Loading summary
Transcript39 lines
- [00:00]
A
John, this is your, your first experience of me recording from the back porch of my father in law's shorehouse.
- [00:06]
B
First time caller, long time listener for something like this. It's always a key part of the summer for me is when Marty transitions to his the shore house and does all the pots on the back porch. That's, it's a real vibe I look forward to.
- [00:18]
A
It's a great vibe. We got birds chirping in the background. There's not a cloud in the sky. It's about 74 degrees. Very comfortable, not sweating. And after the, the heat and the smoke that hit the area last week, this is a much, much needed reprieve from, from the weird weather.
- [00:35]
B
Yeah, you, I'm glad, I'm glad you, you made it through that rough, rough period there.
- [00:40]
A
Did the smoke make it down your way at all?
- [00:42]
B
No, no, we're, we're insulated from all of Canada's shenanigans down in the lower half of the country.
- [00:50]
A
Hey, it turns out forest management control burns kind of important. You neglect them for a decade, bad things happen and then it gets, then it gets blamed on climate change and you're forced to move around your whole life and your portfolio to make sure that we're investing in clean climate tech.
- [01:06]
B
I think the narrative was a little less powerful this time for one reason or another. Fortunately, we're past the ESG days of early 2020s, but never say never. Investment cycles always come back around.
- [01:17]
A
Well, I mean they may come back around, but as you said, it's out of favor now because war is not good for the environment. War is back on the table. Iran shredder hurricane moves back on the table. Boys.
- [01:28]
B
Yeah, we, here we are again. All this has happened before. All this will happen again. As they say. You know, we don't have to hit these too hard. I'm sure anyone who is is a professional doom scroller or a professional situation monitor is, is well aware that where we're sitting right now is, is definitively less encouraging than it was perhaps a couple months ago. Perhaps not totally surprising that we would see re escalation. Maybe you might have thought that it would have waited till after midterms, but I think anyone probably would have assumed that the MoU that was put across the table a couple months ago was not super well defined on a variety of terms which we talked about a couple months back. But here we are again. Straits open, Straits closed, Michael Scott, Snip snap, Gif, Schrodinger's cat type situation. But either way, the Punchline here is WTI Brent up back into the 80s but I think even more notably you've got European nat gas benchmarks all getting back into highs of the early chaos that we were seeing earlier year. So that'll be I think the, the biggest constraint on you know, or the biggest pain point I guess we would say is the countries Europe East Southeast Asia that are maybe less well supplied on variety of key energy inputs really getting squeezed here again so that'll be the, the main vulnerability to, to watch
- [02:48]
A
and stateside we have the Strategic Petroleum reserve falling to 43 days of reserves left lowest I think 83 lowest since 83. Lowest since 83 yeah so got down to 46 days in 2023 so I guess we replenished it a little bit and we've drained it down to its lowest level in over 40 years. So not great Bob.
- [03:09]
B
No, well it might have been a good idea to deal with that before kicking the hornet's nest here but yeah, here we are. But I think you know everyone, everyone knows there's the war is on if we can call it that. Now I think maybe the more incremental or like interesting pieces out of the last week were some headlines just on basically the accelerated urgency to route around straightforward news by various GCC countries. We've talked about these a little bit in the past on the pod, but some interesting analysis out of Goldman this week kind of suggesting that a variety of these projects could get up to say 60% of pre war levels of oil exports in the next two years being essentially straight proof and definitely looks like there's a significant appetite as you would expect to ramp up both investment in these and then also the pace at which they get completed. Now this isn't going to help you if you end up with the Houthis over in the Red Sea causing a lot more trouble and they're Iran aligned. So you're not necessarily out of the woods just because you built some of these pipeline projects and of course they're not going to materialize tomorrow. It's going to take one or two years to really get meaningful benefit. But the world really moving kind of in a direction it seems of minimizing dependence on an ability to this particular maritime choke point which interestingly was specifically called out as we mentioned before in the National Security Strategy document last year as a key priority of the administration minimizing the disruptive impacts of the Strait of Hormuz as a choke point. So a tinfoil hat theorist might say that this was maybe an outcome that certain parties were looking for, but we'll see how it actually shapes up.
