Loading summary
Transcript109 lines
- [00:00]
Michael
Just real quick on the basics. I think this is going to be a meme that's going to come together. You're starting to hear folks talk about it just like deep in the bear, how there's nobody coming to save us. There are multiple folks. I think Lyn Alden had a tweet about, you know, you just got to realize bitcoin's fundamentals set the floor and the rest will take care of itself. I think Parker was on a pod recently talking about it, and then even our buddy Eric Balchunas was saying it as well. Like, I think the reality is risk adjusted. There's never been a better time. The fundamentals are so strong.
- [00:31]
Liam
It all comes down to computers communicating.
- [00:34]
Michael
The information superhighway can be a confusing mix of on ramps and off ramps. Bitcoin is worthless artificial gold.
- [00:44]
Liam
Is it still rat poison?
- [00:46]
Michael
Probably rat poison squared.
- [00:48]
Brian
We need to get into the world of okay.
- [00:50]
Liam
This is actually foundational technology.
- [00:53]
Michael
What the Internet of money does is it creates a single network which can do a micro transaction to a giga transaction. The Internet is going to be one of the major forces for reducing the roll of gun. The one thing that's missing that will soon be developed is a reliable e cash.
- [01:20]
Brian
All righty, gentlemen. Welcome back to another episode of Final Settlement. Today is Monday, July 20, 11:39am Eastern Standard Time. We're back. The list keeps getting bigger. Against all odds, we've got a massive, huge list today. A lot has happened over the past week. Michael, Liam, how are we doing?
- [01:46]
Liam
Doing fantastic. Yeah. Super excited to be back. There's no shortage of things going on and have been seeing some very nice comments in the YouTube channel. So thank you for everybody appreciating the signal that we try to bring week in and week out. There's a lot of prep that goes into this show and couldn't be more excited to be back with you guys.
- [02:06]
Michael
Yeah, I was pausing because there was, like so much that happened since Friday. I didn't know where to start. And the only thing I have is did anybody go see the Odyssey?
- [02:14]
Brian
I haven't, but I've seen some clips.
- [02:16]
Liam
Did you?
- [02:17]
Michael
Yeah, I. I did it. It's actually. It's pretty incredible. Like, I don't actually know fully how
- [02:22]
Brian
to feel about it.
- [02:23]
Michael
I need to go see it again. There's just a lot that was in there.
- [02:27]
Liam
Yeah.
- [02:28]
Brian
How long of a movie is it?
- [02:30]
Michael
They said three hours. I mean, it definitely doesn't feel like three hours. It you fly through it. The biggest suggestion and you could see those Clips on. On Twitter is go try to see it in imax. Because you see the video, the movie, and then you see the clips like it was filmed all in the. The IMAX or the large film. So you just naturally are missing parts of the movie. But yeah, I thought it was really good.
- [02:53]
Liam
Next time there's a rainy day, I'm definitely gonna see it. I was going to, but then it ended up being nice.
- [02:58]
Brian
It's one of those things you'll definitely
- [03:00]
Michael
regret not seeing in the theater.
- [03:01]
Brian
Yeah. Yeah, that makes sense. I mean, it seems like the reviews are mixed. Some people are saying it's quality. I think people Stan, Christopher Nolan, no matter what, to a certain extent. But then other people really hate it and say that it's bad. And so.
- [03:18]
Michael
So here. Here's where I'm at. Dox myself a little bit in like, ignorance and like, I didn't pay attention a lot in school. So, like, you remember Homer, you remember the Odyssey, Iliad. But I think what I've seen, the people that, like, hated it were the people that were super into it. And it obviously skews from the book and you make it in modern times and there's all these stories. I thought it was from just like a pure, like, going to watch a movie and this notion. I started listening to pods before about how the hell did this thing last 3,000 years? And there's a lot of interesting tales that go along with it. I thought that was fascinating. And one other thing, and then before we lose anybody, because I know, I think Brian would appreciate this. And I mean, you tell him is there's a show I saw like some VCs or somebody on Twitter talking about. And it was multiple. It's called the Agency. Easily the most phenomenal, smart show. It's about like. And it's helpful when you look at the world through like a kind of psyop lens where there's a reason why Iran and like, you know, the Middle east and the people there are always pain as a bad guy. Like, there's just real inertia. The past 30 years post 9 11. So like Homeland, all this stuff. So that you have to like, look at it through that lens. Like there's iron component, there's a Ukraine war, but it's all about like espionage, CIA double agents. And it's incredible. Anyway, what.
- [04:31]
Brian
And it makes you think platform is this on, dude?
- [04:35]
Guest/Host
At this.
- [04:35]
Michael
At this point, the Apple TV is just like playing paying for everything. I think like, you can get to through Amazon, but it goes to like, Peacock or Whatever. I don't know, but I'll check it out.
- [04:45]
Brian
All right, that's enough. Movie.
- [04:46]
Michael
And you know you've grown up when you're not there's a movie, there's a
- [04:49]
Brian
movie on the list.
- [04:50]
Michael
You know you're growing up with. You know you're growing up when you're not torrenting anything anymore. Like I used to find my paying for this and now you just click the button.
- [04:57]
Brian
17 different streaming services. All right, let's get into the list.
- [05:02]
Liam
Enough.
- [05:02]
Brian
Enough banter. We're going to start with the biggest story on Friday that bled into the weekend. Kimi Moonshot's latest model, K3 was released, I believe Thursday evening. And basically, you know, all reports are suggesting it's as good, if not better than the Frontier US models. So Claude, Fable 5 and OpenAI's GPT 5, 6 Soul, I got a chance to play around with it a bit. I did subscribe before, earlier, I guess last night or this early this morning. They stopped all new subscriptions because they were basically hitting capacity limits. There was a ton of new subscribers, new people looking at Moonshot and Kimmy. There's a number, number of different angles. We're going to take this but initial thoughts on high level open source versus closed source so the big sort of narrative around, you know, why this is important is that Kimi is not only more open than open AI, funny enough. And Claude, the the weights for this latest model are not released yet. I believe they'll be released on July 27th. So that's when it'll be sort of fully open, but generally more open source than the US Frontier models and seemingly more efficient and so cheaper on a sort of per usage basis. Initial thoughts, guys. And then we can take this in a number of different ways.
- [06:38]
Michael
So I want to go like kind of meta level because I really don't know what to think about it. And then I know Liam has a take on just like the capital markets where value accrues and all the relationship between open and private and we can go deep there at the meta level. And I think this is still debatable or is not like deeply acknowledged, but I believe it to be true, is that there has been, whether it's 2, 3, 5 years a race for ASI, not AGI. I think AGI is already here because the notion of what is artificial general intelligence, I think that if we would have gone back 10 years ago and seen what these things could do. Like this is AGI. Like I these things are smarter than 90% of people, especially if we look at the benchmarks of how dumb people are getting. And so there is a race there. And that's really where if you go back, you know, you can put the trajectory on DeepMind, Google, Silicon Valley, not fully believing that Google should be leading that or whoever leads and gets to asi. Artificial superintelligence. Super yeah. Would. Would own this market. And so you have the founding of OpenAI and so you have these Silicon Valley entrepreneurs understanding this and there's a race for it. And I think globally now there's a race for it. I think my interpretation is there's different implementations on how the hell you get there. And you can see this in the Frontier Labs now we can go into like some of those ideas. I don't. I think it's above our pay grade, but I think that's what's happening here now at a just individual perspective. I deeply find it interesting and I think most people would to the point that you made or in some of the tweets that I saw out about the guardrails. And simply whether it's. It's not even necessarily about being woke. It's just that like, you see with these models, you may need to just like it's not skirting the bounds, but you need to go outside of the traditional thinking and they just will not produce. And I thought that was the most fascinating reason to go download this because I have all these subscriptions and I'm like, well, I would love a bot that just talks to me objectively and not has this alignment and these guardrails. So anyway, that's just more of a meta. Like it's very interesting. I don't really know what to make of any of. Of all this because of there's a race out there. But like their versions of how they're implementing it in introducing it to consumers. There's different strategies and I don't know what to make of it yet.
