Transcript+
This kind of sets up what I've always thought like everyone always talks about $1,000,000 Bitcoin or multiple millions of dollars Bitcoin and that the world would look very crazy there. And they think like whether it was hyperinflation or whatever. And I don't necessarily think it looks like that. I think it looks like what is probably described here and what we see around agentic AI and just like the deep, deep fakes and craziness we see on not knowing what's real or not, like the just aperture and for what can exist and how fast things are moving. That speed will be like a backdrop to a Bitcoin world where the price is $1,000,000. Because if you think about it in a world that is very straightforward, it's very hard to wrap your head around a digital bare asset holding millions of dollars of real world value. But in the world that he's describing here, along with everything we talked on his pot, it would be complete sense that you'd also have an asset worth $1,000,000 sitting there. It all comes down to computers communicating. The information superhighway can be a confusing mix of on ramps and off ramps. Bitcoin is worthless artificial gold. Is it still rat poison? Probably rat poison squared. We need to get into the world of OK, this is actually foundational technology. What the Internet of Money does is it creates a single network which can do a microtransaction to a giga transaction. The Internet is going to be one of the major forces for reducing the role of gun. The one thing that's missing that will soon be developed is a reliable E cash. All righty, gentlemen. Welcome back to another episode of Final SETTLEMENT. Today is Monday, February 16th, 10:32 AM. Happy President's Day to those who celebrate. We've got a big show boys. A lot is lots happened over the past seven days since we since we last recorded. This is our first real ninja launch. We usually riff and 1st sack and get a little prepped. I'm so jacked. This is truly like the best time to be alive hands down. World's chaotic but it hit me right before recording. Like this is like Bitcoin AI and it's funny, they're going to, you know, converge, but it's like Bitcoin back when it's the only other feeling I've ever gotten where it's like the more you lean into it, the more you can not only control your destiny, but really be a leader in the world. When I look back at Bitcoin in 17, I was like, holy shit. Like if I actually, you know, embrace this and go deep, it's going to compound ahead of everyone else. And it's just the same time right now. It's it's crazy stuff. Yeah, it's a very interesting dynamic because I do think, you know, three of us and people in our circles very much are tuned into a lot of this stuff. And like, I would say the vast majority of like, humans, normie people are completely missing a lot of this. And so you simultaneously feel like you're always behind if you're like tracking this stuff. But the reality is, it is if you have any sort of inclination towards using these tools, learning about them, you are miles ahead of the average person who's not. This isn't even on their radar. So lots to get into. I think the natural place to start is the quasi acquisition, I guess, of open Claw by open AI. And this is interesting for a number of reasons. But up on the screen is Sam Altman's tweet about it. So Peter Steinberger, who built Open Claw originally Claude Bot and then you know, Anthropic and Claude kind of came after him to because Claude Bot was a little too close to to Claude. It was using Claude on the back end primarily. But because of that, he changed the name of it and now Open AI sort of acquired him and is claiming to remain keep open Claw open source. But there's a few different ways to look at this, because on one hand, you could look at this as a massive win for open source. This guy built this thing out in the open, fully open source. It has gotten incredible traction in terms of people using it, developers building on it. But then at the other hand, you have Open AI. Hire him, put Open Claw in a foundation claiming it'll still be open source. But as we know, open AI has a sort of a nefarious history of of claiming things are going to remain open and then they're not. Two different directions we could take this, but curious initial thoughts on this. Well, so I think taking a step back, I think a lot of people probably there's interesting overlap with AI and what's happening and then you take it at the bleeding edges of this open claw agentic payment in AI that overlap with crypto and Bitcoin Twitter. So there's naturally, I think, you know, interest or some individuals of probably most listeners have heard of something about open claw. But I think most folks to your point that don't follow any of this have no idea. But then even the majority of people hear about it, but they have no idea like what is it? And so I'll just give a quick recap. I think the the main premise to understand is that you effectively have this like arbitrary agnostic source, which is open claw open source that you're giving root access to whatever platform or device that you're installing it. And This is why you should never really install it on anything with sensitive material or on your main, you know, daily drivers. So that's where you see the the buzz around Mac minis because of the Macs in a really interesting spot too, with a unified platform with RAM in memory, which is what gives it just a lot easier to out-of-the-box have running. But Long story short, you have this open source platform that can now start and effectively integrate and integrate with pulling the AP is whatever foundation models, frontier models you're using. But then ultimately take that and go spin up whatever you effectively need, whether it's emails, whether it's integrations across other AP is. But there's a notion of you have this platform, but you also have the thing that most people don't call out is memory that is stored based on your conversations and it's learnings because these are two big pieces. I will say that if you're on the bleeding edges of using open clog, you've still found some value or can do something productive. What I'm starting to learn is for most individuals, it's even way too much that to start now, if you're even interested in any of this is really just downloading Clod and Clod Co work because it's my understanding from an enterprise perspective, they've been very deliberate and intentional with security and how they access anything on your desktop. Now you're still trusting a third party, but it's a little bit more closed end in what it can do, but also gives you about as an amount enough of functionality in your daily life and how it's has enough memory to store what you're trying to work on and make you insanely productive. It can get you to 80 to 90% of the way there without freaking yourself out and having to go download all these tutorials and mess something up because you have to really set your permissions right. More on that. But I think just from a productive perspective, if you want to play with this stuff, you can get 80% of the way there with, you know, Claude and then if you really want to move the needle now to Brian's point, super fascinating situation that played out for a number of reasons. 