinese local competition on semiconductor technology trying to catch up to leading edge in the West. I think that's a key dimension of everything that we're seeing with the OSIS policy toward China. And I think we got several headlines last week, I think showing that there is a building amount of pressure in the Chinese economy. The EU is now moving to be much more restrictive on Chinese imports and crack down on various elements of a trade relationship. You've got questions over Are US companies going to find ways to basically build substitutes for Chinese rare earth minerals that have been thought of as these massive choke points on AI, on defense, on all kind of key industries, and growing questions around increased friction of the alternative payment rails that China and Russia have been building for bilateral trade. And I think all this just it's worth keeping in mind as we kind of move forward on this great power competition. The CEO of Payments Canada, I don't have it on here, but she also last week had a really good speech. I recommend people read I highlighted in the newsletter where she called payment rails basically weapons of economic statecraft. They're weapons of political influence. She Talked about how 80% of Canadian cross border flows route through U.S. correspondent banks and I think with everything that we're seeing, I think you should expect all of that to be more and more weaponized. So as a result I think things like this are worth keeping in mind as you consider the US's position vis a vis China, what it can leverage and what it can't. Because I think this colors the boundaries and the constraints that China has in responding to the us, and also the leverage points that the US is going to push on, whether it's AI or tariffs or something else.
- [12:57]
A
Sounds like more people are waking up to the thesis that bitcoiners have been putting out there for quite some time. They can weaponize the payments Rails. Who would have thought, you'd think we
- [13:05]
B
would have been aware of that in 2022 more broadly after everything happened with Russian assets and the Ukrainian invasion. But it looks like we're going to get further examples of that here over the next year.
- [13:17]
A
So be prepared. Make sure you have bitcoin in cold storage, money that is permissionless, that you can always access. It's going to be important as we move into this crazy world where you have superpowers competing on that competition. Another topic is okay, China's making their moves and we've been talking a lot about the re industrialization of the United States and you're highlighting the ISM manufacturing PMI which is over 50. So it seems like the investment and the spend in the industrial capacity of the US is certainly rising, getting more expensive and this tends to be a leading indicator because it's tied to PPI to a certain extent too for inflation down the road.
- [13:57]
B
Yeah, I think there's a PPI element but there's also just like you can see under that red line there, we've been bobbing along basically just under expansion for since pretty much the rate hiking cycle began at the end of 2021, we moved down aggressively and by the summer of 2022 we're basically in contraction mode for three full years essentially on the manufacturing side and now kind of definitively bobbing our heads above that for the last five months with a recent month on month acceleration in May. So this aligns I think with some other data points we've pointed out over the past couple months, whether it's freight volumes or other industrial data points. And I think it's not saying that the strategy is necessarily working, that it necessarily will work, but certainly the data that you would be looking at to respond to the administration's current policies thus far directionally are. I think it's interesting because last week we got the GDP print and you actually had Q1 GDP revised down primarily because of investment spending was lower than originally had been forecasted or had been included in the original data set, which is odd. If you were seeing an industrial renaissance, you wouldn't want to see business investment spending being revised down in gdp. So that was kind of a counter indicator. But thus far, I think over the last first half of this year, I think you are seeing green shoots on the domestic industrial side, which points to all of the activity we just talked about vis a vis China and I think where the administration is trying to steer the ship right now.
- [15:29]
A
Yeah, PMI is one of those indexes that people look at in relation to Bitcoin and say if it's over 50, bitcoins typically going to do well as a lagging indicator to that. And so on that note, you have other indicators that people like to look at when it comes to bitcoin. And we had decode highlighting that each time the copper to gold ratio reclaims its prior low, Bitcoin rallies bottom to top for an average of setting projected peak at the end of 2027. So are you calling for a reversal in the bitcoin price here?