- [08:57]
Brian
Liam, what do you got?
- [08:58]
Liam
So there's so much here and honestly, massive props to this team for this model. So to the point of super intelligence and AGI, it's completely here. The model suggested the way that it trains itself for the next generation too. So it's way like essentially telling the entire team how it thinks it should be trained in order to maximize optim, like its deployment. Next it's available on all these different chips for any type of hardware can run it and it's developing its own chip in house in order to optimize its performance later on. So those are all incredibly interesting. The fascinating part is Nvidia is still not available for this company and so it's run with so much less money than all of the Frontier Labs. And the shortage of compute for training and inference constrains itself but it's still pretty much there with the rest of the Frontier Labs, which suggests that maybe there is way too much capital that's going into these frontier companies. And the real thing that we've been talking about for a while and even seen with Deep SEQ is that constraints breed creativity. And then the last part is that ultimately I think that the frontier level intelligence that LLMs create is more going to be akin to oil. Like there is going to be some money made in actually distilling it and offering it to the public, but at the end of the day it's probably more like a commodity where pretty much anybody can run it and there won't be a ton of value derived from the actual models itself. It's more. So how can you get the right data to, you know, train the models to give you the right exact tasks, whether it's one something that we'll get into from the thinking machines company later on or the data brokers that feed and help to train these models?
- [11:03]
Michael
Yeah, I mean I don't necessarily know like this is where it gets a little rough outside of our bounds on competency, but I don't necessarily know if all that's true simply because my understanding of the models and they're growing like the weights and number of effectively like the term's escaping me but like you need the GPUs at least to date because that is the highest performance. Now the TPU stuff is what Anthropic's using per Google. But the reason why these firms have been able to keep up is to your point, the distilling and that kind of if you had it back in. The easiest example of why they're gating a lot of the models is so they can't be distilled older. They're at least trying to effectively
- [11:50]
Guest/Host
because
- [11:50]
Michael
the, and this is a, the, the most fascinating pot on all this and I've referenced it a few times, it's a slog. Was that Leo Ash? I won't, I won't get his last name right. Dharkesh pod from 22 I believe. And the core point is the race is that you're trying to get six months ahead of the, the other firm because that'll make all the difference in the world once you get to like the real like ASI and the the distillation and everything else that will come from it. So the point is that like yes, they figured out to do more with less in the Huawei chips but there's a reality that it's still there are steps behind on by design. And what does that mean and how does that matter? I think like that's the core idea is that there's a level of intelligence that the, that China still doesn't have and it has to do with the export controls but there's a whole like argument around do export controls actually matter? Will they do anything?
- [12:43]
Liam
So yeah, but I, I'm just not sure that the six months is here right now because I think fable came out like what a month ago at this point and Kimmy's here now. That's relatively on par. So I think it's just closer than we probably would anticipate and like we'll, we'll talk about it soon but it seems like Quinn at least is hyping up that they have a new model coming out this week that's going to they think be on par with some of the other models too.
- [13:11]
Michael
Well, what I'm getting at is that we don't necessarily know if how many models ahead like anthropic or OpenAI haven't released and that haven't released and you, I mean we can bet that they do. I mean it could have been 20 years ago that they've had this stuff. So I think that's the core idea is that they don't have. I think it was the semi analysis research on like we're kind of inversely a relationship with the strengths where in the west we have the compute and we don't have the power. In China they have the power and they don't have the compute because of the export controls with the GPUs. And that's why they've had to do more with less because. And then they've also done it from a strategic perspective effectively collapsing which is really smart long term because they own the full supply chain. But I think that's like the, the power dynamic is while they are keeping up and there's a whole other angle of like maybe some of this isn't true and there's a capital markets angle around AI and like you needed something to absorb a lot of this liquidity. That's what I say. It's so hard to, it's so hard to reconcile what's happening. The mental model I have and I'm happy to change it is that ultimately they're they're racing for asi. And I was talking to somebody about this and we, we've hinted at it a bit, we came to different outcomes, but it's like these large incumbents are effectively treating like the dollar doesn't matter. And I come at it from the version of like we're moving away from the dollar. He came in from the version of the debt will be burdened by the consumer. Right. But it's the notion of like the hyperscalers and that the money doesn't matter anymore. And the angle that I've been able to steal is like, because they're chasing si and if you come at it through that angle, um, then you look at the world of like they're chasing this because at the other side of it, like money does it or at least from that money, China's coming at it from a different angle on how do you get there? And so like that's this thing I can't like fully reconcile on. Are they trying to stay ahead on the consumer side? Are they trying to get to as high and they don't care about the consumer side. But then there's the inputs of like how much you're training the models and how many people are using it. And then the last part is like by using open router and sending all this data, if consumers start to go so they like hamstring themselves so they don't get the distillation. But now you're giving more data over to the east because you start to use these tools. I think it's all happening in real time and it's a fluid situation.
- [15:41]
Brian
Yeah, I mean I would say of those sort of goals or initiatives you reference, I don't think any of them are like mutually exclusive. I could be, I think it could be a combination of all those things. But just a few things to reiterate on all of this. Anecdotally from using it over the weekend, the part I was most impressed with was effectively as you're watching it think and work, it is I would say much more robust in terms of its inherent sort of checking itself and iterating in real time. Whereas like, you know, with Claude or OpenAI, a lot of that is still manual or you can do various like looping techniques to get to that outcome. But Kimmy just right off the bat is iterating several times, checking its work, reviewing it and just you know, basically allowing you to one shot things pretty easily. Michael, what you referenced I did want to pull up. So this was a tweet from Callie over the weekend, Bitcoin developer saying, I have a report full of security issues of a software I'm working on, codec. So OpenAI won't fix them because of cyber guardrails. Fable won't fix them because of cyber guardrails. Kimmy K3 fixed all of them. No restrictions, just gets the job done. This will end badly for OpenAI and Anthropic. And then David Sachs quote, tweeted that saying, Kimmy K3 just fixed 15 critical security bugs that Codex and Payable refused because of cyber guardrails. There's no reason to limit American models on tasks that Chinese models handle without issue. We're only making ourselves less competitive. And then, you know, this all relates to a headline from this morning that the Trump administration is contemplating banning, you know, Kimi and other Chinese models because they see this competition happening in real time and they see effectively the usage. And so this was a tweet from Kobayasi Letter with a chart comparing U.S. and Chinese AI models. The proportion of tokens used by U.S. firms that run through Chinese AI models on open router is up to a record 58%. That percentage has tripled since mid January, as you can see on the chart here. So I think this, I mean, the competition is obviously real and I think, you know, it only adds to, I think, you know, what we've seen already and talked about in the past few weeks from the US Frontier Labs partly playing this like, regulatory arbitrage game where they, at least on the anthropic side are being sort of like overly bombastic about these things needing safeguards so that they can basically have the government step in and, and whether it's gating initial models or just putting basically a ring fence around, not only their own models, but then say, you know, banning Chinese models so that they can effectively attempt to sort of maintain their share. When this stuff is proliferating at a speed that's. That's pretty remarkable.
