Episode #539: Zero Trust Everything: Rebuilding the Internet's Money Layer

Welcome to the Crazy Wisdom Podcast. I've got David Lakhmis here, and he is the cofounder of Ika. Welcome to the show.

Thank you for having me. Excited to to talk to you today.

Yeah. Let's go go into it. So Ika, can you share more about what Ika is?

Sure. Maybe I can start more broadly. So myself and my cofounders, we started, DWallet Labs, which is the development company behind ICA, almost four years ago. And the idea there is we wanted to solve so we come from a background of cybersecurity, cryptography, deep tech, mostly, you know, Web two, but we were, on an individual capacity also, very interested and involved in crypto. And when we wanted to, professionally start doing crypto things, then we wanted to find a a big problem to solve something that requires, like, real infrastructure work, and not, you know, just finding a specific, like, user problem and and creating a specific user application.

And in in our view, Blockchain in general and cryptocurrency is basically a segment of cybersecurity. And the reason is because Blockchain allows you to do things with with very rigorous security requirements, right? So, it it it is, and it it it is built on cryptography. Cryptography, you know, until crypto emerged into our life, most cryptographer either worked at university or worked in cybersecurity. That was kind of, what people did.

And so crypto took cryptography, and a few other things, distributed computing, and created, these systems with these very rigorous security guarantees. And so for us, I mean, this was kind of our world, and we wanted to find something meaningful and significant to contribute. And the big problem that we identified that was not solved was, we call it, zero trust custody and zero trust interoperability. And what I mean by that is, if you look at at crypto and crypto adoption, you can see that first of all, the vast majority of of crypto is held by centralized custodians, by third parties, right, centralized exchanges or, other custodians. And that is, kind of antithetical to what crypto was supposed to solve.

Right? Crypto was supposed to remove a third party trust, and, the end result is, you know, the user experience is so difficult and so bad and it's so scary, I can lose everything. So people just wanna have, like, a really serious third party to trust which kind of defeats the purpose. So that's on the custody side and then on the interoperability side, you have all these different networks that started to show up and the moment you wanna do things across multiple networks, you run into the same problems where you need to trust the bridge, you need to trust and you're you're again trusting a third party. So we wanted to create zero trust custody and zero trust interoperability, and to do that in a way that still creates a great user experience and still allows mass adoption and you know, you don't have to sacrifice one or the other.

And in order to do that we had to do a lot of cryptography research, so that's why we started the Devollett Labs. We had along the years, we had a lot of cryptography researchers, and till today cryptography research is the biggest team in the company. And we have people who came from academia, including, you know, renowned, professors like Professor Borkerski, who won the Godel Prize, it's like the the Nobel for, cryptographers. And also people who came from intelligence organizations, and, the goal was to really create some fundamental research, that would be able to solve these problems for the first time. And so that is that is what we did, and the concept we were going for trying to create is a concept that we call a d wallet.

And this is what ICA is. ICA is a network of d wallets and what a d wallet means is it's it's a decentralized wallet. So when people think today about a wallet, they usually think about the user interface. You know, I have this interface, it's it's lives on my browser or in a hardware wallet, but it's kind of a thing that I use in order to interact with with crypto. But what a wallet actually is, it's a signing mechanism, right?

It there's a key and you can generate a signature with that key in order to cryptographically control assets, and that's how that's the basic mechanism that you use in order to interact with a blockchain. And so what we created for the first time was this new type of signing mechanism, we call it a d wallet, that is controlled by both a user and a decentralized network together. So there's this very unique cryptography and we'll I I'm assuming we'll we'll go into more details, but essentially for the first time you can now have this primitive, this building block that you can use that gives you both zero trust guarantees, because cryptographically a signature cannot be generated without the user, but also allows you to implement very sophisticated and robust access control policies, because the network is also required in order to sign and you can enforce that on the network side. So it's decentralized, you don't trust a single entity, and even if all the entities are malicious or are hacked, they still can't do anything without the user. So you have this very, unique and we had to create new cryptography in order to enable this primitive, but once you have this primitive, now you can start building all these different applications that can give you both custody and interoperability in a way that is zero trust but also you can implement implement any logic you want.

So there you can have great user experience and you don't rely on any single, centralized entity. So that is in a nutshell what we've been doing for the past four years. We launched Ika a few months ago, Ika Mainnet, and, now we're working with developers, like Lars that you had on the show, in order for them to use this infrastructure and create this new wave of, applications that can actually, hold up the core values and core principles of crypto, but break out of this niche that can, you know, operate these very very complex and non user friendly interfaces. So that's that's kind of the vision, and that's what we're building with Ika.

Great. Lot of follow-up questions. There's some on tokenomics. I'm gonna throw a bunch at you, and you can you can kinda decide what you wanna take. I can hopefully wrap it up with a with one specific question, but let's see.

So, the biggest thing I've been newly known in my own world is this all of the sort of, disclosure recently about the Epstein related financial networks and their connection to Bitcoin. We don't need to go into the the networks at all, but but just, like, the core question of whether Bitcoin was, sort of co opted by those organizations in order to, make it, defanged, with particularly speaking about the bit the blockchain wars, in, like, 2017, 2018, which was when I was just starting to get a a a a handle on the decentralization. I love the fact that the decentralization was discussed because that is it just feels like, web three point o is in a total state of corruption, at the moment just with, like, the the just constant scams. Like, one of my friends, he basically built this thing called rent a human. And then as soon as he launched it out, it was for agents to go rent a human.

And so as soon as he launched it, he started getting hit with these crypto guys who are just like sending him money. And then when he doesn't accept the money, then they get mad at him and like just like like it's just this and they try to break into his thing, and they're all doing it because they're like, we were sending you money. Well, you're not you're not grateful. So then there's that stuff. There's tokenomics, whether there's a tokenomics and whether that sort of goes into this this types of thing like whether there is a tokenomics, whether the tokenomics kind of goes to that to that 2017 to 2026, wave of of of crypto, whether it's part of that, And I guess we can talk about the cryptographic stuff as well.

So where where do you wanna begin if you wanna lead?

