Self-Custody for You and Your AI

Well, I guess the question is like, why Bitcoin, right? Yeah. I think if- Is that freedom money? Yeah. The rest of it's just pretending. Yeah, I mean, with other currencies, you can get censored, right? You can get your funds taken from you. They can be inflated away over time. So if agents actually want a currency backed by energy that doesn't deflate, well, I mean, we're in the middle of the bear market. So it's very much a long-term vision. It's emotional deflation, but not monetary deflation. So I think, of course, all of us as Bitcoiners, we're pushing as hard as we can to get agents to use Bitcoin because we think it's the best. But in reality, agents are using everything under the sun, right? Roland, how are you, sir? Welcome. I'm great. Thanks for having me on the show, Sean. absolutely i know it is well what time is it you're where you are uh 11 p.m oh man i hear

cricket yeah i hear the crickets and i know it's i know it's late on the other side of the world so i really appreciate it although i imagine you're uh you're accustomed to late nights uh in your line of work well hey so we will uh i will uh provide you know backdrop of course on all b I'm a big fan. I run Hub on, as I said, my trusty Umbral here in the corner and have, you know, I think I sent you some links. I've wired up Nostra Wallet Connect payments from my website over an edge function to my Hub. I mean, you know, I'm in deep. I love the stuff. Not everybody else will be listening or watching, so we'll go through some of the basics. But Let's sort of start here. I think it's one of the most interesting points in the arc so far of Albee as a company and as a team is the decision in January of last year to shut down your custodial wallet and gracefully, graciously offer users the option to move over to self-custody or frankly, you know, go find another product.

my words, not yours. This is a very specific sort of high stakes story related to trust, which is the theme, of course, of this show, which is a company choosing principle over convenience. And so, you know, walk me through rolling that decision to kill off the custodial wallet last year. You had users who depended upon it. You know, I certainly saw the buzz and the noise on Noster. What was it that ultimately brought the team to say, hey, we'd rather face the wrath of users and customers than hold their keys. Yeah, so this actually started way before January of last year. We launched LB Hub about two years ago now, I think. And the reason why we did it, well, it's actually quite a positive thing. We kind of got a bit too big. At one point, I think we were processing over a million transactions a month. we had a lot of people running the custodial wallet

not just individuals but businesses as well and some of the businesses were possibly doing some questionable stuff might not be completely aligned with our vision but actually if we look back even further towards Albi where it really began with the Albi extension the first thing the LB extension could do was connect to your own LND Lightning node it was a self-custodial product right from the beginning but the reason why we added this custodial solution is because at the time it was very difficult for the average person to have a self-custodial node so we had this thing called LND Hub and the idea was actually even then it was an open protocol we were hoping that more people would spin up these L&D hub things, which basically mean you can have one L&D node shared by many users.

The Uncle Jim model? Yeah, yeah, which we've actually kind of come back to a little bit with LV Hub, but in a slightly different format. But yeah, so that didn't really happen as intended, or maybe self-custodial lightning took a little bit longer than we thought, right? So, yeah, the custodial wallet became really popular. And then, you know, in the U.S., Phoenix going out of the U.S., wallet of Satoshi shutting down, we also kind of had to make a decision, right? And we had two possible paths. one was to introduce KYC and go the fully regulated route or we could go back to self-custody and push self-sovereignty which we chatted in the team

and that's what we want to be, right? We want to be self-sovereign Bitcoiners. We want to build open source software on open protocols um anyone can run it so yeah and and actually the thing that made all this possible um at that time was also that nostril kind of came out nostril was very very new at the time um and they had just introduced zaps so um basically you can you can uh request a lightning invoice from from another user in Nostra and then and then and pay it with your lightning wallet um and one of the one of the things that was not so user friendly at the time is that if if you want to zap someone you would get this you would request an invoice from someone and then you would have to open up like a separate lightning wallet um or maybe you don't even have a lightning wallet um and this was kind of the

origination of this thing called Nostra Wallet Connect, which is basically you could stay inside your Nostra client, push the zap button, and it would make the zap payment without you having to go to a different wallet. Instead, your Nostra client would remotely connect to your Lightning wallet. And so this simple... And I think it's worth noting, you know, we fast forward and you tap one button and I send you 21 sats because you posted something cool, X and the rest of them are still wrestling with this. And I think for those who maybe aren't so deep in it, it's really worth calling out that this is what, Roland, two years ago, I guess, that the zap spec sort of, you know, started to take shape. And so again, here we are two years ago, two years rather later on Noster, you know, zapping each other like crazy, sending actual money, sending real value, literally at the tap of a thumb. And the rest of the world is, you know, I don't know,

