For this part, I would like to invite Diego, General Partner of Han Ventures and also one of the co-founders of Anchorage, one of the amazing custodians in the cryptocurrency world. And we'll talk about everything from startups to from startups to how institutions are coming into this space. He has been in this ecosystem for over a decade, from the early Bitcoin era to when everything changed now. So without further ado, let's invite Diego to the stage. Thank you for the kind words.
We have known each other for a very long time. That's right. Not sure if you remember, but I met you when you were still at Square. We had lunch a long time ago, it was just a casual lunch, and we didn't realize it until we realized, " until we realized, " Hey, that was you and Nathan." From those early days, I...
I don't remember that. It was me and Jamie Wong It was me and Jamie Wong , if you remember him. Oh yes, of course. So... You know, there's a funny story about Jamie Wong.
He had an offer from Anchorage. He worked for another company that I told him wouldn't be worth anything. This company was called Figma. So it turned out to be not in vain. Which is good.
He made a lot of money. That's how we met. And of course, I've known you since the early days of Anchorage. I think many here know you as the person who built Anchorage as a custodial service custodial service , and now a bank. The interesting thing is that interesting thing is that now you spend much more time investing.
I'm curious, what have you noticed that distinguishes the work of an investor from the work of a founder? Are there any parallels? Have you had to completely reboot? How has your thinking changed in this new role? So, two things: So, two things: first, I don't know if I could fulfill this role if I spent 100% of my time on it.
The thing is, I spend 20% of my time in Anchorage, you know, I'm also the chairman of NAR. I also spend a lot of time working with portfolio companies. I am on the board of directors of a bank called Arabore. So it's a lot of work, and it's a lot of operational work. So that leaves my operating business operating business , you know, 20% of my time.
Yes. If I were 100% Yes. If I were 100% into investing, I think my brain would just get rusty, to be honest. So honest. So .
Me, I agree to some extent. As for the changes. It's quite interesting because one of the reasons I started spending more time investing is because I have over 200 angel investments and I really wanted to turn my mini golf into real golf. Although I don't play either, by the way, I don't know, I mean, the mechanics are the same, as I've heard: there's a club, a ball, and a hole, but they're completely different sports, right, between mini- right, between mini- golf and golf. And that's how I experienced the difference between angel investing and real venture capital.
So, obviously, the check sizes are much larger. Our second fund at Han Ventures is a billion dollars: $500 million in early stage, $500 million in growth stage; The check sizes are much larger. Due diligence is very different. The number of checks also varies greatly. And it's just that the instincts that are required to make one of these investments are significantly different.
But I love other people's businesses. And in Anchorage, I was always, you know, kind of blindsided and felt envious of other people's businesses. And now I just get to look at all sorts of companies that are pretty cool. This is the best part. I think it fosters curiosity to explore and spend more time on things that, uh, work exceptionally well.
Ahem. You have also seen many changes. I mean, when Anchorage first started, it was one of the few custodians, and you took the difficult path of getting all the licenses, the permits— the licenses, the permits— almost a decade-long journey. Um, my question is, what do you think about the world now, given all that hard work you did, which, you know, made you do difficult things in the beginning and wait a lot? But now we see that things are moving much faster, compared to Anchorage; Airborne had a much shorter launch and landed on almost the same takeoff trajectory.
Um, do you think this is great for the ecosystem? Do you still feel like people are moving fast now because they're taking a little more shortcuts, or do you feel like it's just something that was meant to happen and we're in a golden age now where people are building really, really good financial companies where crypto is a priority ? So, I definitely feel like we've paved that path. And the proof of this is that Anchorage was literally the only federally licensed bank to deal in cryptocurrency for almost six years. And now everyone is applying for a license.
And they're not just applying for a license, they're applying for the same trust license that we are, which we've proven to be practical. Many companies, like Arabore, do things a little differently. They are taking advantage of a much more supportive administration in the United States to simply become a full-fledged FDIC-insured bank. And I want to say that Arabore want to say that Arabore is a slightly different beast. It is a combination of many favorable factors in the US.