1 is this could we don't know the the acquisition price would be the first, you know, billion dollar, you know, company that has built something itself. Also, it's, it's a little, I mean, it's a little concerning, you know, when I think about Open AI and all the different things that have happened with Sam Altman and you know, you can go back to the must stuff and Sam Altman all the way to just, you know, people being suicided and where he's played into that or not in that firm. What's concerning is Open AI. They made this acquisition. They're more than likely going to build it into the platform in some way, whether it's via downloadable application or a plug in. And it also came out over the weekend that Kimmy 2 point O, which is the open source solution coming out of China also put there a version of you know, agentic capabilities into the browser. And obviously I would encourage to do your research to don't do that. But where it's concerning is because if open AI is effectively going to be the default, as Brian was saying, I think most people that even hear about AI are going back to it's like the AOL of AI. It's going to be like open AI. It's a different call out, but it's the realization that if you haven't used this stuff in a while, I was playing around with some on Prem open source models and because they can run via the RAM and it's crazy. What like the old stuff could do versus Claude, like it's just night and day on the level of context. So point being, is anybody there? Just I would encourage you to go check out the latest frontier models. But independent of that, most people are going to go to open AI. They're going to download this thing and Open AI is going to have root access to basically everything that you have on your computer. I was pinging a good friend the other day and telling us he's like, well, this is kind of what Microsoft already does the, you know, Microsoft and, and their operating system. So anyway, I'll pause there because that was just like initial thoughts. There's obviously a lot of other implications and where this goes with the gentic, but I do think it's a, it's a fascinating dynamic. And I, and I question was this always baked into what was going to happen? Because it's also happened insanely fast. And it would make sense to me because opening eye has this play and they're effectively going to have everyone's data because most people are not going to go on the latest models. They're not going to go and buy specialized hardware. They're going to go plug in Google off to open AI and just give access to everything. I think all that is very well said and I think that what they're actually going for too is the data is being is changed from not just your and eyes data, but also the agents data too. And so they're, I'm sure that they're going to incorporate this somehow, but ultimately they really want to, in my opinion, be on the bleeding edges of how are the agents that act autonomously going out and interacting with different interfaces. Because taking a step back, as these models get better and better and all these open source tools, there will inevitably be not just essentially B to B&B to C type businesses anymore, but there are increasingly going to be agent to agent and business to agent, which I think are probably going to be some of the largest industries in the world and and largest companies within a few years. I think that what the play really is, is trying to build on and understand how to optimize the usage of agents to go out and autonomously complete tasks on behalf of whatever employees, business people or just ending consumers on the other side. And ultimately it's it's interesting to see that all the work or not all of the workflows, but a lot of them are going to need to be optimized not just for human interactions and and businesses. But I do think that that's what a lot of the benefits of Open Claw has been, is just the ability to iterate extremely quickly on the open source tools. And ultimately, as the agents continue to go out and do commerce and, and work on behalf of people. I do think that while you and, and I and, and all of these people who haven't necessarily used these tools yet, they need like a perfect interface for people who are a little less technical. These agents will be able to just understand natively how to interact with these tools in the best ways of possible in a technical manner. And so thus, essentially they're going to be able to iterate on these open source models and understand the data a little bit better on that front. It's just a fascinating time. Yeah, One thing I would really encourage anybody listening is to go and and play with this stuff. It really, truly, I was just thinking about it. It reminds me of early Bitcoin when you know, we used to call it, I don't even remember. It's funny because we're all nerds, like, you know, didn't have families and you were maybe married or engaged. And so you could get a Saturday morning and and I forgot what it was called. But like, you know, where you just play around with hardware devices and you're wiping stuff. You're you're check, you're checking you hardware devices, you're checking you open source tools, you're messing around with lightning. And point being is a lot of that serves as a foundation to Liam's point, like a lot of people may never care about this, but if you remotely care about your data, your future technical life, sovereignty, he's not to say you have to go build a lot of this stuff, but it influenced a lot of people's first principles understanding of like risk and security. Because when you understand severing the keys and when you think about how most people don't have a large percentage of their wealth in Bitcoin, it feels risky. But when you start to like play with this download, see phrases, wipe them, back them up, understand multi state, understand config files, it embeds a level of confidence from an allocation perspective or like what you're going to do with your career, how you think about finance. In the same way that when you play with these tools, it's super clunky and you have to just generally be curious because I remember like you're setting up multi sig for the first time. It's annoying. You have all these different things that aren't working. It's the same thing here. And I get annoyed because I'm like, man, am I really being productive? But it serves as this like base layer for a conversation or context into what's coming, how to think about it, how to manage your data that I don't think most people are going to have. And then it just compounds. So that's just something to call out as far as just utilization and and where this all goes. Yeah, very good point. Just there is there's some interesting parallels between you know, basically Bitcoin and and all this AI stuff in terms of DIY side of things, fully open source, that kind of thing and then reliant on various dependencies to help you with those things. So I think we're seeing very similar dynamics in this world. But I want to go back to something you said just around the speed of all the stuff that's happening. Like it was, you know, just back in January that sort of clawed and Anthropic was just all over your timeline, like just completely swamping your timeline. Everyone's excited about it. And this, this tweet I have up here is like telling the story of of basically Anthropic fumbling the bag on Claude Bot in in a sense where back in January, again, all the hype was around Claude code. And then Anthropic's legal team sends a cease and desist over Claude Bot sounding too familiar to Claude Steinberger Berger, who's the founder of it basically changes the name. And then in February, he, you know, we have this announcement that he's joining open AI and they've sort of Aqua hired the tech to some extent, although again, claiming it'll remain open source. But this is just fascinating in my mind because again, if you're in, if you're following this stuff like it is whipsawing back and forth so quickly in terms of everyone was excited about Claude code. Now everyone's saying codecs might be better than Claude code and it's just going back and forth. And it