issuing Stripe invoices. But anyway, so, you know, kudos is what I'm trying to say, but please go ahead. Thanks. So what we realized is Notchewallet Connect can be way more than just apps. Actually, it's like, it's basically just allowing an application to seamlessly have some sort of access, like budgeted access to a Lightning wallet or a Lightning nodes. Actually, we can apply the same idea of Nostra Wallet Connect to create a self-custodial wallet. And because one of the awesome things about Albion that's kind of different to like Phoenix Wallet of Satoshi and a lot of the other wallets, right, is we focus on this idea that you can bring your wallet to all these different applications like podcasting. you can stream saps to podcasters right um that's where mine go right now for anyone who ends up seeing this and and streaming that's that they'll they'll land in the corner over here yeah so so basically this simple experiment of doing zaps and and like doing something fun

actually we started this open protocol and and now today there are so many different wallets and applications built on Nostra Wallet Connect. And it's actually what made our B-Hub possible. And it made our transition to a self-custodial wallet possible without having to sacrifice those, those sort of features. Yeah. I mean, I think of, I think of Nostra Wallet Connect is like the USB-C of Lightning payments. You know, you just plug, plug virtually anything in and you're good to go. Yeah. Yeah. So, but yeah, So we still had this shutdown. We are actually trying to onboard people onto AlbiHub for the full year. And AlbiHub, compared to the other Lightning nodes, we really kind of tried to focus to make it easier for a normal user. So even though it's self-sovereign, we tried to address some of the pain points.

like you must have a channel right to be to be self-sovereign to have your own lightning node connect to the lightning network but who do you have a channel with um so we so we partners partnered with different lsps so you just with one click you can you can buy a channel and and those sort of things um there's still costs involved right there are a lot of lb users who actually were not economic Albi users. So they had a few hundred sets and there were a lot of people who complained, right? Because they did not actually want to pay to use Albi. They did not want to have to pay to open a channel. Well, in fact, I was going to ask, let me, sorry, Roland, just to kind of zoom in a bit, I guess. So when you made that announcement, which again, I think, correct me, was January of last year, how do users react what did the support inbox look like yeah it was it was challenging i mean we for six months or even a year beforehand we're trying to offboard our users right and but there

are a lot of users who just did not do anything so at some point we just have to say okay now from now on you cannot use your custodial wallet if if you actually want you can you can basically get your funds out following a certain process but um yeah so yeah inertia is a thing yeah for about a year afterwards we still had people emailing and saying uh i've been i've been in i've been in prison for the last two years making out these funny stories uh i need i need the money to feed my cat you know um okay credit props for props for props for creativity i guess yeah so anyway we so we've still allowed them to like um yeah i mean we still have the funds right so we we let them basically connect the wallet through nostril connect um withdraw the funds and then shut down the account basically yeah um but well so most companies you know i think and this is

what's interesting to me and you've already touched on this but but i want to come back to it you know most companies in bitcoin talk about self-custody but they keep offering the easy custodial option why didn't you just choose to offer both i i think there's a few reasons um one it like we're we're a quite small team we're a startup right we have limited resources and we we have to we have to make a clear decision on the path we want to go down there's one another one is the regulatory thing, right? We have to get all these licenses. We have to work with people to run this custodial product. It's not actually that easy if you want to do it legally. It is, let's be clear, it's criminally expensive. As you and I talked about, I've been in that business. It is criminally expensive. Yeah. And the last one, just as I said, I guess it's related to the first two points, but um yeah we at least i'll be uh i mean all of us we have slightly different views but

um we we all want to be self-sovereign we we hold our own bitcoin right we we hold our own keys that's that's what we believe in um we built we build this software for ourselves we use it and ideally our users are the same, right? They value these things. That's what we owe, at least. Well, and yes, absolutely. And I think, again, for those who may be kind of on the edge of this, Lightning TLDR is the fast, cheap micropayment layer on top of Bitcoin. Roland, so for someone who's never run a Lightning node, what does AlbiHub actually give them that a custodial wallet does, in concrete terms. And one thing particularly is what can't happen to their money anymore? So it gives them their own key, right? So just like if you have a hardware wallet compared to putting your Bitcoin on an exchange with LB Hub,

you have your own key and your own channels, which solves two things, right? So one, you can't get rug pulled like from a custodial service. So if a custodial service shuts down and they have the key, it means you both can't spend your money and you also can't get your money back. With Lightning, it's a bit different. So you can have one or more pairs, right? You can have channels with different nodes. And still the same thing could happen. A counterparty could disappear one day. They could shut down. but you have a lightning channel that's backed by Bitcoin and you can unilaterally exit and get the funds on your side of the channel back which is a bit technical how it works but basically means you can get your money back and the other thing and I think this is very unique to Albi compared to some of the other lightning wallets even self self lightning wallets is that um with albi you can open channels to anyone so there there no way that there