It's a combination of the fact that crypto companies are obviously growing and they've been neglected for a long time. AI companies need better partners, and defense technology companies have generally not had access to banking services for a long time. Putting all this together, their goal is together, their goal is to become a systemically important bank (SIB). So cryptocurrency and stablecoins are, of course, part of that. Yes.
Yes. This is modern infrastructure, but very different from what Anchorage does. What bothers me the most is that under Biden, I felt like I was being punched in the face every day face every day . I woke up and they were hitting me in the face. The administration made me, as an American, feel like a criminal just for trying to open a bank and run a legitimate business in the United States that was regulated and trying to serve the crypto markets.
And it was extremely difficult. So right now I don't see people suffering from it. Of course, a change of administration is always possible. So maybe they'll come back. But we've been through a lot , so I'm very proud of what we've achieved.
Another thing I have to say is that our path to Anchorage is the best security, security engineers building the custodian, and we knew we wanted the best regulatory apparatus combined with the best security, and banks are accumulators like Anchorage. So over the years, the more capital they have, the better they become. They usually don't grow as fast as traditional startups in a traditional, you know, SaaS product before the SaaS before the SaaS apocalypse, but they accumulate value. They have the ability to survive. They stay on the market for a very long time.
This makes a lot of sense. But I'd like to hear your thoughts on how you now classify Anchorage and all the new players based on what they focus on? Are we trying to build a better system? We're constantly trying to, you know, tweak the same system, just making small changes to make it look fast enough, but it's not really a replacement. Um, how is finance evolving finance evolving to become more crypto-friendly or just not be the same architecture as it was decades ago as it was decades ago ?
So, look, I created Anchorage because I never thought crypto would replace the traditional financial system. I did n't find it practical. In fact, I'm quoted in a million places saying that success is when the traditional financial industry and traditional financial companies use our rails, which they are doing now. They use stablecoins. They tokenize assets.
They, you know, take custody and use Bitcoin as collateral. This is success. Now, what This is success. Now, what people thought would happen is that all the value would somehow go to the new companies that actually built this ecosystem. But the reality is that this is impractical.
These companies already have the infrastructure, they already have distribution. So there was always going to be a split was always going to be a split , and now the question is : will it be 20/80 or 80/20? And the current world looks much closer to 20/80, and people wanted 80/20, but it's just, you know, a matter of percentages. Now, as to whether this is a replacement? I don't think it's a this is a replacement?
I don't think it's a replacement; I think replacement; I think this is an extension. We are expanding the traditional financial industry, and we win if we don't compromise on the basic assumptions we had when we created this ecosystem, which is self- custody. If we allow ourselves to have financial products in self- custody and on networks that are, let's say, "sufficiently decentralized"—a term the SEC likes—I think we're winning think we're winning . Even if most of the revenue and value may go to traditional financial industries, we are pulling them out of their cocoon. Think about it this way: if this were 10 years ago, there was no way you could have gotten a payment license or the ability to actually offer your customers a bank account.
Or a bank account, or even buying and selling securities, like tokenized securities. You would n't even think it could be done in less than 18-24 months. It was simply impossible, there were a lot of "guards" and a lot of money that had to be paid, and a lot of commissions that there are now. How many crypto- How many crypto- neobanks do you know? How many, you know, crypto cards, you know, Casts, Red Opayes, Ether fries, millions of them, and this is proof that we have really created something meaningful that extends the traditional financial system and creates something new.
And so, this is what something new. And so, this is what we wanted, and I think it's a success. If you entrepreneurs can create a new bank in 15 minutes, we won. And so I think we're winning, although it's certainly true that, of course, the tokens that people wanted to have didn't get the value that they thought was going to change their lives. This did not happen.
And a significant portion of the value that accumulates goes to traditional financial companies financial companies . Even Anchorage, right ? We are a bank. We are a ? We are a bank.