I think maybe I can I can do it, according to the order, that you, that you said? So first, I think the more, like, philosophical or, let's say, structural kind of questions. I think this is something that is, it is true for any new technology. And I think, like, I think a lot of things in crypto today are comparable to, like, the Internet, like, the .com era. And at the inter at the beginning of the Internet, it were you it was used for gambling and porn and scams and crime.

Like that was the that was the product market fit of the Internet. And a lot of serious people in the world, very rightfully so, thought that the Internet was a horrible horrible thing that needs to, you know, to go away. Right? But then if you zoom out, right, and you look at the Internet today it's obvious, right? Obviously the Internet has made the world a better place like there is no question about that.

But if you zoom out even at the time you could still get to the conclusion that the Internet would will make the world a better place because the Internet allows, for you know communication or an information to flow freely between people and it can help things become more democratized and it could it help more to like this is something that if you have a, let's call it, practical, romantic, optimistic view of humanity and the world, right? Where, you know, the world is a shitty place and a lot of shitty things will happen, but eventually, like, the human race wants to go towards a better place, right? Like, if that is your assumption of the world, then every type of infrastructure can be used for for bad things and for good things, and And eventually, if it allows for a fundamentally new thing that wasn't possible before and can make the world a better place, then it will make the world a better place. Right? Like, that's that's kind of the and and it is a philosophical, like, point of view.

And the same thing goes with you know we're all also working a lot of cryptography around privacy, right? Privacy as well, right? If you think about it, it's a basic human rights but also it can be used in order to do a lot of harmful things. But at the end of the day, if you do believe that having these things in the world will make the world a better place, then you have to you have to kind of accept that price. And I think the Internet taught us, and in in both kind of iterations of the Internet, right?

Like Web one point o had the same problems and then eventually became, oh, like this is actually good and this can help small businesses and this can help people get be more informed. And then you had web two point o which, you know, you have all the problems that you have with social media and how it makes everything more but but still at the end of the day, it is better to have, you know, free communication between people and and and, you know, the problems in Web two point zero can also maybe be solved with Web three point zero, right? Like all these things you have to kind of believe that the world can become a better place with this technology and it's worth the price that you're paying. So I do think that it is there are a lot of horrible things in crypto, both in the way that the market is structured and in the the players that are, you know, using this for their illicit, or, I don't know, just things that you wouldn't approve of maybe, but it is still as a as a a new thing in the world is a net positive.

It, creates more freedom, and it allows for more liberty, and it allows you to be more resilient against tyranny. Right? A lot of very, very important things that we should want for them to exist in the world. And that so, yeah, that's kind of the philosophical, like, first question.

And it feels like that resilient part is a really key part of because the web three point o what does web three point o promise? It's almost like the original vision of the net the web one point o, but it's with this sense of pure ability to live in that world of that scammy first part of the Internet, which I agree with you. Like, I I am a I think a lot of people might disagree that the Internet actually is this thing that that is, great for everyone. But I do I'm an optimist. I am a techno optimist.

I think this is gonna be really important. Web one point o, web three point o, essentially, now we've got this connective tissue that allows us to live in this sort of, like, this world of both promise and the shadow side of humanity, which is web one point o. Moving through web two point o, which I do believe that web two point o is a sort of authoritarian's wet dream, but it's also breaking down at the moment. And then web three point o, kind of allows for all of this stuff, good and bad and everything in between to take place on the internet with AgenTix agent with agents coming on and the agents are gonna be really interesting particularly with what you said about the d wallets and the cryptographic stuff. And so with Web three point o, we've got this now as sort of like this network that's being established.

It's really hard to get a hold on what is happening in Web three point o. Let's go into the cryptographic stuff because it's you you're solving a specific problem, which I'm very, very familiar with, which is that the wallet is just like this weird thing. Are you gonna are you guys gonna create, like, what in two to three years of where is the wallet gonna what is the wallet gonna look like? Is it gonna be something that humans interact with? Or is it gonna be something that agents only interact with?

That's a really good question. And I I don't think that the concept of a wallet as a wallet, right, like standalone, is something that is, relevant or required. It's like, you know, if we go back to the Internet comparison. Right? You know, when maybe I'm I'm get I'm kind of revealing my age, but, you know, my childhood when I had to connect to the Internet, I had to have, like, a pretty good understanding of like how communication works with my modem and how like and everything was very very manual and you had to really understand everything.

But today, you know, there's the WiFi thing and you have the password and it's over. And you're just connected to the Internet and you don't have to know anything besides that. You don't have to create your own MIRC server, and you don't you like you don't have to do those things, they just work out of the box, right? And I think it's there's gonna be kind of the same level of abstracting this infrastructure where the wallet functionality, which is interacting with the blockchain networks, will just be embedded in whatever you're doing and will have this, like, this seamless interaction as an end user. And the agents, they don't care about the interface anyway.

Right? The agents just they work directly with code. So they don't need like any abstractions on top of just generating a signature to the blockchain. For the agents, I think it's more about how do you as a user, how are you able to constrain them, right, and have these guardrails, which is something that you can do with d wallets and you can create a policy for agents and things like that. But it won't be around the interface because agents don't care about a user interface.

Okay. Great. Great. Good. Let's talk about constraining agents because it's something I'm day in and day out trying to figure out the Open Claw happened and we hosted this event down here in Buenos Aires and tons of people showed up to it because it was happening at the same time as the San Francisco Open Claw event was happening and everybody's posting on about it and everything.

So there's tons of excitement, tons of hype. Anytime there's tons of excitement and hype, it's always like, oh, what's going on here? Like, is this real? Is this not real? And it's that that cycle itself is is speeding up.

But for to constrain our conversation, the the open cloud comes out and it's all of a sudden it's like everybody's like, oh, let's get I I think it comes from this shadow part where it's like, how do we never have to do work ever again? How can I just turn this thing on and it just does it? And it and I'm sitting here trying to adapt to Claude Code as somebody who's non technical for the past like two or three years. Like knowing that this is gonna be really important for me to build all these ideas I have and like how and I got Claude Code and it's working and I'm like every time I trust it too much, it blows up in my face. And not in, like, huge ways and it's getting better and better and better, but just in ways that it's like, okay.

Well, I need to go redo that and rethink about that. And it's like, now we're talking about money, value. How do I actually, like, constrain this thing? And it really interests I think that's why what I was really picking up with Lars' interview was just like, there's something here about access control. I love that you use that word because access control is exactly what I was trying to figure out in my last employment, which was knowledge management for this organization.