- [04:49]
A
Yeah, and I'm not sure if you saw the headline this morning out of Yemen, but it looks like the Yemenis government, or the Yemeni government, excuse me, has announced that they're going to put a naval blockade on Saudi Arabia's ports, which should.
- [05:02]
B
I actually, I actually missed that one. So you're, you're breaking news for everyone here.
- [05:06]
A
Yes. So if you're, if you're thinking about what's going to happen when these pipelines do make it to the Red Sea, as you mentioned, the Houthis, I mean, Yemen, I think loosely aligned there with some Houthi outshoots in the country, looks like they're already beginning to signal like, hey, you're not going to get away with this. We have something to say too. So chaos in the Middle east oil markets. And as this chaos is going, it looks like the US is trying to carry favor with some of the countries in the area and the latest of which is Iraq, which the US and Iraq have a pretty muddy history in the 21st century. But it looks like the Iraqi Prime Minister will be visiting Washington to talk about oil and gas deals.
- [05:48]
B
Yeah, I mean you talk about cyclicality, right? You think about the 90s and H.W. bush shaking hands with Saddam Hussein, 15, 20 years later he gets summarily deposed and we do some nation building and it's a chaotic scene all around. And then 20 years later, here we are again, striking deals and building out partnerships that'll benefit, allegedly US companies like Chevron. So I think these are self explanatory, but I think it's as it relates to a US that's looking to flex its might more and more in energy markets internationally and become more and more of a dominant player across petroleum markets. Clearly one to watch here. Humorous enough that it's all coming back around to being a partner with Iraq, but just the latest kind of puzzle piece. Plugging me in here to the US's clear desire to be more and more influential and puts hand in the till more and more in global oil markets.
- [06:43]
A
Yes. And I mean we've been covering this extensively, which is a lot of the focus on the straighter Hormuz may be focus that's not in the right area. It's what's happening on the edges as the straight of Hormuz shuts down and how is everything being redirected and re architected on the back end and helium, which as we know is, which as we now know, after becoming Experts on Middle east oil and gas strategy is a key input in fertilizer and other sort of other semis too. Other semis. Semis. It's good for helium is very important. That's what I'm trying to say here. Back deck fog is hitting, but it looks like the US Is going to be a massive benefactor in the the helium markets because of all this. Look, if you're looking at the chart that we have on the, the screen right now, it's looking, it says US Emerges as helium winner. We're looking at the import of noble gases from top three sources in 2026. Percentage of if you're looking at 2022 and to Japan was hovering around 25, 30% it'll be around 80% this year. South Korea, it was below 20% in 2022 and is approaching 60% this year. And then Taiwan similarly below 20% and looking like it's going to be just at 60% of their helium imports will be from the US so big winner in the helium markets the US is.
- [08:12]
B
Yeah. And if, you know, if you look at kind of the trend here, if you're watching on video like already kind of in play and in progress over the last few years. So it's not just the disruption that we've seen in the Gulf, but clearly there's been a jump this year, a meaningful jump, especially for Korea and Taiwan across for helium import, the US As a source of helium imports for these countries. And it's the red bar here of China getting squeezed as a player. That was also. That trend was playing out even before this. But I thought it was a good chart to reiterate. It came out of, I believe the UK this week. Something we've been talking about again over the last few months, just the wins on the margin that the US Is seeing, a lot of key inputs where it's becoming more and more of a relevant player and exporter. And I think again, just cuts against the idea that there's no edge, there's no strategy, nothing behind what the US Is doing other than Trump flailing blindly in the Gulf. Not to say that everything has been executed perfectly or without cost. We don't have the chart on here, but there was a really good a paper out of. I forget the organization. It's one of the big basically energy monitoring and commodity monitoring organizations kind of their annual paper on the state of play on critical minerals. So the flip side of this chart is, unsurprisingly China remains well ahead of anybody on kind of refined capacity for rare Earths. I think that explains, unsurprisingly, why you've seen all the headlines we flagged over the last few weeks on the US Doing a lot more on the industrial policy side to build out dual sourcing capabilities and do partnerships with nations like Brazil to further bolster US Rare Earth capacity. So it's pros and cons on both sides, again, like we've been saying. But there's, there's definitely clear benefit that you're seeing to the US from, from this disruption, kind of one way or another.