- [18:46]
Michael
Yeah, I mean, it would. So there's a couple things happening. We've obviously talked about the posturing and that I think anybody paying attention has made sense around, you know, pulling up the ladder behind them and gatekeeping. I think there's a few things that happened the past week. One was they're positioning as it's like a consortium from the industry, but about creating like the guardrails on releasing new models. I think it came out this morning somebody related to like head of intelligence or security around the AI stuff was let go from the Trump admin. And then to your point the open router usage, I think that ties into something related to like we're effectively exporting data and value. Right. Because somebody has to pay for those tokens overseas. And so it would make a lot of sense if all these things that were trying to like navigate and kind of feel around in the dark that they would eventually really make it difficult for Americans to access those models because of all the things we're talking about in the race. And that would be, that would be tough. I mean in why there's so many like vectors and layers to this is because that thinking machines aspect, that layer that Liam called out is there's a fascinating like consumer application here where you can have like a more agnostic open model that is, it's lower in kind of like the output from a lot and there's like, if you scroll down there, it'll show you the benchmarks against some of the other models. But the idea was it's not meant to like, you know, out the gate be at this like, caliber as a, if you go further down, it's really that you can train with your own proprietary data. So you end up with a model that's kind of more bespoke to you. Right, right there. Yeah. So you can see how like the, the model sits below, above 4.6, below 5, 6 via chat. But then there's this version, I forget what they call it, but you can ultimately like provide your data and input. So then it's more bespoke and that's an interesting like consumer behavior. But I think that this is fundamentally different than like the stuff that we're talking about. What's going on with the east versus west and the Frontier labs.
- [21:00]
Liam
Yeah, I completely agree that you know, in an ideal world nobody's pinging, you know, Chinese data centers, especially with the US proprietary information. I think that ultimately though it's, it's tough because. And there, there higher people up and, and people who are really deep in the like semis industry that have just really tried to combat the Nvidia export controls on chips. Right. Because those then are going over to like Australian data and it's being used just remotely to train a lot of these open models that are actually like Chinese run and just because they don't actually like, while the US can't technically send or sell to Chinese firms, they can still use the data centers either or chips just through the supply chain. And so ultimately I think that any sort of restriction on Chinese open weight models or using the data centers in particular would probably just ultimately result in firms setting up entities overseas that are in non combative jurisdictions to the US and ultimately we would probably use them there. I think ultimately information wants to be free. It's almost like the same thing as Bitcoin. Whereas this stuff is just like binary data and it's ones and zeros. And you can ban yourself from using some of these models, but you just can't ban it in general. And some people will try to go around. It's, it's similar to how folks in China can't actually use Anthropic, but they set up all these shell corporations and have, you know, people in between essentially selling or buying anthropic subscriptions and then selling it for 90 cents on the dollars to Chinese firms so that they can distill them too.
- [22:50]
Guest/Host
If you are one of the few people who still have conviction in bitcoin, you still believe in it as a long term store of value, a better savings, technology, freedom, money. I want to speak with you. If you want to speak with me, of course. But if you want to have a conversation, book some time using the link below. Love to chat with you, hear more about your background, how you got interested in bitcoin and if nothing more, just make another connection in the space. It's always fun. It's probably my favorite part of the job. I do like doing the podcast for the most part. I mean, you know, got to deal with Michael, but other than that I really enjoy it and the only thing I enjoy more is getting to speak with people, listen to the show. So book some time, let's catch up, have a conversation. There's a link below to do it and if you actually want to sign up for On Ramp as well, you're more than welcome to use the code tlt. I'm going to ship you merch personally. I'm going to write you a handwritten note and then you're going to get access to everything from brokerage, dca, Earn, custody, inheritance, insurance, you name it. There's more that I can name. Forget more things than we actually do. So yeah, book some time. Would love to catch up and hope you enjoy the rest of the show.
- [23:56]
Brian
Yeah, and a couple other related headlines that we alluded to. So Alibaba says their newest Quinn AI model, the second only to Claude 5 or Claude Fable 5 and I believe that'll be released this week. And then we have sort of the continuation of the stories around these upcoming IPOs. So Anthropic moving closer to its IPO. Yeah, what do you got?
- [24:20]
Michael
Sorry, just one thing I think this stuff mirrors to. It's not the perfect analogy, but if you saw like I just go back to Covid and there was a lot of things done during COVID with policy and et cetera, et cetera. And then. But you, when you peel the layers back, you start to realize like these kind of playbooks have been run for a very long time. And so I think about like the Internet, you probably mapped very similarly. But because we're so close to the digital asset space in Bitcoin, it's kind of starting to mirror very similar in that. And I had this tweet today about the percentage of. There was a graph the other day that was selling like 2.2% of Americans pay for an AI subscription. And you think about like Bitcoin is very close to that is even probably less that are like meaningfully using Bitcoin as like savings technology. And where I'm going with all this is Tillian's point. It's true. Like all this stuff is tack. It's ones and zeros. You can't really ban people from it. You can span yourself from it. But at the same time, we know from consumer behavior agency for hundreds, thousands of years that individuals will just go to the easiest, fastest path. Think about like AOL and the amount of people still use AOL and used it for decades. And so if you take this notion that only 2.2% of people are actually even paying for a subscription, think about how many people don't care, don't even know about this. And if you banned it and you have an IV blocker on it, like they will just never try to use any of it. So I do think that is like an important function of all this, is that whenever you make things convenient or inconvenient to access, how many people are you? And so I'm not saying that that works, but like that is I think one of the bigger issues in all of this. In the same way that ETFs and you know, centralized custody naturally make it easier for people to get Bitcoin exposure, I think that is a little bit not apples to apples in the sense that when it comes to your money and economic reality, like eventually, if you hold a material wealth, but you can make the same case for intelligence that if you're just checking out for your like recipe on a Sunday afternoon, you don't really give a shit who managed that data. But if your whole company is predicated in usage of it, and you have all these different work streams, then you're really paying close attention to this because you're like, a, how can I export all that data? So it's agnostic to the model. And then B, like, if I'm. If you're in an industry or era, that's at the very bleeding edges. Do you really want to be giving up a lot of that proprietary data to that? So I think it's just like a spectrum, like where it matters and how much it matters for different individuals. But the second order effects of making it inconvenient for a lot will be lost on most people. They just will think this is how it works. And when reality, they're getting like, kneecapped on the level of intelligence they can get.
- [26:52]
Brian
Yeah, I think that's fair.
- [26:56]
Michael
Sorry, IPO stuff.
- [26:57]
Brian
Yeah, yeah. So we got anthropic IPO, deep seq IPO and then another project with a raise. So decentralized AI project noose research in talks to raise 75 million at a $1.5 billion valuation. Liam, maybe I'll kick this trio of headlines to you. What do you make of the latest IPO plans? And then also this raise?