How do you figure out access control? It's such an important part. It's such a piece of jargon that no one outside of, you know, cryptographic and software. Yeah. Yeah.

What do you think?

Yeah. So access control is a very, very basic cybersecurity concept that I would argue it's relevant to, like, almost every cybersecurity activity. And I think one of the one of the big problems that crypto always had, it's it's kind of its strength also created a very a very big issue. So the strength of, crypto is it uses cryptography in order to create absolute ownership, right? Ownership that cannot be revoked.

And so the way it does that is with public key cryptography. So with public key cryptography you have a private key and from that private key you derive a public key and then people can verify signatures that you generate with your private key against the public key, but no one else in the world can generate those signatures. So you have absolute mathematical ownership and that's the core kind of, the core thing that cryptography and crypto is built around. Actually like in the in the Bitcoin white paper Satoshi writes, you know, you basically have all the problem of of ownership solved with public key cryptography, right? But the the only problem that isn't solved which is the whole reason we have blockchains is double spending, right?

And because I can I can generate a signature to you and to someone else and there's no way to know, you know, that I did it twice, then you need the ledger? And then who manages the ledger? And if you trust the third party, then you're back to square one. Right? So Satoshi said like, ownership is solved with signatures, but you need a ledger that is decentralized in order to solve that spending, right?

So this is the biggest problem, like the biggest the the most strongest guarantee that you have with crypto is ownership, and then you use this absolute ownership with keys that no one can else can generate, but then what happens is if I have absolute authority with this private key, that means that if I lose it, right, if someone steals it or anything like that, then I I have zero. It's very binary, right? It's either 100% or zero and there's no recourse, there's no back, there's no like, oh I I called support and they recover, that's it, it's over. Right? So with great power comes great responsibility etcetera, but this like this very strong guarantee creates something that is practically unusable for most, right?

Like for most people, for most companies in the world, this is unusable. So this is this is like a big problem that crypto like grappled with as an industry and how to get to mass adoption and how to become the rails of the new world where people are used to like these things where you have access control, every application in the world, right? Yeah. You create a doc on on you create a note on your phone, you can share it with someone, and they can have viewing, you know capabilities, but maybe not editing capabilities, or maybe they can just comment. Right?

Like you're you're used to access control baked into everything. And now, and if you forget your password then you can recover it, and if you like you don't have these like absolute kind of, you know, I 100% own it and then I 0% own it. That this is not a thing. So so this is How

do you guys do that? How do you guys do that?

How do you guys do that?

How do you guys, give the access control to the Exactly. Wallet. Yep.

So that's so that's so when we thought about it, we we looked at the cryptography that did exist today, and this is I mean, we're not the first people to, you know, try to tackle this problem. Right? So the the the best tech that existed that was used in order to solve access control is something called MPC, multi party computation. And with multi party computation what you do is instead of having one private key, right? A single like secret, key that you use in order to generate signatures, you have secret shares.

So you can think about it conceptually like break

Like

a stock exchange?

Like a stock exchange? Secret shares like a stock exchange?

Oh, no. Secret share like a share of the key. It's like you're breaking the key into multiple shares. It's called a cryptographic share. Right?

So it's, that's that's the the the name that it's called, the secret share. And essentially you can think about it, like breaking, it's not exactly breaking the key but you can think about it like breaking the key into multiple pieces. And then let's say you broke it into three pieces and two out of those three can generate a signature, any two, right? That's called t out of n MPC. So n is the number of shares and t is the threshold and then t out of n can generate a valid signature.

So what you have is you don't have one key, you have multiple shares. Those shares have a public key. So you can validate the signature against the public key, but only the threshold can generate the signature. And then how do you enforce access control? Yours you say something like there's a very large famous company called Fireblocks that does that.

And what they say is look, let's generate the key together, like you're the client, I'm Fireblocks, right? I hold one share, you hold one share. Now you have this access control policy. Sorry?

You're trusting Fireblocks. You're everybody's trusting Fireblocks. Right? That's

Exactly. But but they're not trusting Fireblocks the way they're trusting Coinbase for example. Right? Because with Coinbase, Coinbase holds the full key. If Coinbase are hacked or if some employee gets access to the the that they they can steal everything.

Right? Yeah. With Fireblocks, they hold half of the key. So what they're saying is you're trusting Fireblocks in order to enforce the access control policy. Right?

So you can build, you can do you know, your access control policy is for example, you know, I, I want a video call, with certain people above a certain amount or, I want multiple people to approve blah blah blah. I want like I have all these access control policies and I trust Fireblocks to enforce them. Right? Got it. And then every employee in my company can hold the other share because they alone can't do anything.

You're trusting Fireblocks to so then that whatever is held by employees in the company, right, is much less sensitive. Imagine you had a private key, you couldn't just give it to anyone in the company, right? Because then they could do whatever they want and people can steal it from them, right? So it becomes much less sensitive, and Fireblocks are also not fully trusted because even if they're hacked they can't, they can't, you know, sign anything by themselves. They have to have the corporation of someone in the company.

So their attacks still exist, but they're much more complicated, right? You need to hack someone in the company and hack Fireblocks, and then you know you have the both of the shares. So it becomes much less likely. And this is this is quite a, quite a good solution, right? But there are obvious drawbacks.

So one drawback is as you said, you do need to trust Fireblocks, right, to enforce the policy. They are at the end of the day a centralized company, a centralized entity, and, it becomes, you know, a single point of failure. That's that's a pretty obvious one. But also besides that, you know, Fireblocks is a for profit company, it's closed source, they have their own thing, they have a product that they're offering and their product is a custody product. And you can't use that as a developer, right?

And now think about the AI economy, right? Like you can't use that as a composable building block in whatever it is you're building. You only can do the thing, the product that they expose because they're, you know, they have a business and that's what they're doing. And so you can use this for for example, you know, institutional custody, right? It's very expensive as well Fireblocks because there's a very high barrier of entry, right, in order to become a Fireblocks.