- [10:11]
A
Yes. And then we transition to how the Fed is trying to operate and navigate in, in this world. Not only this world, but in a world of their own in which they're trying to drastically overhaul, get away from forward guidance, add more data resources that can react in real time. And we've got some statements from the Fed this week pertaining to inflation. And I think they're really, again, trying to position away from forward guidance and saying they're going to be more reactive with lifetime data, saying we can't get stuck in the cognitive dissonance of fighting the last war on inflation. And a couple of Fed governors coming out and say they believe inflation has peaked and saying it pretty confidently. So we've got the Fed beginning to position its new regime here.
- [10:59]
B
Yeah. Which is never exactly what you want to see the Fed trying to call an inflection one way or another, especially on inflation. It brings back the PTSD memories of inflation as transitory from anyone who lived through 20 and 21. But I just thought it was an interesting set of comments that came out last week. The top one is from new Fed Chair Kevin Warsh in front of Congress, testifying that if we do things right, the inflation surge of the last five years is a thing of the past. And he talked a little about AI as well in those remarks and in the past, if you've looked, he's been very on the train of AI productivity enabling kind of a deflationary impulse that can counteract some of the more inflationary forces that we've seen over the last few years, and not wanting to get in the way of that and hamstring that and really allow that to flourish, to allow for that kind of deflationary impact. And then Fed Governor Waller telling people that we shouldn't be, as he says, fighting the last war, just kind of looking backward and doing what we should have done five years ago in a totally different set of circumstances. And then John Williams saying, like you said, putting in a top, rightly or wrongly, on Inflation collectively. I just think that's an interesting talk track from all these guys that allegedly we're moving away from forward guidance. So I don't know if this is intended to be a forward guidance light or if maybe the WARSH doesn't totally have a handle on how much the rest of his team goes out and talks to the media. But in any case, the trend line here is not suggestive of Fed governor's trying to tell you that they're going to have to get more hawkish and that their key focus is on getting CPI down. Which is interesting given that all of these comments were happening in the midst of WTI ripping back over 80 and all the energy input spikes that we talked about as uncertainty flared up to a multi month high in the Gulf. Again, against that backdrop you've got all these guys telling you, well inflation's probably like it's important, but let's not fight the last war. I think it's peaked. I think we can do enough to make it a thing of the past. Just an interesting dissonance between those two that is I think indicative of how the institution is shading toward positioning itself. There's a lot of internal division and strife apparently among all the different Fed governors and people with a vote here. But I think this is putting into the zeitgeist an expectation for incremental dovishness versus incremental hawkishness.
- [13:19]
A
Yeah, it seems. I mean, and then you factor in the inflation prints of last week too, which surprise to the downside and you could see if the war is blowing up, obviously interest expenses going up. I think this is going to be the biggest annual military spend that we've ever seen. At least that's what Trump and the administration are asking for. So the fiscal side is continuing to completely blow out and I find it hard that they're going to be able to hike rates into that.
- [13:48]
B
Yeah, I mean we've talked about it a bunch here, but the flexibility isn't tremendous here, particularly if you want to enable everything you just talked about and enable warsh's key priority of not stepping on the AI buildout. And perhaps that leads us nicely into this next topic.
- [14:05]
A
Yes, which is Kimmy. Kimmy's so hot right now. Kimmy, come back. Any other fool could see.
- [14:12]
B
Yeah, I mean this is the box for Marty singing. It's an important component of every episode.