really is this war of like not only the tech, but also the talent, right? Like this is a, this is an acquisition of talent to to some extent. Go ahead, Mike. No, no, I was just going to say, yeah, I mean, I think like take this with a grain of salt because it's the same thing with when people look at Apple and Apple looks like they're playing like 40 chess when it when you look at how they think about building hardware and where this is all going. Because to Brian's point, like here, you're reading this. And the way I read it, the way I think about it is like this isn't even true. Claude has a like upper level already enough like iterations on Claude Co work, Claude code eating on the desktop. And they're just taking a more deliberate approach to anything agentic related specifically from enterprises versus this other route. And it, it's a, it's not exactly apples, apples, but it's the notion that apple wasn't playing a playing a game here. They were playing for where Brian's alluding to is these frontier models as they go further out their data size, like from a pure normie perspective, like us, we will never be able to run a frontier model. Most even businesses won't, because when you think about the amount of rack space and power and compute needed. But what's happening is that as those Frontier models get better, the legacy call it 12 to 24 months lagging, you're able to basically, I forgot the term, but you can like start to compartmentalize where I never get that word right the first time where you can find the data and run the models. And so ultimately what it means is you can run these models on smaller amounts of RAM versus what it historically took 12 or 24 months prior. So while you have the models getting faster and better on the frontier side, the older models are getting more effectively like distilled. So you can run on consumer based hardware. And that's where you see Mac studios coming and Mac studios are starting to increase the amount of chips. And then the the RAM, the the VRAM is the main thing. So like this new, I was doing research sort of thing about what to buy and, and it's coming out this year. There's, I believe M5 Max's on the studios, but then also on the computers where you're going to get anywhere between what historically be like 8 to 16 gigs on a MacBook are going to be like now 64 to 128 to like, you know, you've been above. And then on the studios, you're going to get all the way up to almost a TB. And that's how you do right now that I think it'll be like 528 Max, but where you'll start to be able to run a lot of this. Why I'm bringing this up is because when you start thinking about modes and data, where I think we all agree a lot of this proprietary and alpha is going to exist, is based on the prompts and then really the knowledge within an organization. And as people start realizing that because you're effectively taking those prompts, you're taking that data and you're using it to train these models. And so it's going to make more and more sense to start hosting these things on Prem because we went away from this notion of on Prem where you have AWS and Microsoft Azure and Google. And as the market starts to recognize this and it's going to be really like chaotic because you have these legacy firms that all they know is outsourcing their server Rackspace. But now if you want to have a mode, if you don't want to be training somebody else's data with your proprietary information, it starts to pull away. And even from an individual perspective, as you start to think about how do you manage your life? How do you manage? Because there's things around the terms of service of these, I believe a like you're giving it all to train, but also you're, you're liable for anything in a court of law that you're putting into these models. And so it's just a fascinating time to start thinking about what you're going to be doing with your data, how you're going to protect yourself in a digital world. And then also from a sovereignty perspective, when I think about food, you know, firearm BTC, like the notion of hardware, you're going to want to run this stuff because you can see a bifurcation like we truly are going to see this bifurcation between being able to manage your data, being able to run these models, being able to do things that you want in a proprietary private way. But then also The thing is. Check out earlyriders.com for all the latest in Bitcoin investment research. Now back to the show. More truer than ever. It's like this stuff doesn't make somebody that was average a little bit above average. What it really does is take like the top performers and makes them 10 acts better because you can feel the feedback loops as you're prompting them and you're working through how the outcome should be. And you're seeing you're like merging, you're kind of like working backwards and forwards. And then you start to because like Clot as an example, all their code is being built by Clot code. But then the other example is everyone internally is using like 3 to 7 Clot instances because you're just running things in parallel because while you're doing one thing, you're knocking something else out, whether it's the report or design. So it's, it's just insanely fascinating and it's going to happen even faster than I think we all expect because whatever's out is only what they're giving us to be out. It's not what actually is already being used internal. Yeah. I, I want to maybe pull in the thread that you were describing in terms of on Prem versus trusting cloud dependencies, just using sort of like a more enterprise focused version of these tools. Maybe talk a little bit more about like the trade-offs, because you sort of alluded to it. It's like you want to be able to own your data, you want to be able to run this stuff yourself. But then this is where the parallels to like Bitcoin custody come in where it's like, well, then you are the dependency, right? Like if you're not outsourcing the dependencies, then you become the dependency and you have to be very careful about the permissions you set, the access you give these things, and then also the cost of like running it locally as well. So maybe like, do you have any more color on like how you how you're viewing those different trade-offs? Because I think you're right for like the average normally person, they're going to go the route of just, you know, using the open AI, sort of like white labeled version of open claw. But for a lot of organizations or just more technical people, they're going to opt to build this stuff on premises because they want to own the data. They don't want to have those dependencies. But again, there's there's trade-offs associated with that, right? Yeah, I haven't thought too deeply, but I think it mirrors a lot of what we do at on ramp and in the blending. Like there's a barbell, right and there's there's a hardcore barbell of somebody has their own multi sig or self custody set up. And that's for very few people. And I think there will be very few people, you know, and relative to the vast majority of people holding Bitcoin and vast majority of people using AI that will house everything right from I think it'll grow over time because of the amount of compute that can be run on a, on a consumer based device. And then you have the other side that outsources it. And you effectively are giving up all autonomy, right? Because they can not only give you your Bitcoin the next day, but you're also are at the behest of like what the data is feeding you. The joke I like to play and it sounds so crazy now, but we have to remember back in COVID, like Google will tell you a woman, a man can have a baby, right? And so it's like the models will. And