S1E194 - Leshundra Robinson, Jee Vahn Knight, and Christopher B. Davis

We developed blockchain technology before its ideal user was even on the planet, which is AI. Human users noticeably vanished from the system. They started to realize the next wave of users is not going to be these humans coming back. It's going to be agents transacting on these humans' behalf. At Ampli, we are building solutions for a gender capital management. That means what we want to enable is AI agents that autonomously transact money and digital assets. What we do is we introduce separation of powers where we make sure that the agent never holds the keys. Hello innovators, entrepreneurs and risk takers. Welcome to another episode of Web 3 with Sam Kamani podcast. And today on this podcast episode, I am interviewing Patrick from Ampli. And this is an awesome episode. Our conversation went on for nearly one hour because there was so many insights that Patrick was dropping about this industry. We are recording this live at consensus in Miami. So if you are a founder or a builder or someone interested in crypto blockchain, AI, all these type of things, then this is the episode for you. As always, nothing that we talk about here should we take in as investment advice and please like, share, subscribe and follow with all that out of the way. Let's get into it. So Patrick, great to have you on the show. Looking forward to this interview. We are right now here at consensus recording this interview live. And I have so many things I want to talk to you about Ampli and everything that you guys are building. But first, would love to learn a bit more about your personal journey. How did you get involved in this industry? First of all, thank you for having us, Sam. Thank you for having me on your show. So I'm a political scientist originally. Study Political Science in Germany and in France. And I discovered Ethereum and Bitcoin in 2016. And I learned about smart contracts. And as a political scientist at the time, I was mesmerized. Because I realized, wow, suddenly we have this technology that would allow us to solve a lot of the human coordination issues that political scientists have been working on in governance studies for a long time. So that got me really interested on an intellectual level. And on a social level. And at the same time, gladly, I also started investing. Because I also saw the merit of Bitcoin and the whole origin story following the global financial crisis. Yeah, and then a few years later, I went full time into the industry. I escaped academia, which is a very broken economy actually in Europe. I'm sure it's by now all across the world. And yeah, I started advising protocols and projects and tokenomics. And governance, decentralized networks have our governance problems to solve. That's what we do with tokens. We create incentives for positive some behavior. So that's actually way up my alley. And then I met my co-founder, Xavier, in Amplie. I met him actually at the very start. And we started writing research reports on projects that interested us on tokens. We got about like Ravencoyne when it was still being mine before Katrina on exchanges, for example. And then we had a couple of businesses in crypto. And yeah, Amplie is, I think, the one business that will really matter. So yeah, that's the story. That's fantastic. I have lots and lots to talk about. All things such as the whole downturn of SaaS and the stablecoins. And all the things that you guys have been writing on. Before we get into this, let's first talk about Amplie. And what is the key problem that Amplie is solving? So at Amplie we are building solutions for a gender capital management. And we believe that that's going to be a big part of the next major upgrade for finance, in general, for traditional finance. It's going to consist of tokenization of assets and of agentic intelligence managing those assets in the future. And this has been going on in D5 for quite some time. And the problem that we saw was that the agent capabilities they grew much more quickly than the infrastructure and security stack below them. So the agents became more powerful, but at the same time, the way that we had them interact with the assets was a bit finicky. And when we started Amplie, we actually wanted to build consumer products using the agents. We got very, very excited. We were advising one of the leading agent providers right now called Giza. Very early, so we got a sneak peek at what these agents could do. But then while we were building it, we realized, well, I wouldn't want my friends and family to put money into these systems yet. There is still some evolution to do on the whole like execution security and custody security side. So that was how our mission got formed over time. And now we are at a point where we can more comfortably ask the question, answer the question, can I trust an agent with my money? Yes, you can if you prepare correctly. In our case, we introduce a separation of powers where the agent only ever sends suggestions to your system that you fully control and you set up policies that are written on chain that neither the intelligence provider, so like the companies like, for example, Giza has access to not the agent. So you can very, very cleanly mitigate the risk of if the agent ever gets compromised or if the agent ever starts sending signals that would be malicious. And that's like our main value proposition. We make that reliable and secure. You don't have to trust the agent. You can profit from its intelligence. You can profit from its superhuman capabilities, automation and those sort of things. Exactly. Exactly. The 24/7 aspect of it, right? You wouldn't want to sit there and then confirm everything the agent suggests because then you lose all of the upside. You wanted to be looking while you're sleeping and the result exploit happens in DeFi. That was actually a situation where some of the agent providers that we work with, their agents pulled out first before any human could. And I think that's like the main value proposition that we can bring to the table because we mitigate the agent risk and the agent can mitigate all the de-fire risk. Yeah. Yeah. Yeah. There is like, you know, every day you hear a scary report that agent deleted the database of the company or things I'm sure you guys have read all those two. How do you, you know, ensure that your agent, like especially if you're not approving every single transaction. Yeah. How do you ensure that agent is doing the right thing? So in the case of like coding agents, you're actually running into maybe even more difficult situations to find like the right guardrails and to set those up correctly. And somehow we ended up in a situation