So you know you can use it for institutional custody but if you wanna build a, a protocol that allows you to control assets on multiple chains and maybe leverage, this asset against that asset, write something similar to what Lars has built, you can't just use Fireblocks to do that because that's not their product. So those are the two big drawbacks where this is not a composable building block and this is a centralized company that is a single point of failure. And the way that we solve that on the cryptography level is we created a new type of MPC protocol. And that MPC protocol, it's called 2PC MPC. And 2PC means a two party computation, and MPC means multi party computation.

And what it is is you have a two party computation between a user and the ICA network, but the ICA network is a nested MPC between all of the operators of the Ika network. So in order for a signature to be generated, you need both the user, but also two thirds of a very decentralized network. And then you can enforce any type of and and then you can also first of all, you're enforcing the access control by a decentralized network not by a single, you know, entity. But also, you can just use it as a composable building block because you can write whatever logic you want in a smart contract and that will be enforced through the Eka network, signing or not signing. So now you can start building these interoperable protocols or custody protocols and use it in any way you want.

It's completely permissionless. Anyone can build whatever they want, and use that as a very, very basic primitive. So that's how we solved it. We had to achieve quite a few breakthroughs on the cryptography research itself in order to make that possible, and to make it scalable and to make it, you know, be able to support, billions of agents doing, you know, however operations they want a second. So there are a lot of work on the cryptography side and making the research that the actual protocol very robust.

But at the end of the day, that was the problem that we were solving and that's how we solved it.

So you it feels like you're perfectly positioned to answer a question that's been coming up at a bunch in my mind and all laid out, with so OpenAI two years ago had a board coup, which was such an interesting board coup because, it was like such a interesting I was working at a company that was working with OpenAI at this point, and it was so it was like directly relevant and intimate to this to this company that I was working with. And it was so interesting. What was going on? And then about a month ago, I came across this tweet, a really great tweet about the difference between applied versus research. And so research is this idea of OpenAI, Claude or Anthropic both have a lot of research.

And this research is really tied into safety. And so at OpenAI, that just went crazy because all of a sudden you have this guy who's really charismatic and, you know, doing all sorts of stuff. And then it was like a fight between him, which is a board, and the board was all about research. And then and then so there's a board coup and then Anthropic has applied versus research as well. And we're I just had a great convo with a friend today about, in Anthropic, they are hiring philosophers.

Not only philosophers, but they're hiring sociologists. They're hiring, therapists. They're hiring anthropologists. All the humanities is coming back and it's coming back at Claude. And it's all part of this research thing.

And so the question I would love to ask you is, like, you've you've you've been doing cryptographic research, and now actually, correct me if I'm wrong, but you're also doing apply you already applied it. You're applying things. So what this balance between applied versus research, what's your take there?

So I think, the great one of the great things about crypto in general and, like, the promise of Web three is also decentralized governance. And I think at the end of the day, when you launch, an open source project like we've we've done with ECAF, first of all, anyone can contribute. Right? So this it's not I I mean, right now, we're the main contributors and maintainers, but anyone can contribute to ECAF. And not only can anyone contribute.

Right? If, for example, most people, most, so this is a permissionless network. There's you asked about a token, right? There's a token, you can stake it in order to become an operator of the network, and the majority, the two thirds of operators control the network. So even if we decide, oh, we don't agree, we think, you know, as a team this should be the direction.

But two thirds of the network which, you know, control the tokens that are out there in the market, if they believe that there's a different direction, they control the network. We do not control the network, right? All we can do is propose code that we believe is good, but anyone can propose code and at the end of the day the network decides what is the code, you know, what is the code that will be implemented. And I think this is a this is a huge shift in because all of the things that you just described, I mean it's great that they're happening, but they could also happen out in the open.

Yeah yeah yeah yeah yeah yeah yeah.

It could happen in the open and then whoever has a stake in this project, right, can just decide what is this project and what does it do and what it doesn't do. And law, you know code is law, is something that I believe especially with AI is becoming more and more accessible for people, right? Because anyone can take whatever idea they have, whatever they want the outcome to be, and they can create code to support that outcome. You don't have to be like this super developer anymore in order to do that. So you can have an actual literal marketplace of ideas that are succinct and are, actionable and then you can have a system, a decentralized network with governance that actually enforces that, which I think is an amazing advancement to how humanity governs itself.

Yeah. I really like this because it's basic because the governance model we have right now is a a very old it's a very old governance model which is shareholders and this is this is corporate governance not political governance. Shareholders shareholders have a stake, they have voting rights, or they have economic rights, not all the time those both of those things go together. You've built this cryptographic access control and it's a two thirds majority which is super interesting philosophically because I guess the parallel to that would be, I don't know enough to go into where I was about to go of like shareholder governance and how

It's all super majority. Super majority. Yeah. Interesting.

And do you know how it works inside of corporations? Like, how does how do they vote inside of corporations? What goes on with that?

So in corporations, you do have it is actually more similar to political stuff because you have, you have voting by shareholders, but then what they basically do is they choose the board and the board chooses the executive. It's very representative. Right? So you don't have like a shareholder vote on every single thing. Right?

The shareholders, sometimes they can vote on specific big decisions, but they usually vote on who the board is and the board votes on who the executives are and then the executives do the day to day work, right? Like that's usually how it works. And I think this is one of the great things about making things, like making governance decentralized and using, crypto like crypto network blockchains as the fundamental structure is that you can implement code as law and you can put these things that are you know they today they have to be done in a very representative way by by choosing your your representative in order to enforce whatever you want. You can make those happen, in in a way that is very I forgot the word, but it's like, one to one like you can't you can't make a mistake or you can misinterpret it, right, because it's code. And I think that another interesting thing that we're seeing now is things like prediction markets or decision markets where people can very easily vote on specific outcomes, right, which also I think I don't some people think that it will replace governance altogether.

They call it future key. I I don't think that that will actually be the case, but I do think it'll be an important part of governance. Yeah. Yeah. Yeah.

So you will have these decision markets and prediction markets and people will be able to know what is the expected effect of specific decisions in order to help them make decisions.

Oh, it's like decision trees. I've been thinking a lot about decision trees within the, context graphs, and the graph, graph databases in general related to ontologies. Hopefully I don't get too carried away with where I'm going. But, so, knowledge graphs are related to ontologies, ontologies are related to sort of decision trees because I'm I'm I'm making a lot of decision trees with these coding agents. I'm making a lot of decisions.