I'm, I'm a product of this. Like I have a horrible sense of direction. I think at least half of it is because I've just always grown up with Google, Google Maps. So you're just kind of brain is like doesn't really frame to that. And so then I think they'll end up with this middle ground. In the middle ground will be built on trust in brands and you'll start to recognize people doing it certain way. And you'll have hardcore people that are early adopters that effectively put their stamp of approval. And there's audit code and you're either not taking that data or you're able to like right now, what is a general standard? My understanding is if you have proprietary data, you can host these local models that are a lot easier to put on consumer based hardware devices. And then what you do is extrapolate that and then use some of the frontier models for other things. And so you can get like the best of both worlds and you have to be able to know, you know what your device is doing, but you're switching and oscillating between that And so I think that'll be somewhat play out. But I think that middle ground around where we think of multi institution and commercialization, like we see this with Macbooks, kind of right, because Macbooks and iOS devices, yes, like they probably work for the government, but at the end of the day, there's a reason why they're very secure. There's a reason where they've taken stances on the level of privacy that you get in certain respects. I think like everyone's familiar with, you can like spin up emails really quickly on there for to protect your privacy and from security. So I think it starts to blend there, but it's too hard to tell right now. And like what does it look like in a future state? The last thing I'll say it and then we're going to like part of this segment will be the agentic stuff about payments because I just, I started thinking more about it and I think that'll probably play a part in it as well when it comes to how you how you're able to pay for the tokens and the usage with this. Yeah, just to that point too, Brian, if you could click on the second link out there, just because all of this, at least right now, is completely open source. There are so many iterations and on what the actual models look like today. So for example, open claw, it takes about over a GB of RAM and takes about 500 seconds to start off. You need to have like a Mac mini or something similar to that. And if you click on the picture on the right, there are other more lightweight models that are don't have quite as much functionality, but take less than a second to run, take about our 100 X cheaper in terms of RAM and can be run on like any $10 Linux board. And so that's just to say that similar to running open source models that will be a little bit slower and and there's a really interesting company out there called Open Router that just optimizes for whichever type of task. And if you should be using the most expensive frontier models out there in terms of APIs that are for heavier lift tasks that are, you know, actually very significantly like sophisticated versus just, you know, finding some news on Google and using the free open source models and optimizing for costs in that front. In that way, I think that there will be many different offerings for different models and usage of these tools and there will be easy switching costs as long as these things stay open and just optimizing the speed and direction of how and where you use for different tasks out there. Yeah. And Mike, you mentioned this like the sort of a parallel narrative that's been, you know, really accelerating over the past few weeks is the payment side of all of this. So like agents using crypto rails effectively to pay each other, pay humans, etcetera. And so there was a few headlines that we wanted to to bring up here. 1 was Coinbase rolling out an AI tool to give any agent a wallet. And then there was another announcement from Stripe, something similarly using base, which is Coinbase's sort of Ethereum L2 for AI agent payment protocols as well. Thoughts on thoughts on any of this? Because I think the, what what stuck sticks out to me is like, you know, you have a lot of Bitcoiners just assuming that, oh, the, the agents are, are going to use Bitcoin because they're going to recognize it's, it's the best money. Why would they use digital dollars when in reality, like, I think there's going to be at least an interim period here where they're also going to use stable coins. And particularly because, you know, entities like Coinbase and Stripe are building the paths for them to use these stable coins. And so people need to build similar tools on the Bitcoin side to allow these things to happen. Cali had a good tweet the other day. You know, USDC on base seems to be far more common for four O 2 payments than Bitcoin. Recently, even Stripe joined the bandwagon. That's a centralized stablecoin on a permission chain. Agents are starting to use Fiat. It's a huge loss. And in a race, many aren't even aware that it exists. So this is all to say, like, you know, we have to build in parallel with these other firms that are, you know, more on the stablecoin side. It's not just going to automatically happen that all the bots and agents are going to use Bitcoin. So this is this is going to be crazy because I was thinking more about everyone kind of knows like Bitcoin, like let's just go to Bitcoin and it's the it's the easy thing to say. It's like Bitcoin's going to power AI and AI agents and you know, it makes sense borderless, no KYC blah, blah, blah. It sounds, it sounds good and makes sense. But again, we talk about this a lot. The bookends are the easy part where you make money is, is in the middle. And so when you start to play around with these tools, you start to realize because one of the benefits just to call out is like you effectively can produce anything technical now because you can give the feedback. So all this stuff, it requires a lot of technical efforts. But if you're playing around with whatever different, not universes, but like your constructs. So whether you're using codecs or Claude code or you're messing around with Claude in prompts, you can be coding something while running the prompts in a different browser to check on any of the outputs that they want you to run in the terminal. So you're effectively getting this nice recursive feedback loop, which is is helping you understand how to prompt or how to run in terminal. But where you start to really get the visibilities like API rails, because a lot of platforms have API tokens and you have to go set up, set them up. So that's really where like still Clawbot kind of fails because you can go get Clawbot running, but then there's a bunch of different ways to stop it, whether it's captcha, which is coming about where you can have humans like solve the CAPTCHA in real time, but then you have to go get the API tokens, you have to feed it. And so when you start to do this, you can squint again. It talks about getting the getting the your hands dirty. You can start to squint and see how it's really clunky and it's really cumbersome and it's going to start to make more and more sense via what you were sharing with. There's two different things happen in real time. There's Coinbase setting up the wallets and the infrastructure for different web pages to interact with us. And then there's what Stripes doing and some of the payment mechanisms for us. So now you can start to see how streaming payments and things moving over the web will start to unlock these things for your agentic AI, you know, whatever you're using to be able to spin this up relatively faster versus being clunky and having to go