where it comes to the management of assets that we have a technology that's perfect for this situation. It was just blockchain. And we've like one of my favorite things that I like to say is that we developed blockchain technology before its ideal user was even on the planet, which is AI. Like it's not humans. It's blockchain has been historically very cumbersome and a bit, you know, anxiety inducing to use, right? I'm sure if you have been interacting with crypto for a long time, you you you set there and you had these transactions where you felt a bit nervous signing it, right? Yeah. And that happens to me too. Like even after a decade working with this, it's it is still very, you know, there's something that's. It's not human in a way because it's not designed for humans. It's all the big yeah, like big codes and things like that. And it's like we don't want that. We don't deal like that. We don't think like that. Exactly. We think in words. We don't think in numbers. And this is also where the majority of the exploits are happening right now. Yes. I was at Red Summit recently and the security researchers there that were telling us that more than 90% of the money that got exploited last year in crypto was due to human error and social engineering and front end school thing. Yes. So like the private. Agile, Agile, scale, Agile. While the smart contracts actually they get exploit less and less and less because that layer is actually hardening, which is cool. Yeah. So so what we need to do is we need to minimize the interactions that we have where we can't really surface the data correctly. Have an agent do that and find a secure way to control that agent, right? And luckily there's this first interesting synergy where the agent can interact with blockchains on our behalf and make that much more human friendly. And at the same time we can use blockchains to contain the agent's behavior. That's what I'm saying. It's a match made in heaven. It's perfect. So we can already have the blockchain tech. It's been hardened and battle tested by our humans for you know since the Bitcoin Genesis block. And then since smart contracts on Ethereum. And now we at the state we're so lucky that we have the rails already. We just need to assemble them you know the tools correctly which is what we at Ampli have been doing for the past year. Yeah. I am seeing lots and lots of startups in the whole agent space. Yeah. Non crypto as well as crypto like non like off chain and on chain both like I'm sure you're following the rise of open claw and the ability to install and run in your own sort of a local setting in agent. What do you think according to you guys is going to come next in this whole ecosystem? - So yeah, I think that OpenClaw is a great example because last year when I was speaking at conferences of talking and networking events to people, I had to do a lot of education on what the agents could do. Many people didn't follow the fact that these agents and DeFi were already transacting like billions in dollars of volume. It's actually pretty amazing. And they were like, okay, I mean, they saw panels where people were sitting there saying, "In three years from now, I think you will have them start moving money around." Where I was sitting and was like, "No, this is already happening. You just don't know." But then OpenClaw came around. And it opened the eyes of many to the capabilities of the agents. And suddenly everybody was like, "Of course, these things are going to transact money." Very soon. - Yeah. - And we need guardrails because at the same time, OpenClaw, if you've used it, you have this moment where it opens your eyes and you feel the underlying power and potential. And at the same time, you feel the chaos and the uncertainty and unreliability of it all. This is where you realize, oh, we also need the guardrail too. So how can we control this in a way? So for us, it was a really, really good moment. And now, what I think is going to happen next is that we will have a differentiation where you have more and more purpose-built agents. Just for one task that they can do really, really well. And they're trained on specifically on data sets or maybe there's even this area between agentic and deterministic where you might use AI to build deterministic systems to interact with markets. Because LMs are arguably not the best in running numbers. They are language trained. They are more like models that guest the next token. - Yeah. - And in markets, you want to depend on deterministic outcomes. Where possible. If it gets to the execution side of things. So I think that's going to be built out more. You have very, very purpose-built agents. And then my vision would be on a platform like Lack Outs where front manager comes, they connect to it. They can find all these agents on a marketplace that we can benchmark them very transparently. We can show which one is best for which asset, which one is best for which training operation that you might run or which strategy. And then in the future, I believe it's going to be even an agent coming and subcontracting these agents. A more general agent like in LM, maybe you are the next more professional version of open clause might be running at a company and then we'll have an operator. And that agent translate the operator's intent and then hires and these other agents to fulfill with the tasks. I think that's going to be where it's going. - Oh, I completely believe that there's going to be an exponential rise in the agent to agent number of transactions. It's already on the app. - Absolutely, that's actually also what companies in crypto and DeFi are noticing right now. After October 10th with the big Binance Oracle Failure, human users noticeably vanished from the system, not only from Binance, like I've talked to some companies right now and they started to realize the next wave of users is not going to be these humans coming back. It's going to be agents transacting on these humans behalf. And that's how I think optimizing for to get these agents access and at the same time to make that secure. That is a great mission to be on right now. - Yeah, yeah, definitely. It's like it feels like we will have on one side institutions who manage money for people and those institutions will then deploy using agents that money because it's getting ever so complicated. Already even before agents, most of the transactions done on the markets were placed by algorithms because humans are just not as fast. You know, you're talking about milliseconds on that 98% something like that amount were done by algorithms. Now we're just automating those algorithms even further by allowing agents to be the signer of the transaction. - Holy, that's a very