- [14:16]
A
This is a shot across the bow of the Frontier labs in the U.S. kimi K3 came out surpassing many of the U.S. frontier models on some benchmarks and I think the at least reading the tea leaves of people who are using Kimik 3 and comparing it to Fable 5 and Chachi BT 5.6, Sol and Terra, it is a legitimate contender for top frontier model out there. Obviously it is Chinese open. It is a Chinese open weight model. And I think one of the interesting things here is I think we might have discussed it last week, but if we didn't, Dylan Patel from Semianalysis wrote a piece earlier this year that said the Frontier labs have a massive lead and we'll be able to defend that lead because of the relative lack of access to GPUs that the Chinese market has. He was making the case that they won't be able to train the models in the way that Anthropic and OpenAI can do so here in the States because they have access to all Nvidia's chips in China due to export controls, tariffs and other reasons, doesn't have access to those ships and therefore will not be able to to train models comparable with US Frontier models but can be launched. It is comparable, it is competitive. And I guess that begs the question are they getting access to the GPUs, are they finding ways to train these models way more efficiently with with less tokens? And what does this mean for the the future of AI, particularly as it pertains to this juxtaposition of open weight models versus closed source frontier models here in the United States, which are cozying up with the government to try to prevent the layman from getting access to the full power of these AI tools. And I don't know if you have them on the list, but I'm sure we'll talk about them. But there were, there were two headlines that came out over the last five days that really highlight this one pertaining to a dear friend of ours, Kale, who's a developer of Cashew Getting getting a sh.
- [16:20]
B
David Sacks, last night again, you're breaking news here. I missed that one. We should definitely talk about that because that's big stuff if you've been following Bitcoin for a little while. But yeah, I think, I mean there's a ton to say here. We're not going to have time to drill into every possible thing, but I mean there's a lot of nuance within a chart like this and within this whole topic. This immediately kicked off all of the debates about well, are they benchmark maxing and it's not actually as you know, it's not actually as comparable in real world use. We've talked about kind of that dynamic with open source versus true kind of frontier models in the past. I think I haven't played with Chemik3 enough to have an opinion on that. Definitely in the past I've been less impressed for at least agentic deployments by the open source models relative to everything else on the frontier. And I think you've generally felt similarly. But maybe we'll come back in a week or two on analysis there. But there's also the question of. Well, like you kind of alluded to, are there shadow GPU clusters in Singapore? There are all sorts of shell games that can be played to mask who the actual end user is. So has China actually been getting more access to true frontier accelerators? Tpd? Are they using distillation that both the labs and the US government could eventually crack down on more effectively? That's potential too. I think there are good arguments to be had by guys like Gavin Baker. We'll have a tweet from him here in a second. But. But some good analysis that he put out just on the relative cost of Kimmy K3 versus say Fable or Sol on a true per task basis because K3 appears to be much more token hungry. And so if you need a lot more tries effectively per task completion, do you actually get real savings in commercial deployments or not? I think that's very much tbd. All this is to say I don't think that the massive kind of doomer headlines or the desire that I see from a lot of people, weirdly, especially bitcoiners, I think there's this desire to. This desperation to call a top on the AI bubble and to say this is it. It's all going to be one shot by open source models. I think maybe there's an underlying desire to tell people sell all your semis and buy bitcoin. So maybe that's a piece of the motivation. But everything I'm getting at here is that I think all simplistic readings on what this headline means 48 hours in or maybe quite overdone and need a little reconsideration. But you know, I think there's. That's. That's all like one, one piece of it. I have a sense that you want to jump in here on hugging face. So I'm going to pause and pause my rant and let you jump in.