username. And then, but with that, I will say like full stop, like it's all going to be on stable coins. Like anybody saying it's not going to be on stable coins. I think it's crazy. I'd make any bet for to start. And the reason why is multiple one is everyone understands the US dollar. Everyone understands like the US dollar basically, because after that if you go and look at the heavy push on stable coins, not everyone understands stable coins. But it's not hard to understand how it'd be ubiquitous that you can move dollars faster and then you're going to be able to move them online. You're probably going to have wallets in your Chrome browser. You're probably going to have your local bank offering. At point being is you can start to do micro payments with dollars, you can start to move them. And then that's what all of these platforms use. They don't or understand dollars, they don't understand Bitcoin. They're paying for their power and they're doing everything with dollars. And the other thing on top of that is they're probably going to start to embed some kind of like KYC on those stable coins. So you can effectively know who these people are because it's a crazy world and it will happen. But like where you have all this anonymous stuff just happening and that's where like AI will come. I'm sorry, BTC will come in, Lightning will come in and I'm not saying everything will be done this way to start, but you will see the vast purport of the vast majority of these platforms integrate, utilize and everyone will understand stable coins and how you need to get access to them. So you can power anything related to Agenta commerce and AIS. And then naturally though, as you go further out into the future, you can see how ABTC will just will layer into that because of its programmability, because of the need for, you know, the censorship resistant all the other different angles that will come that you'll be able to just spin up what you need via passing Satoshi's through. But I see this very similar to it's just like kind of Bitcoin when somebody buys it, most people don't go and park all their money in 12 words or anything that doesn't look like BlackRock. They generally don't park all their money in anything really to start. They just like, you know, leg in with a small percentage. So I was just thinking more about that and I think it's going to be fascinating to see it play out because I think the web inherently gets better and GDP grows on the web by just effectively being able to put dollars through different applications and letting things happen in the background for you versus like right now, as an example, if you want to run like if so, let's say you go download open Claw. You download open Claw, you have to point it out a model type of device. Don't point it at Claude and Opus 4.6 because it's the most expensive. So I did that. You run out of credits really quickly. But let's just say you do that or you do to the sonnet or whatever was before, you still have to go into Claude's dev website. You have to park whatever amount of dollars as credits to there and then you're basically getting the utilization and then you have to set it up on parameters. If you reduce the credits, then refill and all that. Like all of that requires efforts versus you being able to just streamline and have a wallet that's always like tied and then do it on a usage basis that's like in pennies of a dollar online versus having to set those credits up and maybe never use them, but you have to load it on a website. Yeah, all great points. I I did want to pull up a couple other things related to this actually. Go ahead, Liam. Yeah, I was just, I don't know exactly how this is going to work in a longer time frame, but just because it's so tip of the spear, you know, none of these agents can hold money in cold storage. So they're going to have to use hot wallets and there will be a ton of this is like kind of hackers wet dream just because they're going to be able to figure out how to manipulate a lot of these tools that people give, you know, bad prompts to and just to like send them them actual money. It seems like on the edge cases, there are some instances already of this happening and people are generally using small amounts of funds. But as more and more assets go into this economy, I do think that the amount of security is going to need to have to accelerate really significantly because there are going to be a lot of bad prompts out there. And, you know, just the ability to not hold hot or cold storage as agents or, you know, give some amount of credits at some period of time. It's going to be interesting to see how it lays out. And I think that there will be a lot of hacks out there and could likely be a negative aspect of just more stringent regulations for the digital asset industry as a whole too, as if there continue to be like instances of additional fraud and hacks online. Yeah, that that's one interesting path. The other thing that comes to mind that Michael sort of alluded to is like there is elements for these agents that would make Bitcoin a preferable payment rail just in the sense of its censorship resistant. So ultimately these stable coins can be frozen, right? Like they are not totally decentralized and permissionless. So like, I, I still think this, I agree that this is still a farther way out, but like, I think there will be a point where these agents are trying to do things that may not be totally lawful and they're just operating and doing these things. And so then they will prefer a money that is harder to seize or freeze. So that that's one path. But I, I would agree that like there's going to be an initial period where, yeah, particularly because the coin bases and stripes of the world are building these things with stables in mind, like, you know, they're, they're going to use them. And one, just one thing to call out like I think this is a big fundamental reason we're independent of the fire caster and craziness on crypto and VC and companies. Dan Romero joining Tempo and Stripe has to be a big play. Like, you know, I was on a podcast last week with Robin Sayer and they ask, I don't even remember what it came up, but it was ultimately like if you weren't focused on Bitcoin, like I don't think anybody else here would not focus on anything other than AI. And similarly, if it wasn't focused on Bitcoin and like, let's just say it was gold, you know, tokenizing gold and stable coins because it's really just the surface area for what it was previously not possible has been removed and from AI and how you can do and build things. But then ultimately you have to connect all that plumbing and then how it's being done online. And so I just think that that's also a really big just unlock and you know, you always say like, look where the talent is going. And that's a big way when you see and then that doesn't even account for cross-border and all the real world applications with money movement. This is just talking about. And again, we're not we didn't talk about Agenta commerce and everything that'll be related to how does it just search for you and understand your preferences and your sizes. Like, you know, you think about the clunkiness there with your card and the chargebacks. And so there's a lot, there's a lot of incredible stuff happening. Yeah, before we move past AI stuff, I did want to pull up a couple other links that we had on the list just to sort of contextualize the speed at which this stuff is happening and sort of like the rate of adoption. So this was Spotify reported that its best developers haven't written a line of code since since December. So you know, they're basically revealing that they are using Claude to write a ton of their code. And so you know, things like enabling real time bug, bug fixes and, and feature deployments straight from phones, like, you know, whether it's WhatsApp or Telegram, being able to just do this stuff super efficiently. And so even the best engineers are are sort of actually writing less and less code. The other one which I found super interesting was this chart which shows developers are building for open claw faster than any other OS in history. And so maybe you wouldn't initially sort of look at open claw as a as an OS, but it I mean, it fundamentally is and and people are building on it at a much faster clip than sort of any other OS in history. So I thought this was this was particularly fascinating. Any other thoughts on AI guys, before we go to some some other links, we had some announcements from sort of Chad Phi world. The only thing to call out there's really 2 concepts. 