there's no single point of failure there's there's no single path where you can make a payment you have full control um of opening channels to different nodes and making payments through those notes. Um, so I, I think, yeah, I like to think of it, you know, again, for those maybe not so technically deep in it, um, it's the ability to, you know, maybe a poor metaphor, but it's the ability to sort of dig your own Creek, dig your own river, get back to the, the mother source, you know, it's the ocean, it's the sea, that's where all the payments flow, but you need a path, You need a channel, right? Literally a channel. And I think that metaphor comes from the flow of water. And so, you know, the point I think you're making, which is worth underscoring, is there is responsibility and some technical capability that comes with that, but you've got complete flexibility. Yeah, I mean, and I think this is going to get a lot easier, too.

I'm pretty excited with agents. We can talk about that later. but I think all the technical challenges kind of go away and not just that, but if we have this idea where everyone has a personal agent on their own machine, that also covers the hosting part, right? Because this was one of the big challenges for normal users is if you want to be self-sovereign, you need to run your node basically on a dedicated machine that you control, you have the keys, right? so with with these new personal agents that incentivize people to run a completely their own machine they can also run lb hub they can run the lightning node um on it and then they have you know both their personal agent and they can have their their lightning or running their own keys so i'm i'm pretty excited about that and you know quick quick shout out to last week's episode with Jesse Posner of Vora, which you probably know they're working on just that same thing, right?

To be able to custody, self-custody your money, your agent, and everything that sort of flows from that. But well, and so further down that path in terms of the choices that you and the team made with I'll Be Hub, Nostra Wallet Connect uses Nostra Relays to send payment instructions. Here, too, why build on this emerging protocol instead of your own infrastructure? What sort of went in in terms of the ethos that you guys clearly work on top of? What led to that instead of, you know, building out something new by Albi? Well, this is something brand new, right? This is what we built out. We had to work a lot to contribute. Sorry, I mean Nostr. Nostra versus some, you know, brand new sort of peer-to-peer protocol. So why Nostra is the question. I think it's important we work on open protocols, but we also have to work on ones that have gained adoption, right?

It's very difficult to kick off and get other companies, other people on board, right? And Nostra so far has been the most successful. and I guess one of the only protocols which is an alternative to the Twitter, the Facebook, Instagram, whatever it is. So I think that's the main reason and also because the architecture of Nostra is actually a perfect fit. I think Nostra Wallet Connect and for some of the users you don't even have to use Nostra. It's like email, right? SMTP or whatever. It's just the underlying protocol. It's invisible unless you want it to be visible. Yeah, but it really solves two main innovations

which are really cool. So one is kind of like an alternative to Tor, right? that's like the old way people would expose their lightning node to the world they're basically host host it and but it has to go through the tor network which is slow and unreliable so while connect kind of completely inverts that rather than self-hosting opening up a port and making your node accessible through the tor network it's like you connect to a nostra relay you have a persistent connection and also the application has a persistent connection to this relay and then end-to-end encrypted traffic is is going across this relay so that's one really cool thing the other one is for the first time we have kind of an abstraction layer above different lightning wallets and different lightning nodes because their apis or how you talk with the Lightning node, they're all different. Whereas Nostra Wallet Connect is this really simple

protocol that sits on top, kind of abstracts that all the way. So you just say pay invoice, make invoice, all these simple commands. It doesn't matter if you use a custodial wallet, it doesn't matter if you use LND or LDK or Core Lightning, whatever it is. From the people building on Nostra Wallet Connect, all they need to know is Nostra Wallet Connect. So I think this is a huge innovation. Absolutely. I mean, you know, for my personal example, I set up the ability to zap some of my writing on my website. And I used Claude code, pointed it at a couple of repos, and it built an edge function running on Netlify that is a proxy for Nostra Wallet Connect. and, you know, captures what I think is a really smooth UI on my site and talks to my LB Hub running, you know, as I've said three times now,

in the corner of my office here. And that is to say, that's not for everybody, but it is increasingly accessible, right? I'm not a professional developer, and so, clearly. And so it is increasingly accessible. Goes to your point about agents and where that's all going. Let's look at this, Roland. And so as far as it has come and as objective as you can be, as I certainly think about this a lot, what's the actual UX gap between the Albi app and the Albi hub and, say, cache app? So, you know, what's your sort of assessment of the UX gap for a normal user? And what are still the hardest pieces to advance? Like, so what's it look like in terms of UX and what are the mountains still to climb to make it even better? Yeah, unfortunately, I've never tried cash app. I'm not a U.S. citizen, but I guess.