good point. And it's actually both in traditional finance, like ever since the advent of Renaissance technologies, you have the high frequency trading. - Yeah, that's right. - I think algorithmic trading dominating everything. And even DeFi, like Visa put out a study in 2024, which is more of like a bit of a hit piece, I would say, to stablecoins because they hadn't really accepted that they didn't go out of their adversary or role yet. So they were saying in reality, 90% of the stable-contrast actions on Ethereum are done by bots. Which is like MEV and algorithm trading mostly. - Absolutely. - And then maybe some was trading. But like, now I would look at that and would say, "Great, that's actually another proof that these rails are great for bots and machines transacting value." So now that we upgrade these bots to agents, they can actually do the whole unlocking of the blockchain and DeFi UX on our behalf. - Absolutely. - That's really great. Eventually you could have NeoBanks that have a personalized agent for every user that optimizes across different markets and productizes that into an interest rate. - Yeah. - And that was actually, you know, back to our starting point, our friends and family putting money into these systems didn't happen last year. Maybe won't happen this year. But I think that's where we are going. Once like, fund managers and DeFi have proven that these systems work at scale, we will be able to upgrade phones so everybody has a banker in their phone. - Yeah. So you mean in a way we can have like, I don't know, worlds and agents are managing that world. And then we potentially don't even need that, but you could just choose a world in a DeFi environment. - That's definitely a use case. - I think like vaults and agents have their merit in the different positions in the value flow. - Yeah. - You could totally have vaults that are managed by agents and there are several reasons of why that could be desirable. Like vault chairs have this, like let's say regulatory ambiguity right now. There's not too much focus on them, but you could argue that a vault chair when once a human curator manages them is a security because it wouldn't pass the how it is. So having agent managing them already changed situation slightly, I'm not quite sure how the regulatory data is on which side they would fall. But I would argue that the closer you get to the end user, like for example, a new bank situation, you would want to avoid that the funds of the users are co-mingled and you would avoid that management layer so close to your organization, so close to your users. That's where the self-custodial system where every user has their own agent. And every agent is not, there's not like a general thing, like a general approach, but every agent will do what specifically is best for that user's amount of capital. That's where you get out of that regulatory scrutiny and you can say, if it ends up in a vault on stream, that's a different story. It's not my responsibility as much anymore. So I think we can play around with that and find setups that work in different jurisdictions, for example. It's really good that we have both elements now, right? Let's say the agents, they are one layer above the vaults and they can manage between them and then they can also mitigate the risk of the worlds, which happened during the recent resolve, exploit and defile. I think I mentioned earlier that two of our agent partners, their agents manage to pull out of the contaminated vaults and markets super quickly. Like before any human could, all of the funds that the agents managed were safe, which is a great value proposition in my opinion. - Yeah, yeah, absolutely. What are some other trends you are seeing in DeFi? How do you think DeFi would look this time next year? - Very interesting. So I have this thesis that right now, we are bottoming out of DeFi's first cycle and we are seeing kind of a post.com crash situation. Where you had the first market cycle, then it breaks down, it accumulates, it breaks out and then you have another very traumatic event. And I think that traumatic event was the whole situation with Kepler and Arve right now. And if you look at it very closely and you're like sensitive to history, you can even find parallels in how the events unfold. Like right now with DeFi United and the whole bailout, that reminds me of the newspaper article that is embedded in the Bitcoin Genesis block in the like Chancellor approved bailout for banks. Arve just got bailed out. - Yeah. - But not by banks, but by the whole ecosystem and their partners, because everybody has a vested interest in dense and dense seeding and DeFi surviving. So even that lines up and I'm getting a little bit of goosebumps, lines up with the historical performance of the charts at the moment. So what would happen next would be it starts a secular bull run that lasts a decade at least. And I think there's a, there's confidence there between what the charts look like, what events aren't folding right now and how that technology fits into what is happening. Like Larry Fink from BlackRock, CEO of BlackRock. He's been running around since like 2023, telling everybody that they're going to tokenize every financial asset, every stock, every bond. So how are these assets going to be transacted? Well, they're going to be transacted on the infrastructure that's being tested right now. Very likely on a mixture of proprietary chains like Kanton and settling on Ethereum for also a lot of the transaction value, because Ethereum is the only combination I know of right now that combines the necessary provenance of running for this amount of time, never going down if we discount the Dow Hacker rollback back in the day. And also, I see that Ethereum has the potential to provide the economic security necessary, right? You need to have a sufficient amount of value on your system in order to secure the huge amount of value that the tokenization would require or would cause, right? Like you have a huge amount, if you have all the RWA's, somewhere tokenized, you need to secure them in a way with the economic value of the chain. So I think Ethereum is the very natural fit for that, plus some proprietary solutions, because we know that-- Like Kanton and-- Yeah, like Kanton, like some privacy-focused L2s. Yeah. So yeah, I think there is a very, very strong thesis, maybe not next year, but we are coming out of that spec-breaking event that we just had. We had a long accumulation, like three years of sideways, like as a TA person, that would be a setup that you would love. You know, just being agnostic of the asset class that you're in, it's like three years accumulation, breakout, a very traumatic retest that then doesn't