1 is highly encourage people play with this stuff for the very sake of so you don't lose your job because you get really hard to get axed. If you're utilizing these tools and you become the person people go to, It's going to still be super clunky at an organization to start to embed these things on the other side of it. If you're building in this area, we'd love to speak with you because this really is the underpinning of everything that Early Riders was meant to be. The early adopters utilizing these tools, we'll be able with $100,000, make very large companies. And where you would need that $100,000 in this world is literally around tokens and maybe some hardware, whether it's on your computer being able to run this stuff or if you wanted to host it from a proprietary perspective, that there's not a lot of capital needed. The understanding is there's a lot of like $10 million AR companies out there where people don't talk about them because there's not really defensible moats. So they're just making a bunch of money. And I would just highly encourage if you're building on any of this stuff and also if you want to work with us, because as we think about hiring from an operator perspective, like there's no shortage of people that want to work in Bitcoin. We get no shortage of interest on early writers and on ramp. The best thing to do is show us what you're working on here because that speaks like 100 X on how you're thinking about the world because even about on our team. You know, I went full like 40 hours this past weekend working on it and it's like I had to do it because I'm just naturally curious. But also, if you don't know as a leader the capabilities of this stuff, how can you expect your team to start utilizing it? And so it's the other side. If I'm utilizing all this and I'm as busy as I am and nobody else is, well, then we have a problem on the organizational side and we'll figure that out. But so I just think that there's never been a better time if you're going to be curious and play with this stuff. And it really is asymmetric because to the upside you can benefit, but then to the downside, if somebody has to pick on a chopping block, if your organization has to get rid of headcount, well, why would they get rid of the person that's producing at 10 to 100 X what everyone else is doing? Yeah, just from the early riders perspective, we're really starting to see the do more with less thesis accelerate Open claw built in 82 days before selling to Open AI. Likely people are speculating for at least $100 million. And it's also just the ability to work at lightning speeds too, you know, building that in such a short period of time and having it catch on to the amount that it did is really just fascinating to see. And you know, we're going to see everything really start to accelerate significantly faster than it has in the past too. So just really fascinating time to be alive and use those tools too as as a builder operator and then be able to benefit from it from investing in early stage companies that are going to be able to move and a lot quicker and less bureaucratic ways than these large organisms that are going to come with top down approaches to AI rather than bottoms up from people who are actually using the tools. When it comes to holding Bitcoin securely, Peace of Mind starts with architecture. Onramp's Multi Institution Custody distributes control across three independent regulated key holders in a two of three quorum. No single point of failure. No pooled or omnibus exposure. Segregated client titled faults. You retain full legal ownership while on Ramp coordinates security, compliance and operational workflows behind the scenes. It's strength of money delivered through the simplicity of 1. Multi institution custody is the foundation for everything. We build sound infrastructure that distributes counterparty risk and provides fault tolerant resilience with clear audits and institutional controls. And now on Ramp is piloting flat, predictable pricing, making best in class Bitcoin custody and financial services more accessible now than ever on ramp strength in many simplicity of 1. To learn more, check out on rampbitcoin.com A. 100% totally agree. Switching gears here a little bit, a recurring theme on the show for the past several months to over a year has been Stratify banks, incumbents putting their chips on the table, getting getting things situated to work in the digital asset space, launching stablecoins. Couple weeks ago, we had Fidelity launch their stablecoin FID and then we have this news from over the weekend that Charles Schwab is staffing up to build its own stablecoin as well. So, so you already confirmed on the summer, call it Schwab, Schwab will be launching A stablecoin that's alongside trading for Bitcoin and Ethereum. And so this is just sort of another incumbent that is getting their arms around this stuff and realizing that not only do they have to participate in the sense of allowing their underlying, underlying clients to, you know, have access or exposure of these things, but they actually need to build it themselves to some extent, whether that's a stable coin or custody rails. So I thought this was relevant. And then the other one similarly, we had Wells Fargo posting a head of digital assets role four days ago. The mandate as described in the role is to build a three to five year strategy across tokenized deposits on chain collateral, intraday liquidity, 24/7 programmable payments, fully integrated with wires, ACH Fed, now Swift. And this as it notes here comes right after Morgan Stanley did something very similar and JP Morgan made these crypto leadership hires. The third one that I had on the list was BlackRock doing something as well. I think Michael, you showed that one. Any thoughts on either of those two while I pull up the The Black Rock link? No, I, I just think it's, it's similar as AI, like I think we talked about it in the last year, One of the predictions is digital assets will take the same way as AI did last year in the sense of everyone needs a strategy and that's crazy, but it, it just makes sense. Wells Fargo, who's historically been, you know, insanely kind of against this asset and just legacy. We talk about it a lot when you look at your legacy banking rails and just the user experience and clunkiness. So I think those are huge. And then I tap this mainly because I think we're just going to naturally see more of this convergence. There's a lot here. The one that I joke is like they it's just for Brian is that they did this defy offering for black rocks. It's a fun, they offer some kind of interest, but it was like they built it for trading on Uniswap and they bought the Uniswap token. And I just thought that was funny because it's kind of what we've been talking about. It's like