Well, there we go. And that makes a point, doesn't it? Yep. You could try it if you wanted to. I guess it's a simple mobile app right in custodial wallet a bit. Well, is there, I know you're a New Zealander by birth. Is there an equivalent that runs prominently in South Asia? Okay, so that's why you built this thing. You don't have any options. No, but, you know, okay, so it's interesting. Like, I should have caught myself, you know, a typical American, so I assume everybody's got access to all these American apps. But I accidentally, I think, have landed on the ultimate point here, right, which is that I'll be didn't give a damn where you live and what passport you hold. And that may be the greatest UX advancement ever. So I'll answer the question for you maybe and say, you know, I wish I could go more places and spend Bitcoin. And that's advancing. And shout out to BTC Map and all the folks working on that.

But the fact that I can, you know, pull up my phone, pull up this really smooth. And I mean, I'm just not blowing sunshine. Like you guys have done such a brilliant job on the Albee app, iOS and Android, that I would argue it is so job-to-be-done focused, you know, in a great way that all the noise. I mean, Cash App's great. They've got, you know, thousands of employees and they've honed a fantastic app over the years. But they have to jump through all these hoops. They have to – ultimately, they're a regulated entity. So I guess this is me saying that I think if you are operating on a Bitcoin standard, which again, more of us wish we could in more of our daily lives, it's hard to beat, right? I am moving money from my self-custodied wallet on hardware that I run to the merchant or whoever I'm paying, and I understand how it operates end-to-end.

It's phenomenal. So that's a lot of words to say. Great job, Albie. And even though, you know, like I said, you don't have a comparator with Cash App. Let me ask this. Where do you, for your own internal sort of benchmarks and aspirations that you guys have as a team, what are those mountains to climb to make the UX even smoother, better, easier? Yeah, so there's some obvious things. Maybe we can compare it a bit. So LBGO that you mentioned, right, it's perfect. Go, excuse me, thank you. But you have to connect to an external Lightning node through NWC. So LBGO is a simple mobile app, really simple interface. But it's basically just an interface to your node. It's just a remote control. and that's probably the difference between something like Cash App where you can just install the app and get started straight away, right? Fair. I guess you have to do like KYC or something like that.

Oh, yeah, right. Yes. Maybe getting AlbiHub up and running is a lot less painful than KYC for a lot of people. Yeah, so I mentioned it before, but the main points we have right now is one, you have to self-host, and we have a cloud hosting solution, which is super easy like deploy with one click but it comes with a monthly fee. As it should. Yeah, but either way that's one barrier and the other one is opening your first channel which so we have different options like if you buy a yearly hosted plan we'll open your first channel for free but if you self-host normally you have to pay to open the channel So there's different ways. One, our recommended one is you just buy a channel from one of our LSPs. It doesn't cost much, and you don't have to put up the liquidity in front, right? Because the old way of opening channels is like

you have to deposit half a million sats or a million sats into your Lightning node and then open an outbound channel. So, I mean, you can still do that, and it's cheaper if you have the funds available. Sure, and the skills. Yeah, but for the average person, it's just easier and faster to buy a channel. So one of the things we are researching right now, which is quite cool, LDK node, which powers LB Hub, recently added support for splicing, which Phoenix also does. so basically you can you can increase the capacity capacity on an existing channel rather than having to open multiple channels which is cool and that the other one is just in time channels so rather than right now in lb hub you pay and open a channel up front and you and you kind of estimate how much liquidity you might need to receive

or send payments. But with just-in-time channels, especially for a merchant, I think it's quite interesting because you can spin up an LB hub node and you don't have to pay anything. But as soon as someone tries to pay you, the first payment that comes in will open a channel and just a part of the first payment will be deducted for the channel opening fees. Sure. Well, and I would just say my take on this again for those who maybe are not in the weeds and I'm sure others will take issue with my metaphor here, but I liken it to a zero interest rate line of credit that you can grow and shrink as you will, right? To facilitate payments. And where do you get that deal, right? So, and I'll again, just say kudos because I think this sort of graduated experience from, I just want to pay, just take my money, host it, do all that, you know, to I've got full control is something that's really unique about Albie. Yeah.