- [27:22]
Liam
Yeah, so that company Noose is, I don't know, is it Hermes or. Hermes? Yeah, it's like the open claw type competitor, which makes its money off of using additional premium services for data centers too. And so I think to the broader point, this one's interesting too, because open source and actually using great technologies is a distribution strategy now. And then you can sell value out of premium services too. On top of it, as we've continued to see, there are going to be massive goliaths that will fund things at a loss. And so in order to make sure that you can actually get as many customers as possible, I think that this is a very prudent strategy nowadays. And Kimmy, to the broader point too, got most of its usage when Hermes and openclaw first went viral because it was the best coding agent at the time and could actually fit inside the harnesses. And. And then, you know, once it got more usage through its data centers, I think that's what really enabled the Flywheel to train it faster and faster. But going back to the Anthropic and deep seq IPOs, I think anthropic is going to be really interesting too, because got to imagine that the company's valuation, if it was a public company, would have fallen 30, 50% over the past few days. As you can see that there are cheaper competitors that can do almost as much as Babel can do, but at a cheaper Cost. But the interesting thing and why it won't fail is just the fact that it's a U.S. company. It has forward deployed engineers that can get into the largest organizations and really just allow companies to get comfortable with how the data is being used. If they didn't turn on the subscription or the tracking capabilities, it says that they store customer prompts in their data centers for like 30 days. I think that they would still be at the same valuation or even higher because they would allow for just more companies being comfortable with them. But just the fact that they have competitors that are somewhat equal to them in actual model capability and that they actually say that they're going to hold customer data for 30 plus days makes it pretty tough for US firms to to use them. But at the same time I think that there are a lot of slow companies that will just use them because they have been the standard just like there are there are people that still use AOL to today.
- [30:01]
Michael
Yeah, this is super interesting. We'll have to unpack more. But I think this is part of the whole reason early writers this pod exists is like there's a version of slow destruction of capital happening here where I think it came out there. I forgot where it was after Kimmy. The model came out where their private valuation. Oh, maybe it wasn't a private valuation. Maybe it was just the IPO chances or maybe it was at their evaluation took like a 20 to 30% hit on where they would go. I think that was like a like poly market out there that was you know, X trillions and then it like went down. But to your point, independent of that it just came across. I saw like zero hedge with SpaceX like one of the largest pump and dumps so far. When you look at like where it was trading at at launch and then today. And I think that a point stands Liam, that independent of what's out there outside the US you're going to have the regulatory bodies and the large enterprises and largest capital markets required to use US based firms here you're going to have the dumping on retail and pensions, et cetera, et cetera, that's going to like at least have its valuations for a while. But the reality still stands that if these firms are using not the best technology in the best way at the lowest cost. When we talk about like Noose and all these other solutions that can route and leverage the models at the appropriate usage of tokens, you're just effectively pushing out the end state of these companies and their existence not not being there. And I Think that's the big story of what we're seeing is that there's just a slow destruction of like US capital and people are going on and like we know this with like SAS valuations and the multiples on all these companies but people have to go in. We've seen the centralization, the magnificent 2007 etc and so I think that's just effectively what's happening here that a lot of capital markets are like slow walking off a cliff. A lot of destruction of capital. And then the new stuff is interesting because randomly our buddy Mark Yusko was an investor there, he has his podcast on Saturday and they were talking about it like those guys, they invested, I think they were invested earlier than this. But I think there's a real interesting angle that we're going to see sooner than later around some of these aggregators that start to accept BTC or stablecoins. Because when you think about routing and just efficiency and you can just already squint and see how you're going to naturally need Bitcoin for a lot of this, that there was a report that came out the other day about the GLPs and the, the peptides and accepting cryptocurrency at like this crazy pitch versus anything else because a lot of these places won't let payment processors go. And so you can just start to like navigate where these firms and the people at the bleeding edge of this are going to start to mirror and accept it. But then that's the whole thesis of what we're building and we're talking about is like these companies that recognize where the value is going to aggregate and then really inflation if they're competing in a open field, the other firm that's still trying to make dollars and not really understand inflation and all these other things, they're going to naturally leap ahead. So there's a lot like on the frontier side where the capital markets highly suggest too in your diligence because where these firms trade at their Open may not be 30, 60, 90 a year from now. But the other side of that is these upand cominging startups, they're eventually going to get crypto embedded into it and then, and then rally around bitcoin.
- [33:37]
Brian
Yeah, that's well said. One, a couple other things I want to touch on before we move on from AI. Michael, I think you flagged this and I also saw this independently but a tweet from Gavin Baker, managing partner at Atreides. I'll just read some of this and get your guys thoughts. But the Mega bull case for AI infrastructure would be if market share shifted away from certain frontier labs with 90 plus inference margins towards cheaper models whether open, closed or whether open source or closed. It would increase the ROI on AI spend for end customers by increasing intelligence per dollar which would drive incremental token demand. Margin dollars would effectively get redistributed from the frontier labs to AI infrastructure performance providers. The infra winners would be those with the lowest per token cost and the winners at the monol layer would be those with the highest token efficiency. This is why Jensen and Nvidia are so focused on open source. But it's the. It's most is likely the most important one as I think it is probably less. He is probably less worried about a mon monopsony these days. Lower margin percentage at the model layer equals more margin dollar at the infra layer. All else equal. I think this kind of hits on a number of things that we just talked through. Whether it's just the infrastructure buildout itself open versus closed and this idea of intelligence per dollar. What are your guys thoughts on this before we move on from all the AI stuff?
- [34:56]
Michael
Yeah, I mean I think you. I was looking for that but this is kind of what I was talking about around aggregation and value. I do think the other thing we have to remember and it's not apples to apples but it really helps with the mental model. And I didn't listen to the pod but I saw the clips online. It was the ramp CEO with David Senra and he was referencing you've heard this in multiple functions. He was referencing air conditioning and like how much value did air conditioning accrue to the process versus like what did it open up for the ability to, you know, from productivity and everything to be able to be in inside, in 100 degree weather, whatever. And you've heard others talk about like Coca Cola versus the refrigerator and like refrigeration, what it accrued. And so it's not apples to apples but it gives you that like notion of it's in this like air that you don't really think about or see. And if we go back to 2% of consumers are paying for intelligence today and there's infinite use cases for intelligence, we can't even fathom that. It also does provide a bull case for the anthropics and other firms because we still don't even understand how it'll be used and it'll be ubiquitous across everything we do. So it's, it's really truly like just it, it's. You're gonna go study history. But I mean, I can't imagine anything like even you see people that talk about industrial revolution and the fact is that you never had the speed of which these things are occurring. When you think about anthropics, like revenue, because it's built on everything already that's been laid. When you think about fiber optics to the browser, to the application layer, like all those things already embedded and then they're in our phones, in our pocket. So it's fundamentally different than having to go from like dark ages to figuring out, you know, railroads and then electricity, like all of this is there. And that's why that speed is so disorienting. And then also the capabilities in the race. And so that's where it's like a true. I don't think anybody truly has like a full lens of, of how this plays out or even where we're sitting like right now, but it's going to be very different in the next couple years. And I think everyone reminds me like the bitcoin stuff where everyone thought it was going to blow up and it's taken a lot longer. I think this one probably hits a lot faster, but maybe those are the famous last words. And in three years we're sitting at the same situation, which I don't, I don't think that's the case. Stuff's too powerful and too many people are using in different ways. Like the cat's out of the bag, especially with the open source models being out there.
- [37:31]
Liam
Yeah, I think that this was one other tweet that Gavin Baker had too, but he discussed how Kimmy and any other competitor to the duopoly of Anthropic and OpenAI is literally better for every single company on earth other than those two because it means there's less pricing power. And so everybody in every business is going to use these products and services and have or be on the supply chain for it. And thus any. There's more price, there's just lower pricing per unit for everybody who is a consumer and business using these products and thus they can do more with less and not have to rely on the duopoly of those two companies. And then to your point, like, yeah, I think this is moving way too quickly, Michael. And so I didn't even hear about this company until recently, earlier this morning, but they were talking about it on the Moonshots podcast and there's apparently an open source AI project now called like Liquid AI by some former Microsoft guys that's on parity with GPT5 5 that can literally just fit on your phone. And so the stuff is pretty much already out there and already going on to the smallest size as possible. So very fascinating and should only kind of accelerate from here.