break the lows, let's see. So far, it's looking good. And then you would see how DeFi just turns into FI. Yes, it is. I do think so. It is eventually going to happen. We don't know how it is, because-- But every year, for the last many years, I see more institutional adoption. I have never seen less institutional adoption. It does not matter. Hacks, no hacks. Down to and up to-- So that's why I'm bullish about DeFi becoming just FI. Just like, try to find DeFi merging together. That's right, look at that. I just got back from Bitcoin Vegas. And 80% of the people there were in suits. They were all from different funds looking at deploying and part of their portfolio in Bitcoin. But it is kind of like the gateway to introduction to blockchain for them. You know, once they start in Bitcoin, they know, yeah. Totally. They're okay. This is how wallets work. This is how walls work. This is how, you know, DATs work. They get introduced to a whole lot of products. And then they realize, okay, cool, we can use this to borrow money, lend money, do all sorts of things. And so-- And then that means that's a very positive use case for you guys, because that means once things are on chain, you can have agents interacting with them and all that. Exactly. And I think you see the upgrade of finance is ongoing. It's not like one flip of a switch. And now already, I was in the event on Monday, and did it as a deal summit. And on the panel, they were talking about RWO Teflonization and that upgrade on the markets where they already see, okay, we're not going to have just working hours on stocks exchanges anymore. It's going to be $21.7. And we were scratching their head. Okay, how extreme is going to be the cost of personnel of stucking up all the teams for global 24/7 operations? Yes. You're going to have to trade-- You know, some of them already trading on the other markets at the time, obviously. But it's going to be different when we have unified markets. They just don't close anymore. And this is where the shift to agents, while agentic and the eye capabilities rise, is going to be feel very natural, I think. Yes, it is. You can cut that-- I think in treasuries, like a more than 70% of the cost of treasury operations is in a team and personnel. Yeah. So if you can augment one operator of your treasury with an agent, he inherits some of the agent's superpowers, and you channel that in a way that's secure, which is what we are building for, then you can keep your treasury cost low, despite those, like, I would say, increasing challenges. Yeah. Absolutely. Who do you think is doing an excellent job in DeFi and crypto at the moment? Like, which other companies do you look up to? Oh, yeah. So there are a couple, actually. One of-- since starting to go to conferences, like, starting-- You're trying to build Ampli. I met you run into the same people, right? Yes. All across the world. It's actually an interesting concept. And one of my first conferences I met the guys from Meadas, their tokenization company out of Germany and Britain, and their CEO, he's a belief ex-Coldman. And they took this extremely focused and smart approach to RWA tokenization, where they decided to ignore the US market. And they went to Europe and went to the toughest regulator, which is the Baafin in Germany. Yeah. And they worked with them, and they made it work. And then they scaled-- I don't have the latest numbers. But I know that at some point, they scaled to 700 million in AUM. And a lot of that was BlackRock as a client. So they did this so focused and purposeful, and by the books, which obviously you need to do in order to attract that kind of client. That was always admiring how they were going about things. And then in the self-custody wallet space, we've been interacting a lot with a company called Fortify out of Israel. And they are, in my opinion, the perfect blend of offering self-custody the most secure way while giving access to all the important defy protocols and it's actually where we drew a lot of inspiration from because when we were starting to engineer our security, we were asking the question, how do fund managers right now deploy into defy? And they deployed through solutions like Fortify where they have their self-custody secured by something that's called an MPC wallet and multi-party computing, where the key is charred. And it is pretty much the standard. Like many funds have mandates that would require that. So what we drew from that was we should secure execution by agents with the same technology. And now, on some conference earlier this year, I was talking to a team that's very close to Canton. And mid-conversation, they told me, by the way, hadn't you built on this standard and this MPC standard we wouldn't be talking because 90% of the institutions have mandates that would require that. So it was flat very validated. And one of the reason was that we were watching very closely what these great companies are doing. Yeah, that's fantastic. Where do you see Emily in the next two years or so? So we just launched our first product, which is called the agent control room. And it's how we allow fund managers to discover and deploy and benchmark and control these different agents. And we are already partnered with the leading agent providers in the market right now. And what I hope to achieve is that we now get a flywheel going where more and more agents are going to be live on the marketplace. And more and more capital gets attracted by all those different capabilities that we can start to offer. Where it goes beyond just managing lending market positions, it could go to way more complex strategies, carry trades, things that you could do on hyper liquid, for example, where you have a very, very huge mix of different assets that you could interact with. So like getting that flywheel going and growing, that's like one of them the main things. And then I just, I think that reaching as many fund managers as possible would be the first goal. Because for them, we have a very, very clear value proposition. They're already in DeFi, so they know what they got themselves into. They accept it as risk. But at the same time, they are very open to mitigating some of those risks, if possible. Absolutely, yes. And the agents can do that. I would say that oftentimes you would pitch this whole idea of agents, what we are doing, et cetera, and so we would say, but what kind of yield are you getting? Yeah. And I would tell them, well, gladly, now we are in cybersecurity for agents, so we're not selling the yield. Depends on your risk appetite, right? Same as if you're doing it manually. So the agent might be able to squeeze out a bit more yield on certain strategies, because it's always looking, it's