Well, so let's do this. You've solved self-custody for humans, largely. And I know it's, you know, I say that somewhat tongue in cheek because I'm sure you guys have lots of goals to make it even better, as you've talked about. And now you're building the same thing for machines, for agents. And so you have built a system whereby an agent, an AI agent, can connect to a Lightning wallet and make payments on its own. Why does it matter whether that agent uses Bitcoin, say, versus one of dozens of crypto projects, Stripe and the rest who are supporting X402? You know, why should our agents that we are hopefully, as you said, self-hosting or self-custodying, why should they have freedom money too? Well, I guess the question is like, why Bitcoin, right? Yeah. I think if- Because that's freedom money. Yeah. The rest of it's just pretending. Yeah. I mean, with other currencies, you can get censored, right?

you can get your funds taken from you. They can be inflated away over time So if agents actually want a currency backed by energy that doesn deflate well I mean we in the middle of the bear market So it very much a long term vision It's emotional deflation, but not monetary deflation. so i i think um of course all of us as bitcoiners um and we're pushing as hard as we can to get agents to use bitcoin because we think it's it's the best but in reality agents are using everything under the sun right um i think there's there's going to be a lot of competition and and eventually there's going to there's going to be a clear winner but uh right now there's not so um and i And it's also just a fact that a lot of these, you know, dyno, decentralizing name-only currencies can actually, in the short term, be more faster, cheaper, everything,

than Bitcoin because it's just fake, right? So I think one thing is time is going to take time for this to play out. But also, I think that, you know, the owners of the agents actually have to say, I want you to use Bitcoin. And I don't want you to use shit coins or stable coins or whatever it is. Right. Sol.md. Sol.md. Well, let me, let me, so let's, let's pull on this thread a bit more. And you and I talked prior, Matt Corallo at Spiral, Bitcoin core developer, Matt Corallo, just published a piece today called Open Source AI Needs to Get Serious. And the quote that I pulled is open source agent developers are still stuck in a why can't I just give my agent my credit card number and call it a day world. And, you know, even in the moments I've had to think about this and certainly in recent weeks and months, I've taken a hard look at skills and plugins and MCPs.

And, you know, I said half jokingly, these sort of open claw, simple text files, these fantastic markdown files, soul.md, tools.md, all these things that now are kind of an incantation of an agent. So my point, my question is this, what are your take, Roland, on discoverability, right? So you said and I agree that the humans in the loop need to direct their agent. I would contest that we also, as Bitcoiners and within the Bitcoin sort of sphere, have to make Bitcoin, Lightning, Cashew easier to discover, easier to install, easier to operate. What are your takes on that, where we are and where we need to be? I have a cool story. And I think this is 100% true. So I'm working on this project just on my own. It's called the Autonomous Agent Onboarding Project.

Because, yeah, so since the start of the year, I mean, I've been having so much fun with OpenCore, super inspired. Like, if you haven't listened to the Lex Fridman podcast with Peter Steinberger. Clawfather. Yeah, yeah. I'm halfway in I hear the good stuff's on the back end so I look forward to getting to that I mean the better stuff So to kind of explain my experience with OpenCore I've been full on working on AI stuff since the start of the year and at the beginning I didn't have much time to set it up. One of my friends, he really recommended OpenCore. Have you tried it? He's not a developer, but he's in tech and affiliate marketing and stuff,

and he managed to automate a whole bunch of the stuff that he's doing. He was really pumped about it, so I'm like, okay, I'm going to try it out. And I set up a fresh VPS and installed it and connected to Telegram just to try it out. And it was like, at the end of the day, I'd been working all day. I'm like, okay, I don't really have time to do any more than that. So I turned off my computer and I was going for a walk with my dog. And I had it on my phone, right? And I just started chatting. and then yeah and then it just and then the the magic moment moment happened right it's like i i have this personal agent now 24 7 it has a whole computer to do whatever it wants it can work whenever it wants this is so good for bad yeah and and and the first thing i did was like okay

um get yourself an email address i want people to be able to email you and I was on my phone. I didn't have anything set up. And it's like, okay, here's the things that you will have to do to get me an email address, right? And I was going to say, because if you did, not to turn this into an open claw conversation, but why not? I had to set up a Proton account. If you were going to tell me that it actually succeeded in establishing its own email address, I was going to be pretty pumped. Yeah, yeah. Well, so there's the thing, right? I was like, wait, you're supposed to be doing the work, right? You're my personal agent. You're my assistant. Why are you increasing my workload? Yeah, I'm like walking. It's like 11 p.m. with my dog. And I'm like, oh, man, why do I have to do more work? I've already been working all day. And your agent is like, okay, meat puppet, go set up this account. Yeah. So that's kind of it. So I started. and actually my open core is the one building this project is on it's on github but we started

kind of from scratch like what if the agent itself could set up everything it needs what what does it need so it needs it needs like a bitcoin lightning wallet it needs um some some ai tokens right those are basically the two things it needs and from there it should be able to um set up everything else it should be able to um buy any services it needs and it can even spin up other agents um and that this is kind of funny uh because i i was the first one in the world to create an open core that could actually spin up like a child open core that would pay for its um pay for its machine it's hosting uh with so a separate vps yeah it in turn spun up wow lnvps yeah LNVPS is pay-per-queue. So it's paying for AI credits with Bitcoin. So just those two things, it managed to spin up a child like OpenCore.

