I had turned $50 into $2,000, and by the end of that year I had turned that $2,000 into $100,000. Yeah. So, one of the things that you have to do to be a good investor is you have to live in the future. We're oftentimes investing in things before they have momentum. And I think that's kind of part and parcel of being a good VC is that you're not looking backwards, you're looking forwards.
Crypto venture capital may be dead by 2030. Which one of those is the marketing?
So, neither is marketing. I would say that we have a reputation for saying what we really think, and that is very much what we really think. I seem to be the boring person who's telling everyone this stuff is not going to work. At the time, I think it was popularized by Three Arrows, the idea of the super cycle, and like everybody was using this term, super cycle, super cycle. And I was like, there's no such thing as a super cycle. Assets go up, assets go down, like there will be another bear market. A lot of investors cannot think probabilistically. Very often in poker, you make the right decision and you lose the hand. Sometimes also you make the wrong decision and you win the hand. Your bad investments are never really where you lose the most money. Yeah, it's in the bad passes you lose the most money.
Haseeb Qureshi, managing partner at Dragonfly.
Hey guys, so welcome to the Onchain Capital podcast. We're here with Haseeb from Dragonfly. I've been very excited for this interview for a very long time. We spoke to other funds. They referred you a lot. Said you're a great mind, great investor, and uh today we're here and uh let's deep dive into it. So I think first and foremost, love to hear more about you. Just give me a quick intro about you, uh who you are and what you do.
Yeah, so I'm an investor at a fund called Dragonfly. We invest into early stage crypto and kind of the future of money and markets. Um invest all across the space. Been doing this for about almost like nine years. Uh before that I used to be uh a founder, then a software engineer, and then before that I was a professional poker player. So I've lived many different lives. Um and I I spend most of my time thinking about the ways in which the world is going to change. And uh I've often found myself early to that change, and that's where I feel most comfortable.
Amazing. Let's talk about Dragonfly because I think you downplayed it just a bit. Dragonfly is huge. How big is it at the moment?
So, we've got on the order of about 3.5 billion AUM. Uh we invest mostly into early stage projects. So, somewhere between $3 million checks and $30 million checks. Um we've backed all sorts of projects across the space. So we're early investors into Polymarket, investors into Lighter, investors into Avalanche of course, which is uh I think a seed at Avalanche uh many years ago, and um up and down the space uh you know we're seen investors in Kraken. We've done we've done things from from the exchange side uh we've done the prediction markets, we've done the protocols, we've done the uh the actual applications like uh prediction markets and and Perplexity. So we do everything.
Awesome. Um, for the listeners, tell me some of your wins with Dragonfly just to see the scale of it because you mentioned their Lighter, you mentioned Dragonfly, you mentioned Polymarket. What was the early wins for you?
The earliest wins. Um, well, so Avalanche was certainly one of them. Uh, I also seeded Near Protocol. Um, if I think about the Dragonfly days in the very beginning, actually the first check I ever wrote was I did the seed round of Matter Labs. Um then I did uh that was that ended up becoming zkSync. Um that was a very hot L2 at the time that L2s were very hot. Uh I think so there was um early round of Anchorage. Uh we did we led the Series A of Bybit which is now of course one of the largest exchanges in the world. Um, so in the early days of crypto, it was kind of hard not to end up landing in something that became enormous. And it's a little bit like I'm sure what AI companies must have felt like three, four years ago is that almost everything in that cohort ended up becoming important. The same thing was true of crypto back in 2017. But the delta is that I think we got a lot of the big questions right. Yeah. And the big questions, you know, back in 2018, you know, we did the Series A of Compound. We were early investors into MakerDAO. We did the seed round of 1inch, and realizing that DeFi was going to become really big was something that not every fund understood at that time. DeFi was still in its infancy. There was less than 100 million of total TVL. It was kind of like these little experiments in people's garages that they were playing around with small numbers, and we realized that all this stuff is going to get bigger.
You mentioned one thing right there that's really interesting. You mentioned about the questions you were answering, the questions that didn't arise or what people were thinking about. Right. Those questions turned out to be billion dollar bets. Can you tell me more about your thinking process around that? You know, because the other thing you mentioned at the very beginning was that you see things earlier than anyone else and those bets have turned into massive wins for you. Take me through that thought process.
Yeah. So, one of the things that you have to do to be a good investor is you have to live in the future. And this is one of the things that we talk about a lot when we're looking at investments, especially early stage investments, is a lot of times you know we're using a product and almost nobody else is. So if you think about something like uh for example Hermes agent, which is an AI agent that you know really in the early days of people who are using OpenClaw and Hermes agent, you know you might think like oh these things went super viral, but there's a very small number of users who are using this stuff in absolute terms. Um I remember when we did the seed round of Dune Analytics. Wow. So you don't have Dune Analytics. Um at that time again DeFi was tiny. There was almost no money in the space and so we were using Dune every single day to understand what was happening in the crypto markets and understanding what was happening on chain. Um their user base was maybe a few thousand people. But the reason why we did that deal, I remember it, was that the mantra that we coined in IC is that look we are living in the future and we have to believe that many more people are going to be doing what we're doing today in three to five years. And so if you believe that and if you also not just believe it but you make it true about yourself that you are constantly living in the future, all the things that people say is going to happen, you live as though that's happening to you today. And if you just notice what products do I prefer, what where do I end up wanting to spend my money, what are the things that actually are providing a lot of value to my life and what are things that are kind of implementation details, um that I think is a very good guide to being able to consistently find things that will be the next big trend. Um there are a lot of things that are different in that they're promulgated by VCs. So think about um a lot of the stuff about decentralized social media. Decentralized social media obviously never really worked but it was a concept that was more or less invented by VCs. It was people writing blog posts and doing podcasts and talking about wouldn't it be so cool if you know instead of Facebook, Facebook is terrible, they take all our data. Wouldn't it be great if we had a decentralized alternative and wanting something to be true and almost talking it into existence? Because of course the VCs make the idea. They kind of implicitly call for founders to do it. They fund the founders when the founders want to do it. And it can almost be easy to mistake that for living in the future. But ultimately that's a demand signal that you're inventing as a VC.