going to act way more and be more agile than a human. At the same time, I think that's not the main value proposition. The main value proposition would be it mitigates a lot of the risks that come with being unshamed. Because it's just there. It can process way more data in parallel than us humans, even with the best dashboards could. Yes. So I think that's the main thing. And then that allows you to potentially go further out of the risk curve and increase your yield by other means. But the main value proposition is the agent is a superhuman risk manager. And we at AMP, we make that usable at an institutional scale because we mitigate the agent's risk. Because what you don't want is that the agent's permissions to your assets are stored on some server in some data bank. And then, you know, one year and a half, into it, the North Koreans come and social engineer that team and then everything's gone. Because they had access to the agent and they had access to the permissions, so they could move the wallet. So what we do is we introduce separation of power, so that draws a bit from our background in like political science. - But it's serious, I don't know to say that. - Where we make sure that the agent never holds the keys. And that's already by just on a governance level, but just introducing us as an extra actor in this constellation, we're already making it way, way more difficult to be attacked. Because suddenly the attackers would have to compromise the agent provider to have the agent send different signals and then they would have to compromise us and they would have to compromise the client. So suddenly it gets way more difficult because our system stays self-custodial. The client can secure the upgrade of the policies which are written on chain using the NPCs. So obviously you'll have to keyshodding coming in, et cetera. So I think that's actually how you manage to then get very, very scalable systems. - Yeah, absolutely. Yeah, so I was just thinking like, you know, there are lots of others and lots of people building in this whole space around AI and agents and so on. What is the biggest challenge you think that they have building a startup in this space? - Well, it's a very, very good question because I was going around this conference and I think it's the same, the first time I've been to many over the past year and some, - Yes. - This year already. And it's the first time actually on the flight here, I rewatched the big shot. You know that? - Yeah, I know, I know you. - And then when like Mr. Baume, he flies to the real estate conferences and he runs, it's a bubble, it's a bubble. - I know. - So, sorry. I'm not feeling that yet, but definitely a lot of the stuff I hadn't talked last year or that were in different iterations of a pitch deck. Now you see them everywhere here. - Everywhere, it's everywhere. - It's everyone. - Everything is like, - Less so here. - Occasrating autonomous intelligence. It's like. - Each Denver was 30% just open-cloth. At least 30% just open-cloth, open-cloth, hackathons, open-clothes, demo-day open-clothes that all day. - Yeah. - As we discussed earlier, but that was a spillover moment. - Yeah, yeah, that was a spillover moment. - So I think one of the challenges is how do you find a voice and get cut through that noise? - Yeah. - And then, but then also, and this is something that crypto companies have not been historically very good at, is how do you find a business model that makes sense? And you know, you can, okay, maybe you'll do a token, right? And then the token is your product. But that's also distracting you from what you, from the value that you actually want to deliver. Unless, I mean, I've worked on tokenization and I'm very nerdy about it and I love it. But there are very few examples where there's a good interlock between the product and the value and how the token reflects that, right? - Yeah. - Yeah, we're actually having like this different conversation, but we're having interesting developments in that front as well. But right now, I would say, if you wanna provide value and you wanna grow your company, you gotta find a business model that makes sense. And I think that's something that many companies that don't spend enough time on. It's something that I actually had headaches over at like building Amplie and figuring that out. And one of the key, like technological breakthroughs that we achieved, at least from my perspective, is that we're not only in the business of making things most secure, but we also in the business of distributing the technology and turning it into a product and monetizing it. So we found a way of how to keep the principle funds always self-pustal deal, but be able to take fees on the delivered yield, like take performance fees. And then also do that in a way where every stakeholder and the system can get a piece of those fees very seamlessly with every action that the agent takes. So it's not just happening when let's say the user eventually withdraws, which is not a great solution, right? Like the better your product is, the later you earn any money, right? Because they might not ever withdraw. They just have it running. So finding a technical solution to how you take fees was very, very important thing to figure out. And yeah, I think business model and how do I figure out product market fits and find an audience and do something useful for consumers, offer businesses, whoever clients you choose, that is something that is underrepresented. - That is so, so true. So you know, in crypto or blockchain, all this whole industry, we have very few things that have a product market fit. One of that product market fit is, these are the ones that I have found. Please feel free to share and chime in if you found. Okay, these are the ones that I have found. One of that is transactions. So transferring value from one human to another, stablecoins are great example, millions of people around the world use it. - The best, I see the best. - And their volume is just exponentially growing. The number of transactions, the amount of value transferred every single day. It's now reaching nearly swift, which is a, which took decades and decades to reach this point. And stablecoins have done it in under a decade, you know, in half a decade, what, so it took 50 years. So great product market fit. A lot of DeFi, it comes down to yield. People are there for the yield. If the yield is zero, are people gonna take part in DeFi? I don't think so. I could be wrong, but I don't think so. So the yield is the product. Then there is the whole gambling aspect, which is, I would include Polymarket, all those type of things in that. Like, because people want gamble on trading, or, you know, buying low selling