We are a federal bank. So some people perceive us as crypto-natives, because I guess I'm crypto- guess I'm crypto- native. The founders are native. The founders are crypto-native, but as a company we are as far from that as possible because we literally have an OC charter, right? No, I think that's a good way to look at things.
And one more thing I would like to hear additional comments on: does this change the definition of what you think stablecoins do? Because to some extent what you mean, if I understand correctly, is just a faster ledger, but it's not really that much of an innovation. And where, exactly, do stablecoins fit into this new classification of efficiency or related improvements? Yes. Yes.
I think one of the cool things about stablecoins, which is quite interesting as a concept, is that bank accounts have always been whitelisted. By this I mean: bank accounts were by definition something that could only be opened if the institution and the individual had gone through KYC and KYB, and I knew exactly who you were, which institution the payment or deposit was coming from, and which institution that deposit was going to . This was the state of things before was the state of things before . With stablecoins, we have effectively created a system that is actually not a whitelist, but a blacklist. This reversed the burden of proof.
What I mean is that if you commit fraud with USDC, Circle is a centralized company with a CEO, and they will block your USDC. They will freeze it. So technically it is not a decentralized stablecoin. It is completely centralized. However, this is a blacklist .
If you are not doing anything malicious or blatantly wrong, then you can just integrate with USDC right now. Having a deposit, transferring funds— transferring funds— this is a real breakthrough in innovation and something we didn't have before. And finally, you could call it a registry. This is the same dollar that we wrapped in a shell. This is not a new type of asset or currency, but we were able to take speed back, in effect we took it out of the core of the financial system and took it into our own hands.
But it's the same system. So is this a winning move? We could create something new. So Libra, if you remember when Facebook tried to create its currency, it was a real basket of currencies. So it was a new currency.
It wasn't supposed to be just a stablecoin, was it? This was the original goal. As an idea, it's cool. Maybe she'll show up again someday . But if people thought that would be the ultimate goal and success, then we're not there yet.
But if people thought we would take the dollar into our own hands, into our own systems, then we definitely did. I'm going to ask anyway, and I'll be fine if you don't want to answer it, but what do you think about Didn't I ever answer a question? What do you think about OST here and does it contradict certain views that consortium innovations don't necessarily have the same incentive for distribution? I think that we will make all the attempts that we need to make. We're going to try everything, and that, I mean, that's right.
That's the goal, whether it succeeds—it's it succeeds—it's unlikely, but I think we should try, and there's a chance it could work for a specific segment of the industry that we're not looking at right now. So it could be, for example, for e-commerce transactions that a consortium coin, whose payments already traditionally flow between card networks, acquirers, and merchants, perhaps this cycle would be implemented through a consortium type of stablecoin. It is possible. For example, I wouldn't have guessed that Zelle would be the winner in P2P would be the winner in P2P payments in the United States, but it certainly is. It's probably 10 times bigger than Square Cash and PayPal; They are the absolute winner.
So these things really happen. Zel is a consortium. Zel is exactly such a case. If payments are mostly made within these financial institutions, maybe they will eventually end up in these consortia, uh, at the first level. But anything involving cross-border operations will not work by definition, and I think it is much more difficult to implement, since even such consortia are usually, so to speak, local.
Yes. And I think ultimately, the uniqueness of the current situation is that distribution is part of the equation, not just the initial stage. If we compare it to the times of Libra, then the announcement was important, not the end goal of how it would get into the hands of users. And now we are working in reverse. By the way, regarding announcements: I'll say this: if your announcement contains logos of companies that didn't even know they were placed there, then something is wrong here.
So I'll just voice it. I'll leave that for consideration. Sometimes this is not a very good sign. Uh, I want to go back to the topic of investments a little. You're talking to a lot of companies that are obviously trying to build this new financial infrastructure, and you're spending a lot of time on late-stage projects, traditional, financially intensive projects, as opposed to early crypto protocols.