- [38:53]
Guest/Host
So what we're looking at here is a chart of Bitcoin's drawdowns. 10% below, 20% below 30, 40, 50% below all time highs. And what those returns have been one, two and three years later if you bought the dip. So with bitcoin trading down over 50% at this point, now is a great time to be buying. If you think that this will hold up historically and I'm doubling down, now's a great time for me to continue to allocate. These are cheap sats and if you feel the same way on Ramp Finance is live, you can Sign up in 10 minutes and you can start buying Bitcoin immediately. And so if you do so, use the code paper on signup and we will deposit 50,000 sats into your account after you make your first bitcoin purchase on the platform. Again, that is code paper. All right, back to the show.
- [39:40]
Brian
Yeah, things are moving quickly. One other thing I did want to pull up was this announcement this morning. It relates to something we talked about maybe last week or the week before around with the proliferation of all this stuff, particularly from an enterprise perspective. Owning your infrastructure, owning your stack. So this is a company called openship that released an open source application platform for building, deploying, operating and scaling applications on infrastructure your own. So you can kind of think about this as an open source or self ownable vercel sort of comp or netlify, that type of thing. So good to see that the tools are accelerating as well to allow people to have more sovereignty over their data on the stack. Anything on that before we move on?
- [40:28]
Michael
No, but I think this is a perfect transition to like how fast this stuff is moving I think within the administration. And like what's going on with PayPal and Stripe? I think it's like one of the biggest stories in like Fintech and just Financial Rails. Do you want to kick it off and then I can.
- [40:42]
Brian
100%. Yeah. Yeah. Let me just intro this a bit. So I would say outside of the AI stuff, this was definitely the biggest headline from last week. So Stripe advent and also Block is participating in this potential bid to acquire PayPal for more than 53 billion. The reporting was a little fragmented on this. Some people included Block in the in the pressers, others didn't. But apparently they are involved and this is a large deal For a number of reasons. But I think it speaks to effectively sort of the incumbent tradfi payment. Rails, you know, have been being disrupted for a long time. This could be a massive deal in terms of basically acquiring a number of you know, basically clients or people that use PayPal today for Stripe, who is much more sort of digital asset stable coins forward than than PayPal historically has been. But maybe. Michael, I'll hand it to you and we. Can we riff on this one a bit?
- [41:48]
Michael
Yeah, there. There's a lot here. I mean I think one of the biggest ones is the ARB of Stripe just has this insane private markets valuation that I think most probably recognize they were publicly traded wouldn't have. And so there's an op for what they're doing and we're to make these acquisitions. What I thought was most fascinating was the blocks participation in there because you have multiple firms. My understanding Advent's one of the most sophisticated private equity firms. You have PayPal that's sitting out there that most people don't really know still has like 500 million consumers using it. They have Braintree which is like kind of like Stripe from an enterprise perspective and then obviously Venmo that sits on the user side and the angle ultimately. And it's funny because it's just like we were talking about in the Visa acquired pod. Like you go and look in Visa and MasterCard and really learning how those companies like not only their scale but why they exist is because there's just a fundamental issue or was an issue and mismatching around the merchant issuer. So who's issuing the ability for the like end state merchant. The, the. The whether in this example like a square terminal, the user who is the banker on that side of that. So when somebody comes and uses a card, how do you net settle and then who's the issuer of the individual? So historically you had a bank on one side that issued the card for the individual and then on the other side you had the bank that was managing the account for that. And that reconciliation was this huge fragmented process. And then that's where you ended up with this like almost agnostic Switzerland verse messaging system. And Visa and MasterCard have grown to be these large businesses that aren't banks and they sit in the middle of all this. Now when you start to understand that and it's specifically within the administration we talk about like this just kind of plethora of new technology but also efficiencies for the consumer. You have the notion of stable coins existing but you still have to change a lot of those rails. And so the three pieces when you start to tie them together are PayPal with the users, cash app, square blocks with the merchant side when you think about the terminals and then stripe with the balance sheet. And to Brian's point, you know, they just recently came in on the like stablecoin crypto stuff. They're probably the most forward as the larger companies, but they have a lot of that understanding of how they were bypassing or at least building the rails for, for businesses, enterprises and specifically in an Internet native way. And so you tie that all together and you effectively end up with a solution or consortium that can bypass and develop payments outside of the Visa and MasterCard networks and you can reduce that full flow from one wallet to another holding that. And I thought that is, it's super fascinating and I don't know if the deal will get through but you can see like if this deal doesn't happen, another one will because this is where it's all going where you need the different layers of the individual to have that efficiency, you need the consumer or to have that and then you need the merchant to be able to accept it and they need somebody to like manage those rails. And the crazy part is when you look at the Visa and MasterCard networks, they're all relying on some level of float, like some level of credit before you net settle. And the interesting part about stable coins is they change those rails so you don't actually have to you can effectively like final settle at least on that on the stablecoin side. And so it just changes the dynamic that never existed before. And so I think this is just something that's going to to independent of this deal goes through, others are going to start to put their own kind of projects together. And obviously seeing this is why Visa and then open USD is such a huge focus.
- [45:33]
Liam
Yeah, I think you said summed that up extremely well. I think that you know, six or two years ago this definitely wouldn't have happened under the former administration. And so while the PayPal board rejected the offer for saying that it's too low initially I to think that they probably get something done because PayPal is a company obviously the founders have been gone for a very long time that and doesn't have quite as much vision or really much at the moment. And so I do think that this is really interesting. The one downside that I haven't necessarily figured out exactly how to manage myself is just the fact that many consumers and businesses really don't want to pay Instantly they need the credit associated with that just because of, you know, managing working capital on hand. I do think that that gets figured out, but many consumers don't ultimately care that much about it initially. And so I do think that these companies, especially in the group that they have together here, would probably have the best shot of unseating a visa or MasterCard or. Yeah, MasterCard type duopoly.
- [46:40]
Michael
Yeah, I mean, it's a great point and I think this is really one of the. There's probably multiple reasons, but why a lot of these blockchains and specifically stablecoin issuers have developed their own blockchain because you can just manage the reversal of transactions. And I think there's a new theme and narrative with like perps and defi that you need circuit breakers because of all these different attacks on the Oracles. And so I think it's already becoming understood like that this is all window dressing, but it's an arbitrage. And we've been saying that forever. It's an arbitrage against the incumbents. But I know like Tempo Ark and these other firms, even open USD, there's going to naturally be some version of like reversing transactions.
- [47:24]
Brian
Yeah. Liam, I think you made a good point around a few years ago. Maybe this doesn't go through and we'll see if it goes through. But I think the, on its face, you could look at, you know, antitrust regulation, but in, in a different way as we sit here today. I would say that this deal or this potential deal is actually like pro competition and allows to potentially Disrupt Visa and MasterCard as you described, Michael. And this was just a day or two later that Visa then announced they're launching internal stablecorn platform for clients that provides access to ousd, which we covered a week or two ago. So this is a pretty strong signal that they do see it as a very pro competitive move in terms of Stripe, substantial acquisition of PayPal.