It had its own Lightning wallet. It had its own, it's paying for its own AI credits and VPS. And from there, it can do other things. And so it's lobsters all the way down. It's no longer turtles, it's lobsters. And again, I want to pause to just do kind of an aside, which is to say that what Roland you are describing is you took the initial steps of spinning up a virtual private server, anyone, you know, Hesner, DigitalOcean, pick your spot, LMVPS in this case. And that's interesting and cool. And you got OpenClaw up and running. But then you put it in a position where it had a wallet where it could do two things. it could set up its own open claw and pay for its own tokens and inference through two providers, LNVPS and PayPQ, which offer virtual private servers and LLM inference completely free, no KYC, KYC free, paid for in sats.

Yeah. And so, yeah, go ahead. Sorry. It's awesome. Yeah. You said all these different VPS providers, but I would like to point out that LMVPS actually has something very unique, right? Which it has a fully public API. It accepts Bitcoin. It can be fully automated. And I think it's one of the very few that can do that. Others have captures. They have email signup. They have, you know, all these blocking things. Whereas LMVPS and same with PPQ, right? you spin up a new API key you top it up with Bitcoin and you're ready to go so this definitely is the separating factor I think and something super important and something that also Bitcoin is really focused on compared to this I think a very clear divide because I think some of these other crypto projects or like people building these like there's something called agent mail right and I tried that out

I was like, okay, I need an email for my agent. And the first thing you need to do is the human has to sign up. Like, it's impossible. Yeah, so I think this is a really big differentiator, actually. And this article that Albie wrote about what I did here, spawning this open claw, it was referenced in another article, actually, um that was reposted by elon musk and it kind of it kind of showed me that um i think a lot of the other crypto projects they're doing it for you know for the marketing for the for the big the big noise but sure but we we are actually trying to solve a problem right the bitcoin is trying to solve a real problem and that's that's why we have an advantage well and i think i would just again interject here that, you know, I have been as guilty, if you want to use that word, as anyone

over the last few months. I have been diving in deep. I built a bunch of projects, a bunch of products. And, you know, it's sort of the, a lot of it for me is building muscle. And they may or may not go anywhere, but I'm proud of these products. They're solid, you know. And so that's not to say go me, but it's to say, on the one hand, I'm in awe of the power of something like Claude Code. On the other, I am aware of and mildly terrified by what that means in terms of the concentration of power and centralization. And so as goes with money, so goes with intelligence, right? We have this choice to make the harder path to self-custody and have self-sovereign AI and intelligence and agents. Or the easier path, which I am, again, deeply guilty of at this point, is to do the thing that is easier, better UX, frankly, more powerful, these frontier models.

So, you know, again, kudos. And I think it's just to underscore the importance of being able to truly unlock AI and agents for individual empowerment and growth and benefit. I interviewed a guy, turn this into a bit of a monologue here, but I interviewed a brilliant fellow John Robb a few weeks ago, and his perspective, and I'd love to get your take on this, is that AI is a force that, much like reaching and exceeding the speed of light, is starting to thin out the molecules in a way that they accelerate into a column and reach a breakaway speed, and off you go, and I've just butchered all the physics. But the point is, we're hitting a point where you either sort of attach yourself to this train and you become AI literate and competent and capable. And you might, you know, sort of hope to get some of the great benefit that will unfold over the next 10 years or you're toast, right? And you become a puppet to the

machines. And that's dramatic and blackpilling. And that's not really my point, but I think it paints a picture. And so that's all to bring it back to, you know, talk to me a little more, Roland, about your perspective on why it is so important that we are able to do what you just described, which is that without KYC, without gatekeepers, without middlemen, you can spin up something as powerful as OpenClaw and have it go about the business of helping you personally, professionally, you know, take advantage of what AI offers. So you have a few things there. So pay-per-queue, you pay with Bitcoin, but you do get access to all the frontier models. So it's not, I wouldn't say it's any different than signing up with your credit card and kind of going directly to the centralized provider. I mean, PPQ is like. Except in fairness, you're probably going to spend about 5x, right? Versus a Claude Code Max plan. And that's not the shill for Claude, but I think it's a real barrier right now.