- [19:00]
A
Well yeah, calling tops or like saying that this is the. The death of us closed source frontier models, I think that's way overblown. I think the more important signal here is the kneecapping of These models and the sort of regulatory encroachment that the Trump administration is taking. And I think that's what David Sacks was pointing out in his tweet last night. But just to sort of preface this the right way to your point, like who's. No. If they're benchmark maxing, if they have these shadow GPUs, if they're waging distillation training attacks, I don't know if they're attacks, however you would describe them. But I think that that misses the point. And I think the two headlines were the hugging face story and then Calais. He basically wrote a tweet, I think Friday or Saturday, basically saying, I found 17 or 12 critical, critical security vulnerabilities in a couple of the projects I'm maintaining with Kamek 3. I tried to do them with Fable 5 and with GPT 5.6 SOL, but due to the sort of curbing of those models, particularly when it comes to working on code bases. So for those who are unaware, Anthropic famously will let you use Fable 5, but if you're doing anything that is security sensitive in a code base, they will divert you to Opus 4.8, to a lesser model, to a model that's not as performant. And I think this would be a signal that Kimik 3 is comparable, at least to an extent. But two cases over the weekend, hugging face they had. I mean, this is just a crazy story, crazy story generally, and Kimmy K3 is just like a small detail of it. But they had the, what they're describing as the first sort of autonomous AI attack on their code base. They got a code injection in one of their databases and they had an autonomous AI attacker execute more than 17,000 events over the course of a weekend. And they found the attack and they started protecting against it. But again, they were trying to use Fable 5 and ChatGPT 5.6 sold to basically wage a defensive attack, a security audit of their systems of what was being infiltrated. And Fable 5 and GPT 5.6 said, you can't do this. You're not allowed to work on this type of task with, with these models. And they were forced to download Kimik 3 in a local environment and run it, run the security audit with it against their systems. And it was able to identify with Kimik3 what was wrong. Similarly with Kali, I don't know. He didn't say which project it was specifically. It could be Cashew, it could be Bitchat, could be clawedi. He's running many things. But similarly, was trying to do a security audit of his systems and was using Fable 5 GPT 5.6. They said, hey, you're not allowed to do this task with these models. And he was forced to use Communicate 3 and was able to find 12 security flaws in the code base that he was auditing at that time. So this isn't really a debate about whether or not which frontier model is better than the other. It's what is the nature of being able to use these models, how are we going to be able to use these models? And I think the open weight Chinese model is really proving that what we're doing here in the United States is not going to work long term. You can't cuck the models, you can't prevent people from using them or you're going to get out competed. And whether or not Kimike 3 is actually as performant or more performant than Fable 5 or ChatGPT 5.6 Soul right now I think is irrelevant. It's like if like people are going to want to use these tools in particular ways and if the frontier models here in the US don't let you do that because of fears of the models breaking containment or doing things that the government doesn't like, it doesn't matter if you have an open source competitor that's going to let you do it, you're going to fall back to that even if it's not as performant.
- [22:50]
B
Yeah, I mean I think this is a great maybe segue into that. The Gavin Baker tweet and the other thing on that page, which is basically just getting to your old friend Jevons Paradox, everyone's paradox was the thing that I think bitcoiners were talking about well before it became kind of a mainstream talking point among everyone in the AI world. But in any case, we're not going to read everything on here. I would highly recommend going to read Gavin's tweet and his pinned tweet and everything else that's been on his timeline over the past couple of days. On the right is just an example of this playing out where basically you've got the CEO of databricks outlining essentially this dynamic of as on a per unit basis, this resource effectively gets cheaper. There is that much more or even more compensating amount of demand for volume for that resource, same as it ever was. Same dynamic that we've seen play out across commodity markets over the history of capitalism. And so I think if you split out, who is this good for, who is this bad for in the broader AI complex. And I think this relates ultimately to not just companies, but also the US and China. There's the inference piece and the training piece, and I think this slide is all about how the inference piece is going to be just fine and likely continue to explode. I think open source models, especially a diverse ecosystem of open source models, is gigabullish for most or definitely pieces of compute, broadly different pieces of semiconductors and the power suppliers that will ultimately be required to stand up enough capacity to take advantage of all that. I think it's one of the reasons you see Nvidia promoting their own open source models so aggressively, to kind of avoid the monopsony or the oligopsony of just having a few close frontier players in the space dominating the whole. The whole of everything. But that's the inference side. I think maybe the more interesting strategically derivative question is more on the training side, which you see people kind of wringing their hands once again over this last weekend on the degree to which training will be financeable at all. These massive and increasingly expensive training runs of tens of billions or hundreds of billions of dollars over time to get to the next frontier generation. How can you really