- [48:12]
Michael
So, and you know, like Visa's telling you, I fully haven't reconciled this, but enough to like say that Visa is telling you that their model is going to change forever. And the reason why is because they're doing this. But at the same time, Stripe is a partner in open U.S. open standard. And, and so there's this version of like. And this is, it's funny because I don't know if I, I've shared with you guys privately, but Kai Sheffield, who's leading Visa crypto, he's the one that's like brought them into this world back in the Day he was just interested in this space and he started by writing a newsletter internally. And this is just goes for anybody that's building at a company that is interested in AI or digital assets. Like you're probably the smartest person there and there are a lot of people interested. So it started just as a newsletter, a visa. The reason why I know this and I think this is public he's talked about it was I was early at the block and we were selling research to a lot of these firms and so they would read research. And the point being is that they recognize somewhere through that that they had to be at the forefront of the innovation. And you can look at the rain cards as an example when they were the first primary issuer that's existed that's not a bank to be able to issue those cards. And so they know the world's changing. And I think very similar to the AI discussion that nobody knows like how it's going to really diffuse into the world. And the big part of that and we talk to banks about this is ultimately like how do you get the merchant, the business to accept these things? How do you get the user to change their behavior? And a lot of that comes back to what are the things already installed. You have square terminals installed and I'm not saying it has to be exactly like this, but you have the square terminal installed and then you have people with Venmo or Cash app on your phone or PayPal and you have the login there. So you at least like bypass and can obfuscate how do you download a Net new wallet. And so that's really, I think the V1 for a lot of this is there's going to be more of like the B2B net settling and the user won't really know, they'll just get better fees and they'll be incentivized via some clicks or whatever to get 5% off a transaction. But I think Visa understands that we don't know this to proliferate. So we just need to get involved in different levels of the stack.
- [50:18]
Brian
100%. On a related note, we talked last week about Japan being a little bit more forward thinking in terms of approving or going live with Bitcoin ETFs. And they're also at least on the enterprise side leading into stablecoins as well. So Amazon's largest delivery partner in Japan is integrating a yen backed stablecoin. So this is sort of exactly what you're just describing, Michael. Like more on the B2B side, the actual End consumer may not even realize that stablecoins are being used. But in terms of internal payments to employees these things are being integrated at a, at a rapid clip.
- [50:56]
Michael
Yeah, I think it's super easy for us whether it's the tokenized securities or here where we can say like I don't know, I want it. We all have our own versions of frame of references but when you look there's all these different anecdotes you hear from whether it was Bridge being one of their largest clients was SpaceX because the amount of terminals or whatever you call the dishes they sell and the different third party vendors and Uber's another example. They have a multinational footprint and so they need to net settle. Most people are accepting their payroll in dollars. And so you take that and then this is Amazon and Japan looking at I guess there were contractors. How do you like pay them? You can pay them hourly, daily, weekly, whatever. And this actually goes back to one closer to home here in Texas, Vantage bank with Custodia they worked on that first stablecoin cross border payment or it was really tokenized deposit and it's because Vantage has a footprint in Mexico and they work with family offices and businesses and it was just more efficient to send those deposits on a like T plus one or whatever versus waiting a few days. And so I think like this is just naturally how you're going to see it and then more and more firms it'll be a value prop like why wouldn't you especially if you're an emerging market get your payment for whatever reason. So this is just interesting I think anecdote to see it firsthand. People are using it especially companies the size of know SBI and then obviously Amazon
- [52:17]
Brian
100% not sure who added this one but IBIT options exercise limits to rise from 250k to 1 million contracts. Then we also have this headline. I thought this was a misprint. Citadel securities invest 400 million in crypto.com at a $20 billion valuation. Am I taking crazy pills? How is crypto.com worth 20 billion?
- [52:44]
Michael
Look, there's a weird world out there. Blockchain.com if you look at them and they're fundraising. I'm not saying that at the same scale but you've seen these crazy like series C, series D. There's capital markets, there's forward order flow, there's trading. These I've discussed with some people there like they're. These firms aren't everything that they make out to be. From a like you know, retailer window facing perspective. I'm not saying I understand like why this deal happened. I just know that there's a lot of interesting angles. Crypto.com is a fascinating business because they came out of nowhere, they've been around, they withstood a lot of the volatility and bankruptcies that existed in 22 with all the lending collapses. This is their first formal investment, my understanding. So yeah, I mean super question. I know Liam probably has an angle but the head scratcher is real. But this is a product of where we sit in this world that SpaceX did its IPO and Citadel's investing in crypto.com at a $20 billion valuation. Also we don't even know the terms of how much was cash.
- [53:56]
Liam
Yeah, just before that on the IBIT one I thought was great to see just because it's obviously going to have hedge funds really be asking for this otherwise they're not going to deepen the amount of options that's on it. So they are putting on larger and larger positions. They just want to protect themselves from downside positions. And I bet with the notable 13F filings we'll see more interesting players come out. But crypto.com I think Citadel has made this is their third investment recently. They invested in Kraken also at a $20 billion valuation. Kraken has public financials. They are making like I think $500 million in EBITDA last year. So like they're a real business and going public soon. But they also invested in ripple and crypto.com is another interesting one because yeah they've never raised outside capital but they launched their own token. I think it's Kronos similar to how Binance launched the BNB token and like that just seems to be what the og, you know, firms that weren't bitcoin only just did back in the day and so it is their first outside fundraise because the token's kind of crashing and the broader play likely is that Citadel is paying for their order flow and that's either both in terms of monetizing any buys and sells that they have on their platform and the order book for digital assets as well as tokenized stocks and otherwise this play just doesn't necessarily make sense with just equity and no other sort of voting rights or information rights on the deal.
- [55:40]
Brian
Yeah, I kind of forgot they had a token. But you're right, it is down probably 95% since 21. All right, a couple other digital asset headlines. European central bank taps 36 payment providers for year long digital euro pilot and then we also had Dragonfly first mark lead $38 million Series A for stablecoin startup Velocity with support from Coinbase, Ripple and more. And then this one's pretty interesting from Lynn Alden new private equity firm generally. Or she's basically launching a new private equity firm called Orange Juice that looks for cash flowing businesses, cuts costs and ultimately sweeps profits into btc. So this is something that we've talked about on this show and internally for a number of years around this is this is a sound strategy in my mind. You find something that's cash flowing. This has also been part of our critique of digital asset treasury firms generally
- [56:40]
Liam
is
- [56:42]
Brian
it's never been that we don't think businesses should hold bitcoin. We absolutely do. But you should have a cash flowing enterprise on the side of that that is generating some cash flow to sweep into BTC and build a bitcoin treasury that way as opposed to just being purely financial engineering and having no actual cash flow source. So any of those headlines you guys want to dig into?
- [57:07]
Michael
Yeah, I mean I think the Velocity is just interesting in the sense of like there's going to be more and more and it mirrors funny enough like we've been talking about the open router Hermes models of you need aggregation layers that can help you route. My understanding of Velocity does more of the routing to different banks globally in a compliant way because that's really what enterprises are going to need to do is like plug in and then turn on movement of capital across the world and then that settling into the banking layers that you need. The Orange juice stuff is I think in theory is good, I think in practice is much harder when you. It's a, it's a hard thing to execute on and the main reason why is because it's the same angle, it's just different layer in the stack. Like if you look at it at a barbell you see a lot of the stuff we talk about with like floor forward deployed engineers and these notions of even kind of like what Chamath is doing at 8090 of how do you go and take private equity funding and disrupt businesses and infuse like AI as an example, Net change management. In time I think in some respects you can maybe make the return but in most respects it's going to be a net new emergent company that will just eat the lunch because they'll just be native, they'll understand the different layers. And so the other side of the barbell is that that's kind of like what we're doing on the investing like at the ground level, I think that middle layer is where they're coming at it, which is, okay, how do we find a business with a certain amount of cash flow? Where that ends up a little tough is ultimately you have to have the operational chops and excellence not only to understand that area of the market, but also the tools being very at the front of AI change management. You have to be able to change the internal dynamic so they can understand how to do more with less when it comes to. Because that's a hard philosophy to change an organization. Why they would sweep it into btc, how you can manage that if they're leaving the equity stack. And the last part that usually doesn't get talked about enough is the person that's willing. They're gonna have to like net sell because there's very few people that would be willing to like take the equity while also understanding a bitcoin strategy. And that maybe that doesn't make sense. It's like the core idea of why the stretch stuff didn't necessarily make sense. Or a lot of these firms or there was a couple.