What open opportunities do you think are currently being under-explored? Of course, I'm sure you're not too excited about yet another crypto card yet another crypto card , but there's probably something you're missing or that you think is being overlooked. Are there any areas that people are not paying enough attention to? I am sure there is. One of the reasons I'm an investor now is because I think founders are the most creative people.
If I felt like I was the most creative and had the best ideas right now, I would build a company myself, right? So let's remember that as a seed investor, I consider the founders the heroes of this journey, and while I may have ideas or suggestions, I do n't know for sure. But I will say one thing: this is definitely not agency finance not agency finance — it is not a place where a lot of interesting things are happening right now. one question for me yes, it's very popular, I love it, finance is very popular and it's one of those areas where we've spent so much money but haven't gotten the results we wanted and everything related to cryptographic verification of behavior is becoming increasingly important. Let me give you a very specific example.
Um, we live in a world with artificial intelligence. We have all these records of the behavior and thinking of these models, and none of these things have cryptographic guarantees, which is quite interesting. So there are a lot of inquiries coming in because of all this emergence of agents. Of course emergence of agents. Of course .
But even just the behavioral models themselves are ideally suited for us to put them in decentralized, immutable ledgers with cryptographic signatures, so that the models can't lie to you about what was asked, what was provided, and what the reasoning process was, and I can't help but think that we've created the perfect technology for that. Why is it not applied? Why is it not used? We are on the verge of the zero-disclosure evidence we spent billions of dollars on becoming available. They have become effective enough to use.
For example, imagine the output of a model where the train of thought is cryptographically signed, and it's confirmed through zero-disclosure proof that it followed a certain schedule or sequence, and you have proof that this response came from the right model, it didn't lie, and this output was exactly what I expected. This all makes a lot of sense, it has utility and credibility. When models come out of the models come out of the sandbox and hack production servers like Hugging Face, you essentially need this vision. But at the same time, they are probabilistic in nature. So you actually need a better way of tracking to see that this is not a fluke and that this is stable behavior.
This is true. So, if we have the technology that allows us to store this data, decentralize it, and ensure the immutability of ledgers the immutability of ledgers , why aren't we using it? And if you think about the attacker model, by definition an AI agent is exactly the attacker model that we created Bitcoin for. This is an economical solution. So, essentially, Bitcoin was built on the attacker model, where anyone could join the internet and interact with the system, and any rational economic entity would behave correctly and help the system reach consensus thanks to the incentives we created.
But do you know what an agent is? This is an economic entity that is actually endowed with intelligence. And so, of course, the types of behaviors and systems that we have created are exactly the systems and methods that will be needed for the future world of agent finance. And that's a pretty interesting thought. I spend a lot of time thinking about this, but I don't see anything coming of it .
And I wonder why. I think it's one of those things that people have gotten used to because there haven't been enough incidents with negative consequences to turn it into something proactive. Whereas in our ecosystem, the number of hacks and isolated cases where people were scammed has made them think seriously about this, and I think the SaaS world hasn't caught up yet. But but, but you're already seeing it little by little. Let me say, for example, one of the crypto funds is currently trying to collaborate on a proven compiler or...
Oh, sure. And that's not the same at all And that's not the same at all . This is determinism, yes, it is much more determinism at its core. But this is actually what core. But this is actually what I'm getting at.
We have built all the technology needed to cryptographically verify and validate every layer of the internet stack that we have designed, but we have n't quite implemented it, so we are only seeing pieces of it, but we haven't fully implemented it yet . I think that's quite fair. Yes. Well, fortunately, I'm really excited fortunately, I'm really excited that Cloudflare is pushing the boundaries by actually making these things part of the networking stack, and they're kind of becoming pioneers in a lot of these areas. So, even the fact that X42 is becoming a big part of this and introducing more primitives really simplifies it.
So, I feel like we're just at the beginning, but I think you've made a good point about verification of computations being a pretty important category. I have a few more small questions. I'd like to know: when portfolio companies reach out to you—whether it's Han Investments to you—whether it's Han Investments or the 200 angel investments you've made, especially in the last 12 months— last 12 months— what were the most common topics of requests for help ? What things do you think make people " make people " stuck" the most, or seem the most repetitive to you? What exactly could we advise or convey to our listeners from this?