I'm also outsourcing a huge amount of decisions to it. Then there are the orchestration frameworks which are even more decisions to this thing. Now decision trees are really or traces are really important because you can run those back up as long as it can reason about itself, which it can't, but it's getting there and and like make everything explicit which anybody listening to this please try to make as many things explicit as possible when you're trying to tell this so you have these decision trees so you can kind of trace it back and this is really important for not only the governance. We're we've been talking about for organizations, for republics and everything like that is how did we come to this decision? What were the other decisions that we could have made?

How like how do we get all the all the way here? And I love this idea that Futarchy with the prediction markets is like one aspect of that, right? It's like okay you get you get all these decisions, there's money on the line but then there's also these other sort of insider trading which I guess kind of runs into it because we now know it's insider trading rather than before which was like insider trading where we didn't know about it. And so, and then that's just one piece. That's just one piece of the the giant transition that humanity is going through which basically nobody understands the whole transition.

I I don't think we've ever really understood the transitions that we've been through before but like at least they were a little bit more manageable. Now it's just like

It just it just took more time. So you had more time to it's like the frog in the in the boiling water. It's now it suddenly became, like, really boiling really fast. Like, you need to, you know, you need to ease into it.

Yeah. And but but the agents can help help us do that. Right? So there's gonna be there's gonna be a feedback loop because now the agents are making it easier for me to track the decisions, but I can't track all of them. I have to, like, specialize, but then, it just totally throws the whole specialization thing in the window.

Actually, I would love it. We can we can go here unless you wanted to say something about what we just talked about. But specialization as, the like, special specialization as the only thing that anybody should do feels like it's a product of the industrial age, and we're headed into the industry information age. Might be my own personal cope because I'm not a specialist. But, what do you what do you think about generalism versus specialism in this world that we're entering?

So I think I would I would, it's it's hard to do that because, AI I mean, let's let's say LLMs, right, to be, like, very concrete. Right? LLMs have, really created a very demonstrable impact in a very short time with things that, you know, two, three years ago would have seemed just science fiction. Right? So so it's hard to do that.

But if you if you zoom out a bit, right, technology has always done this. Right? It always took a specific type of, work that had to be done manually and just made it much, much easier, like either completely automated or, you know, 90% automated, whatever. Right? Like, this is what technology does.

And it it happens since we, you know, since the the agricultural, you know, revolution all the way to, to today, right? And I think just when we see the new class of, let's say software engineers, right? Being so impacted and, you know, I can give you examples from our company. You mentioned about being explicit. Right?

You can't get more explicit than a scientific cryptography research paper. Right? Like, this is the most explicit thing ever. And when when a human being needs to implement cryptography into code, it is one of the most like there are very few people in the world that can actually do that. Right?

Yeah. And and it takes them a long time. And then you give it to Claude Yeah. And it it is able to do that. It's insane because it is very, very explicit.

You you don't get more explicit than that. So you have a protocol and it implements the protocol. And, yeah, it's not like the most optimized and there's still it'll get better, but it still is able to do something that would take a human being six months. It does it in two days. Right?

And it's not just any human being. It's like the smartest, you know, most, like, specialized, most, you know, serious people who know how to, you know, both math and computer science and can input like, this is a huge and for us, it was a big, a a game changer, honestly. Like, we're faster than than before, like, an orders of magnitude faster. And I think this is, the this kind of shift is I guess what people, that worked at a factory felt like when factories started to become automated. Right?

It's like this is this is the same type of shift. So I I don't think that it this is new in, in quality, but it is just something that we didn't expect to happen and it caught us by surprise and it is affecting a lot of people. Right? Maybe I don't know if actually more than, you know, previous waves of automation, but definitely different classes, like different types of people that we thought would be affected. And you can think about, like, the services industry and middle management and lawyers and accountants and, like, all these things that we always thought will just be there forever, and we now can very clearly see a near future without any of those people.

Right? So or with a a a friction like, a fraction of of the amount of people that were required in order to do those things. So I think humanity will reorganize itself around this just like we did with every new piece of technology. And at the end of the day, technology helps increase the the quality of living, not decrease it. Right?

Even if we'll have a transition period where there's gonna be a lot of uncertainty and people will lose their jobs and it we'll have, like but at the end of the day, it will increase productivity, so it will increase the quality of life. And we'll need to just rearrange humanity around these very, very rapid changes. Right? It's also faster than than anything ever before. But I but I don't believe, like, this is inherently different than anything that we've, you know, experienced in the past.

Yeah. I love it. It reminds me of a book Future Shock, by Alvin Toffler. I've had a few I've had a few instances of Future Shock myself since the LOM showed up, and then I've seen it in other people as well. And the it it it I I agree with you.

It's not anything novel. I mean, it's novel in the sense that how scale and how quickly but these are all things that technology does is like speed, efficiency, cheaper, better, faster, cheaper. Right? It makes everything better, faster, cheaper. And I think that we are gonna enter a world where everything is better, faster, and cheaper, which is so interesting given the last fifteen to twenty years of intitification.

And but, everything I've seen so far, it's I'm building software better, faster, and cheap cheaper. Not only that, I guess it is a little bit of a of a zero to one situation because I am now in a position where I can build software where I could not build software before. I guess I could have hired people to do it. So there is a little bit of a connectivity. But, yeah, better, faster, cheaper.

Technology is doing that. But that is a future shock. And I feel like, Neal Stephenson, I think his book, The Diamond Age, really gets into this world that we're entering of, like, we're gonna create so many weird social, technologies for getting into this world in a way we're all gonna kind of become full aristocrats. Right? Like like, where we're gonna be able to go do these things.

But

this this has already happened. Right? Like, if you look at at your life today or my life today, we're aristocrats. Like compared to four hundred years ago, we're more than aristocrats. Like their their aristocrats had a shitty life if you compare it to our lives.

Right? Like this this is always happening. What I'm saying is that this isn't new. It's just very fast and it's affecting way more people at a time than what we experienced before. But not changing anything significant.

And I think eventually, the trend is always upwards. And if we are able to be more productive as a society, right, as an as a market, as, so we'll, a society, right, as in as a market, as, so we'll be able to create to produce more and to increase the quality of life of everyone. How do we, make sure or how do we solve the distribution of of resources in you know this changing world? I don't know it's a big problem. I think maybe crypto can be a big part of that, right?