these things don't have value. They're not needed. But like these people don't know. We just saw, you know, Harvard, who we think knows things, bought Ethereum. Like this stuff's just going to propagate for a while because people just don't really understand, you know, that one thing could just service everything. And honestly, it kind of can't right now because the people that understand like Uniswap and how to do that stuff, they don't understand how to do this on on BTC. Like there's just this interesting angle where people that come from, I remember being in New York at the time, like the Columbia's of the world, they like got the shiny object of crypto and they built all these primitives around D Phi that will probably live in Bitcoin, but they went to like crypto and these other things versus building on the sound foundation. And so I think we'll just see more of this. But the the net is positive, whether it's digital assets, crypto or stablecoins and AI or just cross-border. All these things are bullish for Bitcoin because it opens up the aperture for people to realize that digital assets are here here to stay. And then people will see the ruggings, whether it's via inflation of the supply, inflation of the monetary, you know, nominal value. So your, your real returns are holding, your purchasing power of holding stable coins is going down as compared to Bitcoin. And then it's just one quicker step to go from that into BTC 100. Percent I I think there's I've certainly come around to the notion that you know the the sort of crypto journey is no different for an institution or a track by firm than it is for any individual over the past ten years in the sense that you're going to try out these other things you're going to think they have value, you're perhaps even going to build on them. But there is that learning curve of recognizing what is the solid foundation to build around. And so this is obviously still very early stages as as a lot of these firms get involved with this stuff, they're going to have similar learnings. And whether that takes months or years, not exactly sure, but there's going to be that similar curve. And, and I think it could be a feature, not a bug when you think about it, because if you, if you think about like you do need, you do want a better Internet, you want to like whatever more open Internet, more GDP flowing through it. And so while Bitcoin is better than stable coins, the reality is most people are going to build on stable coins via these primitives. And so we all benefit from that in the same way that if somebody wants tokenized securities or they want security access, whether it's private or Publix, and they want to be able to leverage them or do other things like these primitives that they're building on are going to do it. They're just not decentralized. But I think that's a feature for them. It's not a bug. Meaning like the same way Tempos is creating the solution, because if somebody hacks your stable coin wallet built into it, they can just burn those stable coins like an FDIC and just revert it back. Most people want that in the same way that if something nefarious happens on Uniswap, like there's just an admin key. And so I think like once you realize that, you're like, oh, well, this is how it was always intended. It was just the crypto like kind of, you know, the the the Etheds or whatever didn't really understand that they thought they're building decentralized infrastructure. They're not. They're just building the the other the same thing, but like with different goals or different whatever it is insight. And so I think like it's just plain out the way it was and and crypto was like a test net for Tranfine what they're going to be able to do here and take over. Agreed and a lot of that is still a tough net for what's going to happen on Bitcoin and those products and applications as well. And the other thing just to mention too is while there are a lot of folks in the Bitcoin space who have at least started out with some DSS around altcoins or stable coins too, they've been a consensus of 1 and so they've been able to easily dismiss that one. It didn't necessarily go as planned, but when you have a large organization with a bunch of different committees and etcetera, it's going to be more difficult to give that up. Even when you see it not necessarily working out and playing as a thesis intended. There will definitely be some net benefits, but there are also going to be a lot of just warranty that's, you know, wasted on things that don't necessarily come to fruition as well. Yeah. All right, Maybe before we wrap, I did want to just briefly touch on something that went super viral over the past few days, which was a Ray Dalio X article. It seems like, you know, X is having a good amount of success pushing the articles because every every week it seems there's something that goes mega viral where there's, you know, your timeline is flooded with people quote, tweeting the article. This was one of those where, you know, he's really just talking about his thesis that he's had for the past, call it 10 to 15 years around the shifting sort of world order monetary monetary system, sort of long term debt cycles. I think I guess the new thing that he's saying here is this was always a six stage cycle as he described it. And he's saying basically we're we're now in stage 6. So for the past few years he was saying we were in stage 5. Now he's saying we're in stage 6. And I mean, the take away from me, like if you look through the comments or a lot of the, the quote tweets on this from sort of Bitcoin Twitter space space is like, you know, he gets it. He gets what's happening here, but he doesn't see Bitcoin as a solution, which I think has been the the case for a while. It's not that he doesn't understand Bitcoin at all. He's had some exposure over the past several years, but you know, not nearly as big as his gold position. And so I think there's still a lack of understanding there in terms of Bitcoin being the the solution or at least an alternative in this shifting monetary order, the shifting world order. Curious if you guys had any thoughts on this, if you even read it or if you if you just think it's it's noise at this point. He's old man yelling at Cloud about something that he has been yelling at for 10 years. I mean, it's similar to what he's written in the past. So I don't think that the thesis is any different. I think like the CBO recently put out new estimates on where they expected debt and deficit to go in the longer term as well. And it's just more of the same. I mean, the thesis is what people have been saying for the past 10 plus years. It is definitely accelerating, but it's accelerating. It's like a a very slow moving train. And you know, people at least right now on the government side of things and a lot more of the optimists are at least seeing AI as a potential way to get out of this problem just due to the productivity increases. But the thesis is still exactly the same. I don't necessarily think that anything is new from this one too. It's just him reiterating it in a more cohesive manner too. Yeah, I think I'm glad you recap that, Brian, because I kept seeing it and I was hoping I'd have to read it. We had, you know, we Nick said he he listens to the pods, so hopefully he hears this. He had shared this article over the weekend in our Slack and and all I got was like the link and it says it's official. The world order has broken from Ray and I my my comment was laugh out loud. Thanks for the heads up Ray. Because like to the point that we talked about it's like what kind of anybody paying attention understands that there's just natural friction happening. I think maybe to