- [15:59]
B
Definitively? I don't call bottoms, I don't call tops publicly. Maybe in the 1031 team chat, I think we've all been known to call bottoms and tops and super cycles and super cycles and everything like that. So I'm definitely not calling for any of that. I thought it was an interesting data point here. Copper to gold is kind of this loosely held relationship that is a very blunt instrument to think about where the business cycle is, whether it's inflecting upward or downward or not. And so the data set here suggests, as you said, there might be some green shoots in play. If you think that this is. It's an N of what, like three? So it's not a super powerful data set, but I think it's directionally interesting to keep in mind and to watch, especially relative to some of the other data we're about to go through. On the flip side, Michael Howell, the kind of liquidity, global liquidity godfather, has a view that we might be heading to more bearish territory for liquidity for financial market assets like Bitcoin. Because, yeah, you may have this major industrialization spend, but it's going to be, in his view, drawing liquidity out of financial assets and into major kind of industrial buildout. I would think the kind of pivot point is how accommodative is our fiscal and monetary policy to kind of, can they be accommodative enough to allow all assets to essentially benefit from this upward impulse, or is it really going to Be kind of net contractionary out of one pool of liquidity into industrial activity. So I think that's a tbd. But with some of the data that we have on the next few slides, I think it's at least worth keeping in mind and watching.
- [17:36]
A
Yeah, and the next slide we have here is the amount of bitcoin in profit versus the amount of bitcoin in loss. And another historical indicator again and of a few data points considering the relative nascency of bitcoin. But typically when the total supply of bitcoin in loss begins to cross with the total supply of bitcoin and profit, that is typically an area that marks a bear market bottom or consolidation level. And having been in this for 13 years, I will say just anecdotally pattern recognition coming in. It does feel like we are at that sort of chop solidation, low setting territory, could last months, could last for the rest of this year, but it feels like that's where we are.
- [18:22]
B
Yeah, I mean I think again not trying to call a bottom or timing or anything like that, but I think data like this would directionally suggest and same as the next slide if you want to pull that one up too from James Van Stratton. This would all tell you you're probably closer to the bottom, a lot closer to the bottom than the top. Doesn't mean you can't go lower, doesn't mean you can't have a violent wick down. A final just absolute capitulation puke. Maybe that's what we've seen over the last couple of weeks. But having invested in different markets markets in my career, the very common human thing is when something is just constantly going against you and when it's painful to hold, you get to a point where you're just like, just get me out of this. I don't want to look at this. I'm tired of it. I'm angry at myself for buying way higher. Just gimme out and I'll hold cash. I won't do anything. Maybe I'll rotate to stocks, maybe I'll just hold cash, but I got to get out of this thing. And typically when you feel that it's the wrong time to sell, you missed it. It not financial advice, but don't try to ameliorate or correct for the fact that you should have sold whatever if you were going to sell bitcoin, yeah, you should have sold at 125, you should have sold it at 100, you should have sold at 80. Well now it's at 60. So don't sell to recapture the opportunity to sell at 100 or whatever because you missed it. Now, if you think we're going to zero, if you think we're going to 20k and you can buy back cheaper, then make your call on that. But I think if you look at these last couple of charts, if you look at sentiment and everything we've been talking about here, I think probably it is fair to say we are closer to the bottom than the top. Not a bottom call, but do with that what you will.
- [19:56]
A
Yeah. And then it's important to pair these technical levels and these indicators with some qualitative fundamental news. And I think we've gotten that out of the institution. Charles Schwab launches 247 Bitcoin futures trading on thinkorswim and we have better and Coinbase issuing the first crypto backed conventional mortgage and Fannie Mae really leaning into this. I think the last couple slides here you're trying to make the point that there's no reason that Fannie Mae or Scott Besant needs to be championing bitcoin and broader crypto. I would hope it wouldn't be that way, but it is. There's no reason for them to do that if they really didn't believe that there's a there there and it's something that's strategic for the U.S. yeah, totally.