- [59:31]
Brian
There's not a lot.
- [59:32]
Michael
But we'll talk about investing in real estate and putting in bitcoin. It's like you need the person that's going to develop the real estate to be on par with the strategy. There's not enough people that are world class that are operating in that world will also understand bitcoin. If there was, we'd be at much higher. It's just the TAM and we haven't penetrated that area of the market yet. So you add all that together and it's really good in theory. But when you get into practice, it's really hard to execute. It's not saying it can't be done, but I think it's really a function of time and a lot of these things are right. It's just too early. And I think this is one of those things as we've seen, might just be a little too early. But one last thing is that the scorecard's easy.
- [60:11]
Brian
Right?
- [60:11]
Michael
It's like 10 years from now, $40 million. What's the Bitcoin price? Does the enterprise value of those companies or the return to investors end up more bitcoin or less? So we'll have like an objective measure. Anyway, that was the last one.
- [60:25]
Liam
Yeah. I think that the idea is very sound of wanting to create a bitcoin company or like Berkshire Hathaway company of bitcoin. It's much harder to do it on paper. I think that it's interesting too if you want to have a long term never sell strategy and have number of different investors on it too. Ultimately like funds and other people at some point if they're investing on behalf of LPs probably need to get them back capital and if they don't plan to go public and maybe they do, but secondary transactions are sometimes tough unless they're just spitting out a ton of dividends. But the last part is, yeah, I think that it's almost similar to the those forward deployed companies that are just like, okay, we're going to take AI and infuse it into a call center and then it's going to drive down costs. Like it sounds great in theory, but I'm not sure that there's that much differentiation in that strategy. If almost everybody can do it, at least in theory, there's obviously a little bit difference in how many people can execute it at a world class level. And so it's ultimately just up to the people and can they do it or not with this type of strategy? Yeah.
- [61:45]
Michael
And one other angle, just to contextualize it is it's truly more art than science today. And that's why I think that it ends up being hard if you're not doing it like because every organization's a snowflake. And what does that mean is so it's more art than science if you go on the bitcoin side because do you save more bitcoin? You need the entrepreneur that can, can operate at the highest level on do I invest in growth or do I save in bitcoin? Right, because you're having, you're doing cost benefit analysis of like do I stack that or do I go and grow? And so that just really requires somebody to do it and that has done it. Manage the volatility of bitcoin. And then similarly what Liam said on the, it's art when it comes to AI diffusing and implementing. Because if you go to an organization that's running well, well you have to be very careful on how you're going to change that up or they'll leave. And then you have to know where to park it and how to do that analysis. And we're like at the bleeding edges of where AI is diffusing into the market on the enterprise side. So there's very few people that have even done that yet. So you take that all together and it's really again great in theory. I think that if you did one company to do and you took entrepreneurs that had done it before you get a decent shot at it. And that'll still be hard. But to create a holding company to do that just becomes I think pretty difficult.
- [62:59]
Brian
Yeah, all fair. I mean I think the 11 point you make Michael, around buy in from the entrepreneur and being willing to do that analysis of spending on growth, hiring or just saving in bitcoin. I think the dynamic changes a little bit if they're just buying them, buying these companies outright and, but then the, the counter to that is well then maybe you lose some of that operational expertise of whoever the founder or entrepreneur of that original cash flowing business was.
- [63:27]
Michael
Exactly. And you have to be able to run it at a very high level on manage both respects, the bitcoin manage treasury from buy versus growth and then also on the infusement of AI and there's very few people on the planet earth. You put them on like a hand that can do both right now just because it's all net new. So that's why I like the idea and I think it's easy. Like if anybody wants to start a business like that, you probably raise if you know the right networks because it's a great, at least on the AI side the bitcoin stuff is a little, little difficult because if you went into a market they'd be like why don't I just buy it? And the last thing is to your point Brian, like if somebody really is a world class entrepreneur and has a great cash flowing business, they can effectively get the debt and go buy a shit ton of bitcoin.
- [64:07]
Brian
Yeah, you can do this yourself for sure.
- [64:11]
Michael
That's and that's the hard problem with all this is the people that will do it are not the people that are world class. They're trying to get out. And that's the thing that the market, it's like the second, it's that last layer that we talk about that people think oh you're hating or whatever when we talk about Stretch. It's like when you go into a room and you're like wait, so I gotta buy this like speculative thing and then I gotta layer all this other stuff on. It's like I'll just not do any of that.
- [64:30]
Brian
So yeah, okay, moving on, a couple of Tether related headlines. So Tether's USDT hits two year countdown threatening its position on US crypto platforms. If you remember when the Genius act was signed into law, I believe it gave Tether three years to basically get compliant in the US So we've had one year. So they're on the Two year clock and then similar or sorry, not similar but different headline related to tether Bolivia weighs adding Tether's USDT to its national payment system. Thoughts on either of these tether headlines.
- [65:04]
Michael
I thought they were interesting in like they were juxtaposed, right? So tethers kind of. I think the main point of what came out was that they haven't formally produced anything and they have little uptrend on the USAT side in the US and then there's the other side which is that like gray market we've talked about that they're really excelling in doubling down whether it's investments or to your point, in Bolivia sitting alongside the US dollar for their global payments or their payments from a sovereign for their own people or globally is just fascinating to watch play out. And I think it'll be fascinating to watch play out over the next two years to see if they're going to double down on the US or say outside. It came out last week that Binance was actually coming back into the US heavily. So it'll be interesting to see how those markets form.
- [65:48]
Liam
Yeah, I completely agree. It's not on the list, but I think it's worth at least touching on clarity because I think that they're going to speak this week and probably have a vote I think within the next four weeks. That came out that Russia is I think voting on bitcoin and digital asset regulations. They've had a ton that's come out over the past few days or few weeks of just net new companies offering bitcoin and crypto digital asset products. So I think that, you know, while the odds, I think right now are like something like 35% that it will get passed before the end of the year. You never really know with these types of things. Anything could really happen, especially as they see the rest of the world kind of coming together and trying to get the regulations right. And the US obviously doesn't want to necessarily, you know, be a laggard in these types of situations. I think it's probably just down to the ethics situation, which we've talked about before.
- [66:42]
Michael
Let's call it here. Does it, does it get passed this year?
- [66:50]
Liam
I don't think so.
- [66:53]
Brian
I want to say yes, but I just don't trust that our government is competent enough. They should, they should pass it. But I don't, I don't, I don't think so.
- [67:04]
Michael
I think it's done.
- [67:05]
Liam
All right, well, well, we have a perfect sample size. It's right with the odds. It says 2 to 1 odds right now. And I, I do think if it wasn't for all the AI stuff which will just soak all the headlines out of the room too as well as the fact that like Trump launched a token and shit that they would get it passed. But there's just going to be too much distraction with, you know, at war with Iran. Yeah, it's just going to be tough to do by the end of the year.
- [67:32]
Michael
I will say that it is a bullish catalyst if it does because it's one of those opposite of buy the sell by the rumor sell the news of like nobody's expecting it. It's fallen by the wayside and Spidey senses. I've been saying this since it started and it's picking up now. I think it gets done because it's just like. Well, especially what we've seen, global aids, they're acting like it's going to get done and I think we get it.