But but this these are problems that eventually will be solved because you have to have like a society where, if there's more prosperity then it also trickles down, although I hate the word, but it it does eventually trickle down to everyone. You have to have that in order to continue growth. So we'll find the way to do that, whether it's UBI, whether it's other ways. Right? Whatever it is, there will be ways in order to facilitate that.

Yeah. I really like that because it goes back to the Futarchy kind of like it's just one piece. UBI might be one piece, but we'll develop new pieces to fit into that. I want to go back to the crypto because it's so interesting. And I love that we're in a down market now because I love the down markets.

The down markets are the most interesting, interesting times to to to think about where this technology really can lead. It really feels like the last year or so interest we we, I've been running an AI group down here for the past, two years, and we've partnered with a crypto center down here. And I've was introduced to the fact that Argentines run, like, 50% of the crypto companies, or part of the 50% of developers. Like, all the people who build building it there, they were doing the same stuff that was going on in 2017 when I was in Silicon Valley. I had no idea.

And it's one of the largest markets, and so Ethereum is setting up down here. And so we've been partnered with this crypto world and so I've gotten back into it after a little bit of a disillusionment. And it really was really weird how all of a sudden DevConnect happened and then like just like a light went out. All of a sudden everybody just stops talking about it except for the Twitter accounts that were going crazy with all the stuff I had mentioned earlier. Like that was and that was that's not good for for for the, people's understanding of crypto.

Do you think we're we are at a sea change of, like, Bitcoin, Ethereum? Well, actually, that's a good question for you. Do does ICO represent a threat to Bitcoin or Ethereum?

So I I don't think so. I think, I think we are very, very it's like, trying to think how the Internet would look today back at 1999. Right? It's it's very hard to know, you know, what will be what will still be there in ten or twenty years. Right?

I do think that you have some very fundamental things that, crypto is useful for and things that are basic infrastructure in crypto that are required. So, obviously, the the concept of, an immutable decentralized ledger. Right? Like, that's that's obviously and and Bitcoin embodies that, but also others. Right?

So I don't know if Bitcoin will be the one that wins. It has a lot of traction and brand recognition, but, you know, those things can also go away. I don't know. Yep. Then you have, like, the infrastructure for creating shared state.

Right? So, like, l ones, smart contract l ones like Ethereum or Solana. And they they, you know, there are a lot of different ones. There's Sui. There there's a bunch of different, l ones and each one has different approaches.

I don't think this is a winner takes all but it's also almost impossible to know, you know, what will who will be the winners in in ten years or twenty years. Right? But this is definitely a problem that needs to be solved. So it's either one of those players or or another player that solves that problem, right, that will be there. And then you have problems around like the things that Ikar is solving.

Right? Like let's call it, secondary infrastructure. Right? Like how do you solve custody? How do you solve interoperability?

Because you won't have just one blockchain, right? And then how do you solve privacy? Right? Blockchains are public and you need to solve privacy. So you can do that in a centralized way and relying on hardware blah blah blah.

Or you can use cryptography like zero knowledge proofs or FHE, fully homomorphic encryption. And and so all of those things are things that will exist. Who will win? I don't know, it's impossible to know right now. Obviously, we're aiming to win our categories, right?

And we built right now we have the most advanced cryptography and the best solutions, but we we know that we need to keep working in order to, you know, stay ahead of the curve, right? Because eventually, and you talked about the down market, right? Eventually the down market goes up again and when it goes up, you want to be there to capture that activity. So you wanna have the best, most scalable, cheapest, you know, most secure, most decentralized solution out there, and you want to get as much traction as possible. So you start creating that moat that today Bitcoin and Ethereum and Solana and all these networks have, that makes it harder for others to take their place or at least gives them a better chance to still be there in ten or twenty years.

So I think this is kind of how we're thinking about it. There are major, problems or major aspects that need to be solved. And if you're solving one of them, then you need to do it in the best way possible in order to get as much traction as possible to still be relevant as, you know, as we move forward.

So let's go into the interoperability part because we talked about the access control a lot, but it sounds like there's a whole other part where there's interoperability and that is so important. And I've studied this a lot, not to the degree that you probably have. But, what, because so what what what type of interoperability does ICA, create?

So, this goes back to the the zero trust guarantees that you have with the block Right? So when you operate on a blockchain, and this is actually a bit of a misconception that a lot of people have. So people think that because something is decentralized, it means that instead of trusting, one centralized entity, you're actually trusting a lot of entities. But that's not true because the way that blockchains are built, and this is actually in the name, a chain of blocks, is that you don't even trust the operators, right? You don't even trust validators or miners because you have a chain of blocks that are cryptographically, you can verify them from Genesis, right?

From the moment and you can verify that there was no illegal transaction that was included in the state. So no one can cheat you even if all of the validators of Ethereum or all of the Bitcoin miners, right, want to include an illegal, transaction. Right? So let's say they want to take all your Bitcoin and send it to themselves. So they sign a trend like, they include a transaction that wasn't signed by you.

You can verify the entire state from Genesis, go over every block, every block is cryptographically connected, and verify every signature against the public key. So you have this structure where you actually don't have to trust anyone, right? The decentralization is just to remove the single point of failure, but you never have to trust anyone, it's zero trust. And that guarantee is very strong, but it also creates huge problems for interoperability. Because the moment you operate with two of those systems, now you need to trust someone to tell you what is the state of the other, right?

So if you're operating right now on Ethereum and you wanna do stuff from Ethereum to Bitcoin, you don't have that zero trust guarantee. So you have to trust someone to tell you, oh this is the state, this is the correct state of Bitcoin, right? This is, and then you can do this stuff of Bitcoin but you have to trust them. So interoperability by its nature introduces this trusted third party into this trustless and and zero trust environment.

Mhmm.

And the way that you can solve that with Deakin for the first time is a d wallet is a, signing mechanism. Right? It can control assets on any blockchain. But it is controlled by both the user and a decentralized network, which means you don't trust anyone to tell you the state. You're always required in order to sign transactions.