tie this all to the beginning of the pod is that this kind of sets up what I've always thought. Like everyone always talks about $1,000,000 Bitcoin or multiple millions of dollars Bitcoin and that the world would look very crazy there. And they think like whether it was hyperinflation or whatever. And I don't necessarily think it looks like that. I think it looks like what is probably described here and what we see around agentic AI and just like the the deep, deep fakes and craziness we see on not knowing what's real or not. Like the just aperture and for what can exist and what how fast things are moving. That speed will be like a backdrop to a Bitcoin world where the price is $1,000,000. Because if you think about it in a world that is very straightforward, it's very hard to wrap your head around a digital bearer asset holding millions of dollars of real world value. But in the world that he's describing here, along with everything we talked on this pot, it would make complete sense that you'd also have an asset worth $1,000,000 sitting there. And then the last thing I'll say is when we think about, because he probably didn't, because a lot of stuff is for his investors, he probably didn't outline the social strife we see in different sovereigns when it comes to privacy, autonomy and all the different things that we're seeing across like Netherlands. And you know, they realize income and what's happening in Iran across across the world, and that this is the value in borderless central censorship resistant money is that people that are hosting models that are deemed to be, you know, a threat or risk to the local sovereignty can do that and accept money that can't be censored or seized. Because you can imagine how stable coins and other things will naturally seize or they'll get the addresses, they'll have IP blockers, they'll have all these things that won't let those payments go through. But that form of money will naturally end up in it in that way. And it'll make a more robust and freedom oriented technology, kind of like area of the market that most people will realize. And eventually it's just like, I just use that because it's a better tool. And you'll have swab in between stable coins in dollars and they'll look like stable coins, but stable coins and Bitcoin, it'll, it'll move again. I don't know what it'll look like, but you can start to understand, well, if this is all true, then you can't just recreate the existing stack, which is what's happening now. You'll start to have an asset appreciate, but then it'll also start to be more ubiquitous across all things. And so, you know, it's always fun to throw out the $1,000,000, but like I really think that's how you end up seeing a fast approaches to the crazy numbers that you see in Bitcoin is once this gets realized in the same way that Harvard, I think when what we talked about, we're probably still three to six months behind, you're going to start to see a lot of hardware fly off shelves, especially with there was a really good article around the RAM. I forget who is the manufacturer, but it's ultimately like all the stuff that goes into iPhones and computers from a memory perspective is super like short in the supply chain. And I think you're going to start to see that search a break. And that's when people are going to start hoarding a lot of this stuff. And when you see that happened, you're just going to have like different spikes and different things. And obviously like production will grow with that and then they'll start to manufacture. But I think you're just going to start to see a lot of weird stuff play out across all sectors. Yeah, just scrolling back up here to, to this section here, I think speaks to sort of what you were describing where like, you know, he's, he's outlining very clearly here why you need neutral permissionless assets in this in this world where it's sort of multipolar deglobalization. No one trusts each other. You're not going to use U.S. Treasuries as a reserve. You can't at this point. And so that's why I, I think, you know, he's sort of explaining why gold's gone on its run over the past year because there is an increased need for neutral reserve assets that can't be frozen or, or harder to seize or freeze than let's say U.S. Treasuries. But he's describing why a lot of the reason why Bitcoin exists in this world and has a massive value prop in this world. And so while he doesn't mention it here, a lot of people read this as sort of the writing on the wall for, as you mentioned, the value prop for a a massive repricing of Bitcoin as we sort of hurdle into this new world. Yeah, maybe I'll pull up one thing because it's kind of, I don't expect to, but this we, this we, this came out last week. But also it kind of ties into this new world we're building for and going into. So we had Nick Delozer, who's our COO and CCO and on ramp on last trade and, and he put out a great piece highlighting a lot of the technical, operational, legal work related to building multi institution and how somebody can onboard in minutes. But one of the other things that was released here was that you can actually download a wallet configuration file. So you, if you're familiar with like third party, like collaborative custody providers, there's the nunchucks cost as they, as you effectively need that wallet configuration file from a collaborative custody provider because that ultimately is the math to how do you reconstruct the wallet. So you know, by leveraging somebody else's stack. It's helpful because you get the platform, the wallet coordination to make it easy, but God forbid if anybody ever went away, you need to be able to reconstruct that because just having your hardware devices isn't enough to root back into. How do you build a wallet and then spend the BTC? And so Long story short, there's open source tools like Sparrow Caravan, there's others where you can upload that, but we actually just shipped where now you can take out the wallet configuration file from your multi institution wallet and then go and verify that those are all segregated on chain. You can even go, you know, legally to the other institutions if we weren't around to be able to move the assets. And I think this kind of ties into a lot of what we're talking about because this is kind of like the end state. As you start to think about a world where you're rebuilding trust and systems, you're going to naturally have people you trust because they can aggravate and create better experiences. But they're also kind of similar to what we talked about with Claude and Claude code and Co work on your on your desktop. How they kind of shield off and make sure you approve permission before they access things in the same way here, you'll start to coalesce and brands will start to be created where they're building things, they're creating delightful consumer experiences, but at the end of the day, they have like really robust backgrounds and security that you ultimately don't have to be reliant on that. That has never existed before. And so anyway, I just think we're going to see more of this. We'll continue to build this stuff for people that are going to be holding large amounts of money in BTC, but naturally know as the world gets more chaotic, they may not want 100% of it on their person or on a third party custodian, which I think we're all pretty bullish on. As Bitcoin appreciates, more and more people will come to this realization. And so that really underpins everything we build and then also the things we invest in. All right, gents. Good RIP. Good tight RIP. Good stuff, guys. See you guys next week, guys. Thanks. 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.
Transcript source: fountain