underwrite that if the ROI is going to be meaningfully undercut by whether it's a distillation attack or a fast follower or however you want to think about it. But if open source can come in and constantly just ride on the coattails of the gains that are made in training of these new models, then how are you even going to get further model gains over time? Because there won't be an incentive to continue spending on that. And I think this is like to Your point on OpenAI and anthropic in different ways, cozying up to the government. I think it just kind of gets to a point of view. I don't necessarily even think that that's right and that's how it's going to work out. Especially if you think that the Frontier Lab's ultimate strategy is to verticalize everything and to stand up kind of their own accounting and law firms and all these different professional services they could go attack as vertical infrastructure on what they've built. I think you can very easily justify meaningful ROI on that basis. But let's even just game it out and say that that's actually what happens and that you have a meaningful headwind to to training spend for that economic reason. I think you just have to ask yourself, well, as the tweet on the left here suggests, K3 shows that scaling laws are continuing right Effectively larger models are still giving you concomitantly better performance and throwing more compute at training, throwing more resources at it. Growing the size gives you the ability to get better models. I think if scaling laws generally are still holding and there's still meaningfully more juice to be squeezed out of massive training runs, there's still technically progress that can be made. It kind of doesn't matter if the private ROIC is there, because neither China nor the US nor anyone else who's trying to play in this game is going to be able to just concede and let the other side get to it first and own this incredibly important technology. And they're not going to, as you see on the right here, these projections of spend from AI and anthropic over, especially OpenAI over the next five years. We're talking about hundreds of billions of dollars in incremental spend. And this is very likely not even counting all the derivative spend that needs to come in from the power and infrastructure side, which is very clearly critical to the way that the Trump administration is thinking about reindustrialization. It's critical to the way that Kevin Warsh is thinking about managing monetary policy and inflation is having to spend here, building this out and then getting return on that. I think what I'm basically getting to is even if you think that that's what happens, the training just gets kicked to the sovereign level.
- [27:40]
A
Right.
- [27:42]
B
In no way do I think that the US is ever going to look at this and say, well, we could continue to dramatically expand model capabilities if we just kept spending on big training runs, but it's hard to finance because open source makes it tougher to capture those returns. And so we're just going to concede on that. I think what you're looking at here on the right side is even if you're super bearish on open source's impact on Frontier Labs, what you're looking at is just another line item on the federal budget.
- [28:09]
A
Right.
- [28:09]
B
Like someone's going to be spending that money. Right. Whether it's OpenAI, Anthropic and their investors, or private financiers or the government, I think that curve is going to move that way regardless, as long as scaling laws hold up.
- [28:23]
A
Yeah. And it'd be remiss of us. Not to mention. But I think a whole new sort of twist to this juxtaposition of the superpowers in the US and China and this age a AI race entered the conversation last Thursday night when President Trump gave the speech on Chinese meddling in the 2020 election, which is probably a, A strategic.
- [28:44]
B
It's interesting timing, right?
- [28:46]
A
A strategic announcement based off of everything going on right now in the. The AI world.
- [28:52]
B
Yeah, it's, it's an interesting time for sure to, to, to bring that up out of nowhere.
- [28:57]
A
But speaking of interesting times, interesting times in bitcoin, hovering around 64,000 up from the lows of the prior week right now. And our good friend Alex Thorne and his team at Galaxy Research are just putting together a chart to see what indicators that have historically triggered a bottom in bitcoin have been hit already. And so we've had four that have hit two that are approaching and about seven that are not yet there.
- [29:26]
B
Yeah, if this is wrong, you can put it on Alex and Galaxy and don't blame us. But just an interesting kind of data point or set of data points, mile marker on where we are in bitcoin's cycle, such as it may be. And I think certainly looks like if you kind of just eyeball this closer, much closer to the bottom than the top, although not fully there yet. And I just think interesting to. As the market has completely moved on from bitcoin, not in the meta at all, not in the zeitgeist, bouncing around this very kind of tight consolidation range in the low 60s. If you compare that to everything else we've just talked about with war, defense budget, the Fed's current positioning on inflation, the potential need, if you're super bearish on tier labs, for the government to effectively pick up that tab and backstop them, as has been hinted. If that's your belief, then it's a really interesting time, I think, to consider all those topics against the backdrop of what bitcoin's currently doing. And this last headline that we have from BPI on the strategic Bitcoin reserve Armo bill finally moving to committee for the first time. Interesting set of data points here as we look at bitcoin's breadcrumbs against everything else.
- [30:41]
A
Yes. And actually was in D.C. last week recorded with Connor Brown. So if you want the inside look from the Bitcoin Policy Institute's perspective on what's happening with bitcoin on Capitol Hill, that'll be on the TFTC feed. This feed, if you're listening to this on the podcast on Wednesday. So go check that out and we'll be back next Monday.