high, whatever, like, you know, leverage trading, all that, all that's gambling. I consider a lot of that retail side of things. There's the gambling. And then the last one is the investing for capital gains, whether it's in Bitcoin or by institutional capital or by other things. Apart from this, I'm yet to find another use case that's got really good, strong product market fit with billions of dollars transferring every single day and tens of millions, not hundreds of millions served, like stable clients, every single week. I think your list is very good. What we could add would be front loading capital liquidity, pretty much what the ICO did. We, I think, over time after the ICO we found like different iterations of that use case could have been very good at that, like making raising capital easier and also, yeah. - Yeah, don't get like, DGE's have helped a lot of startups raise capital for the better and for the worse. - Yeah, but it is getting harder now, that model is kind of showing cracks. And a lot of VCs are no longer in the space that used to be there two years ago. That's partly because the L1, CYL2s protocols, their TGEs haven't returned them a good return. So they couldn't no longer exit and invest in the next generation of startups. - But I think there's going to be future iterations of this. - Yes. - I think last year was when, when Sailor, he went on a podcast with Balaji and he proposed this like ultra-financerization of all US businesses. And if any site, he basically proposed to tokenize everybody's business, small business. And then have this, or G of investing in each other I guess. But I could see how something like that could eventually manifest. And again, it would be an interesting moment of liquidity unlock for many people, again, for the better and for the worse probably. But I think that that use case can still develop further and take different forms. Like what would you say to that? - Yeah, I give you my two cents on that idea. - Sure, I actually am. - Yeah, so there is a very interesting idea. But you know, whenever someone has tried to tokenize either real estate or physical business, it is a very cumbersome process just because of legal reasons. And every country wants to protect the businesses that holds their real estate, the land. Sometimes it's tied to residency with all sorts of things. The easiest assets to tokenize are financial assets because they're already in a digital format somewhere in some database. - I agree. - So that's why stablecoins is the prime example of the United US Treasury being tokenized. And now it's 300 billion or something market cap off. - Totally. - Any private equity would be the first one to be tokenized because it's already on paper. It's already liquid in a way and it's not physical. - And it doesn't get so disorganized. I mean, imagine talking to really small businesses, how do you liquidate that? Like and how do you prove provenance and-- - And prove stock levels and everything. - Yeah, exactly. - Very good point. - Very good point. - And back then. - I don't think Sailor answered that during that podcast. - Yes. - It was more of a big concept. But getting back to the stable points, I think because you're also getting back to it, that's such a powerful concept, right? And we already saw how users, everywhere across the globe find different applications for this, you know, within their own circumstances, how users in the global South have been using them to slide step local inflation successfully. And now they're even graduated to being part of the USA's just geopolitics, right? - Yeah. - They decided to not go with central bank digital currencies, unlike Europe. And Trump said, let's let the private sector find a way. And what is happening is that they are, they don't need to find sovereign states to buy the government debt anymore. It's actually going to be the individual citizens of the other countries that prop up the dollar. And that is the dollarization through stablecoins. And the sense is on camera saying, the stablecoins are going to help us find more buyers for, and we can print more debt. And you combine that with the whole Venezuela Iran situation. And that's how they save the petrodollar and go into another round of a couple decades probably. - Yeah. - Yeah, the significance of this can't be understated. And then I'd also show them. - But then at the same time on the other side, there would be other things that would try and compete. Like in Europe, you had to use USDC, not USDT. So like different jurisdictions will come up with their own stablecoins they would want to use. So even though they might not be government held, like I mean, they might not be CDVCs, or CVDC, central bank digital currencies. They would be other, they are denominated in Yen or Ramimbi or whatever other currencies. - That becomes super important when you look at payroll, for example, right, using businesses. You don't want that sidestep into the USD in that on your balance sheet. You know, if X-trade and every employee salary, that would be pretty bad. So I'm actually, I know a couple of companies in Europe that work on that, like all the Unity, for example. They probably use USDC. - Yeah, circle has their own a Euro stablecoin there. Small amount of adoption. And then all Unity, I think they're backed by Galaxy. They've also working on their own Euro stablecoin. And it becomes important in these commercial situations. But at the same time, I know the TEDA is also finding ways that they're not in my car, right? They're finding ways on how to get back into the market there. - Yeah. - So it's continue to be interesting. And I think that whole use case also, because we were like, what other use case and market fits do we have for blockchain? Or maybe what could occur? I think that with proof of work and proof of stake, we have very interesting new primitives, right? That we could mix and match in different technologies and different points of value in the future. And I think that as we become more and more eye-native and more and more globalized, eventually, there is going to be another cycle of stablecoins that are not backed by the dollar and that are not backed by a sovereign currency at all. I think we had that phase in 2022, 21 with Terraluna, which obviously was not the fit. But there are other stablecoins out there. And I believe that if you extrapolate AI development, eventually, maybe I'm going a bit too far in the siphon. - No, no, go ahead, go ahead. - Eventually, if you assume a sovereign entity that's AI, that's not human. Is it going to be interested in transacting value in the currency that's backed by human wars? I don't think so. I don't think it would be interested in using the dollar. So