- [20:44]
B
I mean I think this stuff has caught my eye because I just frankly find it odd relative to having been seen a couple cycles now. We remember very clearly when 2022 happened the rapidity with which corporate partnerships or government interest, like if some of our portfolio companies or people that we knew in the space had been building rapport with different government officials or executives at big public companies or whatever, the switch just turns off once you're into deep bear market mode. Emails don't get returned, things get put on the back burner. It costs them nothing. It costs an incumbent. Whether it's an incumbent company or a senator or a governor or whatever, it costs them nothing to just put this off for another year. They have 30 other things that are more important to them and more core for them. And if bitcoin's not ripping, if it's not making new all time highs and everyone's palpably excited about it, it the natural easy, low friction move is just like, let's put it on hold guys, we'll revisit it in 6 months or 12 months or whatever. And I think with that as a background, it's very interesting that as price action has just been Awful unmitigated bear market type action for the last six, six months basically since October, you've just seen all these initiatives continue to march forward from institutions and people that don't have to be dealing with this. Like Schwab is a massive $13 trillion asset management platform, very old, very white shoe, caters to very mainstream audience, has absolutely no obligation to continue moving forward with any Bitcoin products. It could easily just table it until next year. Coinbase is obviously very bitcoin levered, but better. Their partner in the mortgage product is not. They're not a giant of the mortgage industry, but they have a totally different business that has nothing necessarily to do with bitcoin or crypto. Certainly Fannie Mae could easily, the FHFA could easily table this indefinitely. It could just go into a black hole and never be thought of again. Who cares what Bill Pulte said? They're not under any obligation to do anything with this. And then the last one here that you see on screen is Scott Bessen, the Treasury Secretary. He was teed up with this question, to be clear by Senator Tim Scott, who is part of the digital assets push. But with bitcoin going where it has gone, there's no meaningful political credibility that Tim Scott gets from bringing this up. And Besant could have easily been much more evasive than he was. This is really interesting language if you're trying to just move on from this. And that whole Bitcoin thing was this embarrassing thing that we had to do to win a voting bloc to win the election. But we plan to never think about this again. You could easily just completely avoid this, throw your hands up and forget about it. And that's not what any of these people or institutions are doing, which is meaningfully different from prior cycles at this time. It's just like I just have to ask why, what is the reason that you'd be doing this if you did not in this environment, if you did not see some kind of ongoing durable benefit trend tailwind that you want to latch onto?
- [23:39]
A
Yeah. And for those of you who are listening to this, not watching it, we have a quote grab from Treasury Secretary Bessant's response to Tim Scott's question, which is, I couldn't agree with you more that economic security is national security and part of that is our digital assets initiative. The strategic Bitcoin reserve is new technology, new ground. We're making sure we use best practices and things will be durable for the future. So they're signaling that they're moving forward with the strategic Bitcoin reserve, which is very interesting.
- [24:13]
B
Yeah, we'll see. Right. There is an executive order on the books. There have been multiple pieces of legislation suggested to codify it. I don't necessarily know know exactly what form it'll take or how that will happen, but this is something, again, that we could have abandoned long ago. This could have been a meme from late 24, early 25 that no one is under any obligation to talk about or think about again. And so the fact that even in the middle of this awful drawdown that it's still being apparently taken seriously is interesting. At the very least, if I could leave people with one thing, it would be, I would say the show Mindprint Hash with Matt Dines and Cameron Otsuga at Build Asset Management is, is very underwatched, very underfollowed. I think Matt is a great Twitter follow, has a lot of great Pepe Silvia type takes on the government's interaction over time with bitcoin. Not saying I fully endorse or agree with all of them. Not saying he's right on everything. But those guys are taking a very heterodox view on how things are developing and they're, I think, identifying a lot of the same headlines that we're identifying and drilling deeply into them. So my alpha for the week is go subscribe to that show and go follow Matt on Twitter.
- [25:20]
A
Twitter levered USD usts usts on X. We'll link to it in the show notes. We'll be back next week.