- [67:59]
Brian
All right, you heard it here first. Okay, a few more things. I know we're over time but there's a few more things on the list. Early riders. We put our mid year investor update out last week. Liam, maybe I'll hand that to you in a second to talk through. Anything you want to talk through there. But then also on the onramp side, promotion this summer going on. Bitcoin is back on sale and we're getting back to the basics. So we launched 0 fee DCA and 50 off all buys and sells. No cost IRA setup and discounted multi institution custody. So check out this landing page onramp bitcoin.com lp backtabbasics. You can open a free account and get started with code basics to unlock all of this. Michael or Liam, anything on either of these two things.
- [68:42]
Michael
Just real quick on the basics. I think this is going to be a meme that's going to come together. You're starting to hear folks talk about it just like deep in the bear, how there's nobody coming to save us. There's been multiple folks I think when Alden had a tweet about you just got to realize bitcoin's fundamental set the floor and the rest will take care of itself. Parker was on a pod recently talking about it and then even our buddy Eric Balchunas was saying it as well. Like I think the reality is risk adjusted. There's never been a better time. The fundamentals are so strong and so we have a lot out there. So yeah, excited to get that rolled out this summer.
- [69:19]
Liam
Completely agreed. And yeah, in the mid year letter too, talked a lot about just the fact that bitcoin has almost gone mainstream and now it's deep down in the bear again. There's a real lack of attention. I think something like 80% of all BC dollars have gone into AI this year and that's just related to infrastructure. I think that there's a real lack of focus, understanding. And on bitcoin a lot of over leverage has been washed out. I think that Larry Fink also said that when looking at the bitcoin side recently too and he's more excited about where it will be in one year. So yeah, I highly recommend checking it out for just what we've been up to and you know, the types of deals and verticals that we're focused on for the rest of the year and into next year too.
- [70:06]
Michael
Yeah, kudos to Leah. It was an awesome letter. We had a lot of investors reach out thinking like the insights, they're going to join our private call. And yeah, I think about similar with bitcoin from a infrastructure perspective. There's never been a better risk adjusted time because whether clarity gets passed or not, we know the market's going to continue to move forward. And these companies, they have certain pieces, right, but they haven't put the whole piece together around movement of capital, Bitcoin as a store of value and then leveraging deflationary tools, specifically like AI to build the companies of the future. So super excited where we're headed.
- [70:39]
Brian
Sweet. Okay, maybe time for one more thing. Michael, I think you said you had a take on this, the Netflix AI filmmaking startup. Was it this one or you know, rays.
- [70:50]
Michael
So they were, they were, they were both tied together. I, I've just naturally, you know, it's all interesting, right? Like Ben Affleck's interesting, he starts an AI company. I think what really came on the radar was Ben Affleck when he was on Rogan, he had just like a very, you know, nuanced take on AI. And then it came out that his company was acquired by Netflix. It just came out this week that it was over half a billion dollars. And it was a recent, you know, company I think started a few years ago. And the core idea was that. And I'll butcher this so you can go do your research. But the core idea was that there was like a unique model set up for every film because every film has to require the different lightings and all the different things that a producer would need. And so they took and distilled like whatever models and then they were able to create this like, construct. And there was a, a anecdote or case study around some, some deal that Netflix had rolled out that was like 12 episodes, that they cut the production cost in half by leveraging this tech. So that's one side. The other side was another firm that is in the, the suno that is in the. It's in the, you know, news right now because they're going through litigation. It's similar to like Napster, Spotify, kind of like music rights deal where they allow you to effectively take anything and create music from it. And they're now, they had raised like $250 million, I think, last year. And it's really changing this whole notion of like, what does it mean to be creator, musician, art, rights to the music? And then it's being published. And I'm sure a lot of people will like this because we're, we're not Luddites. We're just like, this stuff's moving so fast is if you know something's created by AI, you're like, why am I going to listen to it? But I don't think that's how it's going to be forever. Especially when you hear some of these tracks that have gone like, you know, mainstream or viral. And so I thought it was just unique to look at how like, AI is going to, you know, permeate and change almost or all industries and creatives. Like, people think that creativity is tied to the human. And I think where it all ends up is there'll be the drive and like, willpower that is like, human in nature. But eventually these tools will come up with creations that are so much more like, outside of what we could have ever fathomed. Still working through it. But I just thought it was interesting both of these in the same week were in the news.
- [73:02]
Brian
Yeah, for sure. On the music side, just anecdotally, I remember a year or two ago some things were going viral as like, high quality AI music. And I think we're sort of magnitudes better than that today. And if you just try to forecast going forward, like, I think you're right that at some point it's probably just going to be very hard to distinguish. Whereas, like, right now as we sit here, you can pretty much easily know when either a video or music is, is AI generated, but that those lines are just going to continue to blur and it's going to be more. Liam. Anything?
- [73:37]
Liam
No, I think it's fascinating. I think that, I mean, I listened to already using like 11 labs like a Lot of like, articles that are just AI written. And so I ultimately think if there's a human in the loop that's actually creating or ideating on what the music should sound like and not actually just executing itself, it's similar type of concept too. And so I don't see why at least this would be. This wouldn't be popular with at least one subsection out there. So. Interesting to see. I haven't really listened to any AI music, though. I'm not gonna lie.
- [74:09]
Brian
Yeah, that's fair. I mean, the other angle I think about is like, I think right now there's a perception issue with something that is whether it's text or music or film. It's like, there's a perception issue of, like, okay, if it's AI generated, it's lower quality. But even if we get to the point where it's such high quality from like a signal perspective or an education perspective, I think ultimately people stop caring about whether it's AI generated or human generated. Because if you're getting value from the piece of content, whatever it is, I think that perception issue starts to fade over time, basically, as these things just get better and better and higher quality.
- [74:47]
Liam
Agreed. Especially if you can have like, other people vet it for you that it's not just like actually complete slop. Because sometimes, you know, when you start to look, look at something or listen to something, it can sound like it initially.
- [74:59]
Michael
Yeah, yeah, we, we. This is a more of a TLT thing. But you can just see the market bifurcating because of capital and inflation. That what used to be valuable in a human world is going to be like, for the upper middle class, to wealthy people. And you think about, like, music as an example, going to live shows. That's where like, you're in this barbell of people that can make their money because all the rights are going to go to like, zero effectively online. Because it will get there. Like, I've heard things and I'm like, ashamed I got to turn it off. I'm like, I'm not listening to that. But it's good because. And you've seen like, Rogan pop stuff up or they'll do like, recreations of like, like best hits. And so, yeah, like, the world's just going to be fundamentally changing fast.
- [75:41]
Liam
And.
- [75:42]
Michael
And I don't think we're have a choice to like it or not. To your point of education. I went down that rabbit hole, Brian, after you shared it. And that's also just a complete thing. We're not prepared for that's. On kids education. Forget about, like the whole notion of MBAs and what's going on with people in school. Yeah, 100%.
- [76:02]
Brian
All right, boys, that's a wrap. Thank you as always like and subscribe Rate 5 stars. See you guys next week.
- [76:09]
Liam
Thanks Johns.
- [76:10]
Brian
Thanks for listening to this week's episode of the show. If you found the information valuable, please share the episode with a friend or leave a rating on your favorite podcast app. All the links we discussed in today's show will be in the show Notes inside your podcast app. Before we finish, a quick reminder that On Ramp Media is for informational and entertainment purposes only and nothing should be construed as investment or legal advice. Regardless of where you are on your Bitcoin journey, we'd love to hear from you. Visit on rampbitcoin.com contact to schedule a consultation with one of our private client advisors.