using something that's based on math and secured by blockchains that it might itself be able to run the validators for. That sounds like a much more interesting proposition for such an entity that's not human. So I think that there might be a cycle eventually when we find other ways of transacting value, like units of account, that are not necessarily backed by sovereign states. And I think that's very interesting. Maybe Bitcoin can also play a role in that. Bitcoin is not great money, we know that, but Bitcoin could be very collateral. - It is, it is fantastic. It's a great sort of value. There's nothing that comes close to it. So portable, mobile, decentralized, so good in every other way. But it feels like from what you are saying, that we are pretty much going back to the credit system that was before buttered, like everything started with credit system. People didn't do butter. In fact, they just remembered that you gave me something and I will give you something back in return. So that was how butter was only for enemies because you didn't trust them. There was no trust that you will run away. You'll never give me back something. So you buttered at that point. Otherwise in the communities, inside the communities, it started with credit that you gave me five bushels of wheat. So you can have my card would something in future when I have it ready. So that's how it worked or when the crop was ready and stuff. And people just built it based on trust. Now we have blockchain for trust and all that. I mean, and then we moved to electronic database and credit system and now we are going to decentralize database and credit system. And then we make all these instruments on top of underlying that. Yeah, and we've been solving the trust issues with like, you know, that's one of the main things that blockchain is built for, right? Yes. We don't need to trust the millmen anymore and we can facilitate trust action between two parties that don't necessarily. Exactly. And I think that a unit of account that works in that on that technology could make sense as an exiteration. Not right now because we are in the adoption cycle of the government backed stablecoins right now, mainly the USD, but I think that could be interesting. And then also we can also, as I said, like proof of work, proof of stake, merge mining, we can actually back value with useful activity, right? Like there's a new proof of work chain actually that recently launched called Pearl that we at Amtie watch very, very closely. And it's actually merge mining with compute. So again, if you think about like, what is the most valuable thing in the next decades, it's very, very likely going to be energy to spend on compute. So if we can tie that to currency and value exchange, and maybe make even that more liquid by doing that, I think that's also a very interesting use case. That might not be very fleshed out and explored yet. It's happening right now, but I am watching the break. So if there is two questions, if there is a young, sorry, this has gone for, because I'm really enjoying this. This is great. It's great. I think. This is a lot deeper topics. And then the general sort of interview. But there's two things. One of that is if there is a young person wanting to study, what should they study? What should they read, acquire knowledge? And then the other side on the founder, what should a founder look at building? Because this comes back to your article that he wrote on X, some time ago, Twitter, on the downturn of SAS and how. I mean, that is very well known. Like you can look at the stock price of Figma or any SAS product. Some of the other stocks that I had, like zero, which is accounting software half of what it was back in its heyday. Everything that's perceived as a dashboard is-- Every dashboard is down pretty much. So it's like, what should a founder study-- I mean, build and what should a student study? Yeah. Those are like super tough questions. I was actually talking to the daughter of a friend recently. We're just about that because I study political science and philosophy and a little bit of law. And I ended up being a tech role, I guess. Yeah. And that's not a very linear path. And I think that's going to continue to be the case. The paths are not going to be linear. And one thing that's-- I love the smart AI influence, as we tell you, is use your domain expertise and then translate that and leverage the new tech technology. And I think that's for a while that is going to be the case. And I also think that studying computer sciences and learning how to code, despite bipocoding, learning architecture of coding is not going to be obsolete. That is quite the opposite. It's actually the most leveraged people are developers right now. Or CTO, he always was a beast, but now he's unstoppable. And I think that's something that might be something that's where people have a misconception, indeed, that they think that when working in mode, there's actually a situation where companies they tend to increase the budget of the departments that get more production. They don't like-- Reviews and owls. So it's like, oh no, every dollar I put into this or into this person turns into more dogs. So I should give them more dollars and find more people like him or her and I think that's something that somebody that's young and trying to find orientation right now could take into account. Did that answer both questions? That is such a good point. So that's on the student side and on the building side as well. What should they build? Oftentimes, it's really like about listening to whom you think your clients could be and taking their needs seriously. Like we did surveys, we went to the conferences, we talked to all the different service providers and what they do and we talked to the fund managers, we talked to the neo banks and financial people. I even went to Money, Life and London and there were most big banks and regional banks. Some of these people couldn't get away from me quickly enough. Because the regional banks, they are the concept of risk regarding AI is, let's do a pilot where we have an AI agent making this part of our back office a little bit more fissure. And even just an agent that would be a chatbot, that would be perceived at high as high risk because it's touching the end customer. For them putting money into the hands of an agent was still very far out there. For me that was valuable information. I know the path will be the fund managers and then there and eventually I can talk to them. So I think taking in a lot of information and taking the need seriously and trying to build towards that, that's the best thing you can do as a founder. Fantastic. And what's your biggest headache right