So in the context that we gave before, right, let's say you're on Ethereum and you want to do something on Bitcoin, you can actually control the d wallet that holds the Bitcoin and control the d wallet that holds Ethereum from the same place where the user always has the final say from a cryptographic perspective. Which means that you can do things like swap between native Bitcoin and native Ethereum, and you can make sure that you know the Ethereum doesn't go through if the Bitcoin doesn't go through in an atomic way. Right? For example, but you can do you can implement any logic that you want that requires multiple assets on multiple chains and the user is still required to approve them so you don't have to trust any third party.

Sounds like you're, counter positioned against Chainlink because Chainlink feels like, it's sort of like a nation state centralized bank, type of thing that

So Chainlink yeah. I mean, Chainlink, just like any, and I'm talking specifically about the bridge component of Chainlink. Right? Just like any any other bridge, they, you know, you have wormhole, you have layer zero, you have a a bunch of different bridges that are supposed to connect between networks. But they in fact serve as the trusted third party that you need to trust that tells you what's the right?

So what they do is they tell you, you give me your Bitcoin. Right? Not Bitcoin. Actually, most of them are not active on Bitcoin because they need to have smart contracts. But let's say Ethereum.

Right? Give me your Ethereum on Ethereum. I will hold it in an address that I control and I will then issue a derivative, a wrapped asset on Solana, right? Wrapped Ethereum and I will give it to you. And now people who you don't actually do things with Ethereum, you do things with a wrapped Ethereum and you trust them to keep the actual underlying Ethereum so when you give them the wrapped Ethereum back, they'll actually give you your Ethereum back, right?

But for example, there were a lot of hacks like bridges are one of the most hacked things in crypto because it creates a honeypot, It creates a single failure.

Wow. Interesting.

And one of the and one of the most interesting things with this zero trust dynamic is if you today, if you break Ethereum, right, let's say you manage to take control over more than 50% of the network. It's it's almost impossible to do. But let's say you are able to do that. Right? You found, let's say there's a bug in one of the implementations of Ethereum and you were able to take over validators.

You can't really do any damage to anyone on Ethereum except if you're using centralized entities. So for example if you off ramp to a centralized exchange and then take out the money, then you can do damage. If you use a bridge, you take you you can fake transactions and drain all of the, assets that are held within bridges because you can fake as if I returned wrapped Ethereum now give me back the Ethereum right? So the only thing you can actually steal because all the other things are reversible in a Blockchain. So you can the moment you get back, you can say okay, these things are just being reversed and I I just don't recognize, right?

You have the zero trust guarantees, where you don't have to recognize illegal transactions. So the only things that are actually vulnerable when you hack a blockchain are the things that are connected to the things that remove the zero trust, like centralized exchanges, centralized custodians or bridges. So with EK, you solve that because you're actually controlling with cryptographic guarantees that the nothing can be can be generated without the user. You're controlling a native, address, a native account on Ethereum or on Solana or on Bitcoin.

Okay. That's brilliant. So the wallet is really interesting because it's it's approaching the wallet the way, well, so let's go into the other wallet, which is, I mean, there's a bunch of wallets right now. What I can't even remember the name of it. MetaMask MetaMask, MetaMask.

All that programmatic stuff is all happening in the browser. Right? So the it's they they they've sent they're a very basic wallet. What based on that small insight into the my head about what MetaMask is, I don't actually know that much about it. What what's really going on with the current wall wallet eco economy economy?

So with MetaMask, we you can think about it if you break down what what they have. Right? So MetaMask has, first of all, it's it stores the private key. Right? That's one thing.

So first

of all, you have the private key stored. Let's talk about the MetaMask extension. Right? That's the easiest way to talk about it. So it's stored within your computer's physical memory, right?

So that's one thing that they have. The second thing that they have is an interface for the user to be able to generate signatures, right? So you have the the key, you can generate signatures with that key, and then you have a mechanism to broadcast those transactions into different, Blockchain networks. Right? So you you have the key, you can sign a transaction, and then you can broadcast it.

And MetaMask, it it does a, you know, it does a very good job to provide that functionality, but that functionality in itself has a lot of, drawback. Right? And and it's, it's not something that you can expect, you know, the majority of people in the world to use. And if we're talking about things like AI agents, giving an AI agent control of something akin to a MetaMask is very dangerous because the the agent can hallucinate, the agent can be prompt injected. There are a lot of things that can happen to the agent once it controls real money that you actually wanna make sure that there's some guardrails there.

You don't wanna give them full control. So while MetaMask does what it does really well, what it does isn't what we actually need in order to create mass adoption or in order to bring agentic operations, into, you know, into our into our lives and to use crypto as underlying rails for those things.

Beautiful. Okay. So we'll take another about ten minutes left. This wallet thing is really interesting. I wanna dive into this because I'm I'm I'm now thinking I wanna use ICA to, start making experiments.

There's the Hearst. Hearst was, there's the x four zero two, I think it was called. This sort of protocol level, like, how do we use the Internet in a way they thought of back in 1995 or 1996 of how to transact value and getting the four zero two error message. And so they built these sort of microtransactions and yet it again, but it feels very I think it's called like AstroTurf where, like, they try to make it work with a bunch of money and, like, putting a bunch of things, and it doesn't really, like, catch on. That feels like where the state is right now, basically.

And so where how can I how can I start to transact? Well, I really like this bridge, the the lack of a bridge, because that means that in the access control actually, those are the best questions. What what can I give my agent access over that's a really safe way for me to spend a little do a little bit of an experiment to try to figure out how to actually give my agent the ability to transact? Yeah. What what can I do there?

So, this is actually, there's a there's a company called HumanTech. Mhmm. Funnily enough, they're also building, tech for AI agents, not just for humans. But they're they're building using econ. They're they created something called it called a WAP, Wallet as a Protocol.

So today you have things that are Wallet as a Service. I think the most famous one is Privy. I don't know if you've heard of them. They were acquired by Stripe. Mhmm.

A lot of these companies were acquired by larger, in fintechs because they want access to users. Right? Because they're centralized companies, when you create a wallet with them, you basically own the user. Yeah. And so what they're building is the decentralized version of that.

Wallet as a protocol instead of wallet as a service. And what you can do with Wallet app as a protocol, they'll have an specific, like, agentic wallet, offering there. And basically what you can do with that is you can have as a as the user, as the owner, the admin, you can have a backup of the full key, right? So you know that you can always do whatever you want, right? You keep that in very deep cold storage, but you know that no matter what happens with the Eka, with your agent, with whatever, you have your full backup.