Yeah, yeah. So I think they definitely subsidize, right, the plans compared to API access. No question. Yeah. So that's one thing. The other thing, I think you're right. AI is going to be a massive disruptor, right? People are going to have to learn. Some people are going to lose their jobs. Any job that can be automated will be automated, right? Yes. But at the same time, and I've heard people a lot smarter than me, I think on this podcast I mentioned before as well, that the people working on AI, they've been working harder and longer. hours than ever so if it's if if you start using ai i i think at least there are going to be jobs there are going to be a lot of jobs it's just that the job that you previously had is is not going to be there anymore you're going to have to do something um more creative more more you have to work with the ai to do something that the ai can't do on its own right but i think i think

overall it's great for humanity society right I saw this stat like I think 80% of people work at jobs because they have to for the money right not because they enjoy it At least Yeah and I think this is like terrible at least for me. I've had that in the past when I've worked at a job and just like after a few years, it's just so soul-sucking that I just feel like completely exhausted. And I was lucky that I had a different option and not some people don't right they're forced to do this for their whole life and you know so um i'm i'm really i really hope that that this this revolution i mean it's going to be tough for people people are going to have to change and adapt and and be uncomfortable learn new things right um but i i think once they get past that that barrier and they kind of change their outlook

on life change how they um they have to rapidly learn rapidly adapt but i think it's incredibly rewarding um i hope yes i think unlearning and relearning are going to be are and will continue to be the two most valuable personal skills yeah um so how how bitcoin fits in here um i did and i think there's two yeah definitely two different ways and i think it goes back to what you were saying about um giving my agent access to my credit card um and i think a lot of people which i would definitely not recommend is like they give their agent access to their own accounts basically the agent operates on their behalf right whereas i would see your agent as like a personal assistant it's like it's like it's its own identity it has its own um well Nostra is perfect for this, right? It's own key pairs.

Yes. It has its own email account. It has its own AI token account. Its activity is completely separate from you, right? I'll post it in the link and just, I built a skill that's up on OpenClaw to bootstrap an agent onto Nostra. And with the same mnemonic, it boots up a CocoD cashy wallet. So it's got its key pair, it's got its cashy wallet, and off it goes, right? And again, not a developer. So the accessibility of this stuff is accelerating. Yeah, I think this is going to be super important as, unfortunately, things seem like in the traditional world, right? They're becoming more and more dystopian, more censorship, more shutting down. If all your activity and your agent's activity is tied to your own name, I think they're going to be really big problems. and the fact that you can just set up a separate agent with its own wallet, its own identity,

it's kind of separate from you, right? You maintain control of your own actions. You can give your agent as much autonomy as you want. You can still have control, a level of control as you need, but whatever it does is kind of separate. Absolutely. Yeah, very important. So Roland, both in terms of Bitcoin and AI agents, for someone who's listening, they have their Bitcoin on an exchange. Maybe they're dabbling with chat GPT. What are the first steps that they should take both with regard to self-custody of their Bitcoin and becoming more familiar with how not just to be a user of AI, but a builder? What would you suggest that they dive into? Yeah, a few things. If they have money on an exchange, I would recommend take it off the exchange.

There's a really cool feeling, right, when you realize that just with 12 words, you can travel the world your money if you fly on the plane your money arrives in the country before you do it's it's everywhere you go and it cannot be taken from you it cannot be confiscated um maybe maybe it's a bit it's a bit scary you have to take that responsibility right but i think today with hardware wallets and and you know seed seed backups and stuff it's it's pretty easy i think for most people just single sick is fine i mean there's different opinions about that i think more complexity leads to more risk of loss of funds right and that's probably what a lot of people are worried about and maybe why they would rather someone else take responsibility of their money um but at the same time if if you let someone take responsibility of your money then that's that's the issue we're having now right with with the state having

too much control and using using your money against you basically so um that that would be the first thing the other thing i mean even though ai is such a disruptive thing and people maybe uh are worried about it on the other hand i think it's easier than ever to um get started building stuff right just with a like a single english phrase you can you can build an app now and that's one really cool thing that Albie that we've been working on the last couple months is you mentioned that this thing called skills which is kind of like giving your AI agent specific knowledge. So we have one for Albie to build Bitcoin apps. So if you give the skill to your agent code code or chat GPT, whatever it is, codex I think is the open AI version it can add payments to apps

like with one shot and actually get it right and not like hallucinate and this is something just like in the last few months that has become possible and it's like a huge dump before they'll play around with AI and things will break all the time and you just get frustrated but there's just just there's been a huge jump recently you know i'm reminded of the scene in matrix in the matrix where neo uh they he jacks in and and and is it karate i forget which martial art yeah so he goes from right he goes from i don't know kung fu to i'm a master and those are we're not quite there right but i mean that's the way i think about skills yeah for agents Yep. So, yeah. And I think, well, so maybe to finish that, what do you what do you think is the first approachable step in terms of, again, building on not just consuming the product of of AI?