now as a founder? Biggest headache as a founder. It is, I think everybody's biggest headache is that we are in a time of acceleration. We are in a parabola when it comes to technology but we are also in a time of social, terminal and weakening of institutions. And that means that there's also weakening of trust structures. And we need trust and reliability to do business. So it's always, you're always feeling like you're moving on like moving sand, right? Yes. The standing on moving sand. And that's something where we are trying to extrapolate into the future and understand what's going to happen next. Like right now for example, we are fundraising in our pre-seed round and it's very, very difficult to figure out if waiting is best, if doing it now is best, what's going to be the sentiment in half a year. Right now I'm going around the conference and I'm feeling, well, this is like bubbly enough so that we are building something that could be perceived as very valuable right now. Half a year from now, that could be even more extreme or something drastic similar to how open-clock came out, some spillover moment could have happened where the sentiment has changed. It's very, very hard to tell. So you get to play it by ear and that's daunting sometimes. Yeah, it's like some other VC story that you know, just get money whenever you can as fast as you can. That is very, very good advice. But you also don't want to be random with who's money you take. Absolutely. Right, one of these strategic about it, etc. And you also want to project power and strength and go into these situations from a position of strength. Absolutely. You want to get much better terms. You want them to use a crypto firm, a term formal for you. Yes. And give them good reason. And so I think being strategic about that is important while always be raising, yes, that is the thing. Yeah, always be raising. That's what VC's do. They're always raising from LPs. It's not their money. So at the same time, I would also argue for always be looking for how to get profitable. Yeah. You don't have to sell so much of your company if you make a lot of money. Exactly. And then if eventually, if you're profitable, they'll change you. They'll find you. They will not even have to reach out to anyone. And surely maybe an offer comes that you want to take. But the more you own of your company at that point, the better. Absolutely. Right. So yeah, that's also something that I think is a bit underrepresented in the, there's this whole like bubble of just VC cycles, right? Yes. And but I think your goals should still be to deliver value and to just make money. And the strong, getting back to crypto, the strong tokens we see are the ones where you couple some way of reflecting value back into a token with an extremely revenue driven and successful business model. Binance, hyperliquid, right? Like those are the tokens that that that drive or hold up better than others because they couple a successful business with a lot of revenue with some mechanism to reflect it back into the token. Yeah. The burn might not be the best, but yeah. That's, that's fantastic. So with all that, what's your ask, are you guys looking for VCs, hiring, partnerships, anything, feel free to share? Yeah. So, and also second part of that question, where do people find you connect with you guys? Totally. So, in terms of hiring, we are still in the, the phase where we're trying to stay lean. Eventually, we're going to be looking at like growth and sales. Now that the products are running out, that's always something that's, that's important, especially if you are working in the B2B space. We are currently raising a pre-sea browned and we are welcoming investors that are strategic in terms of like, ideally, we want investors that are interested in using our products to match capital or that have strong connections to clients because that growth angle is really important for us. And then we're also still, still like looking for clients that are potential clients that are interested in testing out the agent control room and want to understand really what could an agent do here in my setup with the assets I manage and the risk I am managing. This is for our audience actually, do you have to be an accredited investor to try out that agent room or can it be an individual? Actually, we are currently looking at funds using it and high-network and network and individuals. But there's this like path to, to adoption where I think that needs to happen first before you can surface the risk to retail. Yeah. Very likely that's going to happen to distribution partners like new banks. What you can do today though is you can connect to Giza, to ZFAI, to sale. They're all amazing, amazing teams building agents and they're active in B5. So if you already have a crypto wallet and you have, for example, stablecoins, you can connect to them and work with them right now. What we did is we upgraded the whole security stack to the level we believe that institutions would require to do that for hundreds of thousands of people eventually. And to feel good about that situation. So yeah, that's- So new banks would be the primary people you would love to connect with here. Yeah, that's like the path. Like Emily started out wanting to build a new bank based on agents and then we realized this is too early. We need to build the security stack and maybe us for at the time with our capital and our reach we're not going to rival Revolut next year. So we realized that the mode is going to be in being that infrastructure partner to a new banks in the future. And I think there's a path there over the next two years where it's going to happen. That's fantastic. Thank you. Thank you so much for your time and what I'm going to do is going to leave links to everything that you guys are doing at the show notes so people can read more. So don't just say that you're out. Yeah, that's fine. I'll put it underneath on the screen as well. So thank you so much for your time. Sandra was great. Thank you so much. That's it folks for today. And if you got value out of this conversation, you enjoyed it. Then do not forget to share it with your crew and leave a comment. Like and subscribe so you do not miss the next episode because we will be sharing a lot more insights and industry secrets as always and also nothing that we talk about here should be taken as investment or financial advice. If you'd like to be a guest on the show or if you're building something really interesting then I want to hear from you. Leave it in the comments. Reach out to me. You will find all the details in the show notes description or just search for Sam Kamani. I look forward to seeing you again in the next episode. Peace.