But then what you do is you import that key into a d wallet, and that gives you the network share that Ika controls, and you write a policy for that and they have like interfaces that are very easy in order to create those policies like spending policies or maybe whitelisted addresses or protocols that, you know, that can be engaged with. So that is enforced by the network and then the agent gets the user share. So the agent can't do anything by itself, it has to abide by the policy because if it requests a signature that doesn't pass the policy then IKAC just won't sign. And then IKAC also can sign by itself because, you know, it only has the network share so even if IK is somehow compromised, you're at, you still have a zero trust guarantee. So as a user, this is kind of the the best of all worlds because you don't need to trust anyone, you always have full control yourself, but you can have autonomous actions being, being done for you in a way that you know that there are strict guardrails that can never be, you know, reasoned around or hallucinated around by the AI.

And you don't have to trust. So today, in order to do that, you'll have to trust an entity like Coinbase or or Privy, right, and give them full control of the wallet, and that is something you don't wanna do. So this is the way to do that, and and again, they're using ICA's infrastructure in order to create that as a specific product. So you'll actually be able to just connect, plug your your agent in, you know, you get as a user, you get the the the key or you put you input the key, and then the agent gets their share, you the policy gets rid of to a smart contract and is enforced by ICA, and it's it's just, yeah, completely hands off from from there on.

Yeah. So it's the access control. I was just, right after we get off the call, I'm gonna go back to what I was doing, right before we got on the call, which is access control for a new server, that I gotta set up so that I can, because I learned or actually one of our AI Whispers, part of our community, he was a fellow non technical guy who picked it up really quickly and who's an entrepreneur and just knew how to use it. He, we're gonna learn from his mistake because he made a mistake of of getting the server up and then he got hacked by a crypto guy. A crypto guy went in and hacked his thing to mine crypto and, it's all because he gave root to the server and then he gave sudo to the parts of the server as well.

And so like I learned from that basically I just have to lock that shit down basically. No any sort of process and it was really funny because, here's a perfect example what I was saying earlier with the agentic stuff is that I asked it to set up a Docker, and then I started asking. I have a planning agent and I have a coding agent. And so the planning agent told the coding agent to go set up the Docker. And then I have a conversation with the planning agent.

I say, what do we need to talk to about security? I'm about to do this thing. How do we get this going? And it said, well, on the Docker blah blah blah blah blah, Docker docker docker. And then I asked it again, you know, did you give root access to this Docker file?

And And then it said, no. Definitely not. No. We haven't done that at all. And then I go and check the coding agent.

It definitely gave root access to the Docker so I would have been setting that up with a total root access into this thing if I had not double checked. You just always gotta double check, double check, double check. And 100%.

And by the way, I think one of the one of the big things is going to be, you know, like cybersecurity red team agents. Right? Because Yeah. Cybersecurity is complicated and people even even experienced developers who don't have experience in cybersecurity Yeah. Do a lot of those kind of mistakes.

Right?

Yeah.

But but you do have like the industry evolved in a way that it it became pretty mature and you have like all these different cybersecurity solutions and the more the company becomes mature the more you you adopt them. But with with AI agents people are just like building things so fast and no one is looking at cybersecurity so that's definitely gonna be a big thing. But I think a very interesting part of of that connected to ICA is, you know, we're talking a lot about crypto use cases, right, custody and interoperability, but I believe that a threshold signature network that can enforce access control on how signatures are generated is relevant for so many things in the world. So like the world, people don't know, but the world runs on cryptographic signatures.

Yes.

When, you know, when you, enter a website and you see the the little lock, right, that's cryptographic signatures because it's end to end encryption. When you, get an update to your smart TV from Samsung, that update is signed by Samsung's key because otherwise it could be a malicious update and your, your TV won't install it. Like, there's all these things in the world and this is called the, the thing that approves like, you know, Samsung's, approval of of the update, it's called the CA, a certificate authority. Right? So certificate authorities also in the Internet, right, when you have, the HTTPS, the TLS, that is signed by your certificate authority.

So when you see this little thing it's because one of the very few certificate authorities in the world, right, there are very few of them, has signed the certificate that yes, this is actually the website, this is actually the server that owns the the the NS, right? The the the domain, right? And they sign that this is actually their website. So now think about having a decentralized certificate authority, right? Right now, today, if someone is somehow able to hack one one of the certificate authorities in the world, the whole Internet is broken.

Yeah. You will be able to impersonate as any website you want because all the all the browsers, they all respect that certificate authority signature. So if I am able to somehow get the key of one of those certificate authorities, I can impersonate any website and any anything in the world. Right? And all the browsers will respect it.

Now imagine a world where you have a decentralized certificate authority. Now instead of breaking one, I need to break tens, maybe hundreds of different operators in order to be able to impersonate a website. Right? So like there's a lot of use cases around cybersecurity, whether it's internally in how you manage access control of, of agents or or of different permissions that you have internally or externally on the Internet, and how you push out, you know, updates or things like that, how you maintain open source projects. Right?

Right now if you have a maintainer, they have the authority, and if someone is able to hack them they can impersonate them and then everyone gets the malicious update to an open source library. What if you have a decentralized certificate authority and not just one maintainer that holds it, right? That there's all these things that I can solve in the real world because decentralized access control is a huge, huge deal, and if you can remove a single point of failure, it becomes, a really significant upgrade to security.

Yeah. The decentralized access control, that's the jargon. Okay. Well, thank you so much. That was a brilliant interview.

How can our find our listeners find out more about you and more about Ika?

So, you can, follow Ika on x. It's at Ika, I k a, d o t x y z. So Ika dot x y z spelled out. That's, that's our profile on x. We we usually communicate there.

My profile on x is d three h three d, and, we're, yeah, we're, you know, the the website is ika.xyz, ika.xyz, and we're going to soon launch, another project. This is actually the first time where I'm talking about it publicly, but we're going to launch another project around private cryptographic privacy called Encrypt. So that will be at encrypt.xyz. And, and we haven't really talked about it publicly yet, but we will start talking about it soon. Great.

Thank you so much.

Absolutely. Pleasure.