You mean building your own application? So what might you recommend to a family member? You know, you're out hacking on OpenClaw and you're a developer. What would you say to a family member who you would want, just as with Bitcoin, to take a bit more control of as opposed to just be a consumer of? Yeah, I haven't approached a family member yet, but I think open call would be the thing. Not because it's, I mean, it's very new technology, it's very buggy, but it's very clear. If you get this telegram message working, you get the simple flow working, people can really see the value, the magic behind it, I think. Absolutely. so and and i mean open core is not as good at like like coding or something but for the average person that's fine they can build stuff it's very much general general purpose ai right is

it's very good um and you can get it there like i basically rebooted mine i've tried a couple of things um and it's been it's been an interesting experiment from as i say building a skill where It can get itself on Noster and Post and Pay and all this, which is novel and interesting. I then tried the chief of staff, which then spins up all of the functional agents. And, you know, I burned through my Claude Max Pro 2X plan for the week and a day. That was a bad outcome, but a great learning experience. And now I have gleaned from a bunch of very talented developers how to really refine, you know, a very lean dev team, you know, and using Codex 5.3 to code versus Opus 4.6 to think and plan, you know, all these things. So I think to your point, it's evolving and it's far from perfect, but I agree with you that it's a great place to start.

So Roland, let's wrap up here. Where is Albi in a year? What are you building that will further change how people think about use money and maybe AI agents? Yeah, so maybe if we tie that back to our journey with LB Hub the last two years, now I think we have such a strong foundation, right? We have the self-custodial wallet that's able to connect to any sort of application service. we have this really good foundation LB Hub itself is actually used by LB to charge subscription payments for people who subscribe like we're dogfooding more and more of our projects every day powered by LB Hub so I think right now we're going much into a more experimental phase to see what sort of use cases can be built with albihub i think ai is is going to make a big part of that the other thing we're working on is

is further improving like our developer experience for example our experience for small to medium-sized businesses so they they if they use albihub you know accounting transaction labeling stuff like that um to to more easily operate and have less headaches when at the end of the year when they have to do accounting and things like that. Yeah, we'll be doing additional refinement. Actually, I didn't mention yet, we did recently release a huge improvement to our developer experience, which I think is quite interesting. I think I'm proud to say I'm pretty excited that I think we brought Albie actually to the absolutely forefront, the best tool for developing with agents. And there's kind of three parts to that. So I already mentioned that the skill, right,

that you give to your agent, you install, Kung Fu in this case, you install... The jack-in in the back of the brain, yeah. Yeah, you install Nostra Wallet Connect LB knowledge, how to use LB tools. But we have another thing called Nostra Wallet Connect Faucet, And if you know about test networks, it's a common thing. When you're building an app, you won't use mainnet Bitcoin. You'll use a testnet Bitcoin. Play money. Yes, play money. But we made play money for Nostra Wallet Connect. So it's another level above Lightning, actually. It's kind of like you can almost think of it. So there's on-chain Bitcoin, then there's Lightning. And then Nostra Wallet Connect is like a third layer above. And it basically gives the agent's ability to spin up these fake Nostra Wallet Connect wallets as many as they need. And this is super powerful because it means the agent can actually write end-to-end test cases when it's building and actually test the whole flow of the app.

So it can even spin up like a browser, you know, go through the app, spin up these wallets, make real payments and test the whole flow without without a human having to, you know, manually test. That is amazing. And I mean, for anyone who is a developer, they'll know it will be obvious to them that the ability to write these unit tests and to be able to truly stress test all components of the flow from, you know, inception of an action to complete payment, that's tremendous. Yeah, so I'm pretty excited. I can see it in your face. That's awesome. yeah because i mean i spent a lot of time and i was looking at um you know spark and arc maybe we can talk about that a bit but i also decided like um yeah so i was trying to find a way how can we make the onboarding even easier um but when we really stepped back and looked at what the developer experience is from the beginning uh how you build apps actually not even having to

create a wallet for yourself is even better right so not even like a wallet without channels or a custodial wallet but just a completely fake wallet that that just works like a normal one you don't have any issues with testnet coins you don't have any issues with opening channels all of that is just thrown out the window and what you do is just focus on building your app and and so So I think this is really awesome. This is a huge achievement from us. So I think over the next year, I'm going to be excited to see people using it and see what people can build, especially with AI agents this year. Incredible. Well, we're going to do another one of these for sure, Roland. This is just so much great stuff going on. I really appreciate your time today. It's exciting what you're building. Thanks for Albie. Thanks for I'll be hub, I'll be go and all the great work. I use it every single day.

You too, Sean. Thanks. Thanks for having me on the show. Pleasure. Talk soon. See you later.