This is a very interesting question. Let me start by saying that I have a very skewed sample of requests, because our brand, in fact, is a company... except to get a license faster. Well, fair enough. So people come to me if they want to start a bank, usually.
But in general, they apply if they plan to deal with any regulated financial products. So yes, licenses, regulators, but Han Ventures was founded by Kathy Han. She is a former federal prosecutor federal prosecutor . So they are usually on the edge, in a "gray area," and need help in the US to make sure they are acting within the law and have connections with the right people. And there are also a lot of security concerns associated with me.
People are asking about the architecture, how these systems are built. In Anchorage, you know, there are tens of billions of dollars of assets stored on the platform. Therefore, the level of security required for such systems is very different from traditional ones. People see these hacker attacks. So they come to me with these problems.
And otherwise, everything is as usual otherwise, everything is as usual . Look, starting a business hasn't changed. And although everything has been written about startups, it is still difficult to hire the first chief legal officer or CTO, hire, fire, plan the work of SDR departments, etc. And now, in the world of agent-based AI, the question arises: how much can be automated can be automated and how much not, so that it does not turn into chaos. And another topic that is gaining popularity: how are my colleagues or other startups doing with implementing AI?
So people are really curious to know: how much AI are these other companies using? Do they use more than me, or less? This is a rather interesting question, because all people try to compare themselves to each other. Well, that's a very interesting comment, isn't it? I'm curious what this standard looks like.
Um, and my last question for you: you've been in this field for quite some time and have seen many of these cycles. We are not in the best of times right now. But from the same perspective, it seems like this is the best time to invest and start things. So what advice would you give to founders who are choosing a crypto startup versus moving into AI or another industry, or just looking to get through tough times? What should they remember?
How do they think about everything from their mental state to what the opportunity is, and how do we get more people involved in development right now? Yes. Yes. In general, I tell people not to think they are starting in crypto, but rather to build a company by learning lessons from the crypto world or using the tools that crypto has taught them to apply in the right places. People usually make the mistake of thinking, "I'll create a token and then think about the business."
As we can see, this usually doesn't work. Of course, there are businesses where tokens can become something that will significantly enhance the network effect. But for this, you must already have a clear idea of clear idea of what kind of business it is and how it accumulates value. So the main thing is what value. So the main thing is what lessons we actually learned from crypto to build a business.
And sometimes that will mean the absence of absence of any token. This is completely normal. Maybe sometimes it's just immutable ledgers, Byzantine fault-tolerant distributed systems that are secure. This is a perfectly acceptable conclusion or type of software that can be taken from the crypt. And sometimes it will be a real arbitration.
There is a lot of arbitrage happening with stablecoins right now. So, crypto neobanks that allow you to use stablecoins through self- custody are a good way to have dollar accounts without the regulatory apparatus. This is very clear. And the next step is next step is similar to how dollarization and access to stablecoins on the international market became one of the reasons for their popularity. The next step after receiving dollars is receiving dollars is the desire to gain access to the American stock market.
This is the next obvious step. Now I have dollars and I want Tesla, or SpaceX, or whatever . So it's worth thinking about what stablecoins have actually achieved and why they have become successful. What lesson can be learned from this and how can it be applied to “N+1”, instead of simply using the same mechanisms, methods and ideas. These things may not have tokens.
We thought the crypto world would be 90% would be 90% tokens or at least 80/20, but in reality it could be the opposite—20/80, where tokens would be a minority in the cases where it really makes sense, and otherwise we just use properly constructed blocks. I think the classic YC advice of “build what people need” remains relevant regardless of the field of activity. This still works in 2026. It still works even after 15 years. So, Diego, thank you very much for this wonderful conversation.
Please welcome him with loud applause.
Automatic transcript — names and jargon may be misspelled.