Do you not feel that's the same thing on blockchain though? Like the same thesis back in 2017, 18, people just talking about the blockchain all the time, the technology, the banks going on chain and so on as well. Do you not feel that people had this sort of like imminence of like this is going to work and make it work?
No, because even in 2017 it was true that it was it wasn't that VCs tried to make blockchains useful and used, it's that VCs noticed that blockchains were being used and we started using them ourselves. And you know again if you cannot understand why somebody would want this product, you're going to be a terrible investor in it. So, you need to be able to cultivate and if you don't have that yourself, hire people who do who have the native understanding of why this thing is valuable and why it's not. Look, there were many attempts to build enterprise blockchain before the current crop of blockchains that actually are getting usage and getting assets tokenizing, getting stable coins on chain and blah blah blah blah, right? And that generation of enterprise blockchains are basically dead. You know, most of those things we don't even remember anymore like the Ethereum Enterprise Alliance and you know, Corda and all this stuff. All this crap that we don't even bat. It's not even in the history books. Um, it all didn't work because again, it came top down rather than bottoms up.
Okay. One of the things you just mentioned there that you did, you wrote something that stopped me this year. You told that like you wrote that when your fund does a deal, you sign a legal contract, not a smart contract, right? You run a crypto fund. So what does it tell people if you don't really use your own product?
It's a good point. Um so we use our products in certain cases but only in very simple cases. So you know for example when we have vesting, very often our vesting is happening through a smart contract rather than happening through just a legal agreement. Um but you're absolutely right when it comes to the deal docs, when we actually sign the investment contract. Yeah. That is not written in a smart contract. Now why Mhm. So there's a number of reasons why it's not happening on a smart contract as opposed to a legal contract, but the main one is that when a smart contract goes wrong, we kind of don't know what's going to happen, right? It could there could be a bug here, there could be a bug there, there could be some bug that allows them to take all the money or to unwind our our shares or something. Who knows, right? Whereas when a legal contract goes wrong, Yeah. it's usually pretty predictable the ways it's going to go wrong. And ultimately, that legal contract is going to be interpreted by a Delaware judge. Mhm. And we have a pretty good mental model of what Delaware judges do. There's a lot of case law about what Delaware judges do, but what will the blockchain do if there is a bug or some implementation uh mistake in this particular smart contract? We don't know how it could go wrong. And so the the claim that I made in this post is that this feature largely arises because as humans, we understand written language much more intuitively than we understand code. Okay. When code goes wrong, we don't really understand how it's going to go wrong. Yeah. I think that's going to be less true for AIs. Okay. For AIs, yeah, they actually look at language. Yeah. And they're like, "This could be interpreted this way, could be interpreted that way." You know, look, I've analyzed the entire corpus of all the judges that are in Delaware, and sometimes they say this, sometimes they say that. There's genuine ambiguity here. But when they look at the bytecode, Yeah. they can say, "Ah, I see exactly what it's going to do in every single scenario." Yeah. And a lot of the bugs that humans introduce will never get introduced by AIs, especially when they can use techniques like formal verification to ensure that the code does exactly what they say it's going to do. And so my claim is that I think a lot of the ways in which humans have found crypto and smart contracts to be kind of unintuitive, to be difficult to use, to be scary even in that, oh, if I make a mistake, is it going to mean all my money gets lost? To AIs, they're actually going to think the opposite. They're going to think actually with crypto, with smart contracts, I know exactly what's going to happen. And I'm much better at reading and understanding code than I am at understanding a legal contract that's written in human fuzzy language that's going to be interpreted by a judge who I've never met. Okay. That tells me that I think the adoption of crypto is going to accelerate the more that we use AIs. Yeah. Because AIs are going to find this a better way to implement agreements with other AIs relative to how humans implement agreements with humans.
That that's a great point because I feel that at the moment, I mean you mentioned at the very beginning you're very well versed when it comes to AI, you understand how it works, the language models, you specifically understand what tools people are using. You mentioned Hermes bot, there's Victor now, there's these all these AI assistants coming out from nowhere from Instinct and Nous and so on as well. Dragonfly, let's talk about your your investment thesis because you know so much about AI. Are you investing into AI?
So we invest in the parts of AI where we think we have an edge, and that really means means one, the intersection of crypto and AI, and second, the ways in which AI is going to change the future of money and of markets. Mh. So um we're not doing you know some random Neolab. We're not going to go try to invest into you know the next uh um core innovation in AI or the next data center company or the next chip manufacturer. Um but we do invest into things that we think for example compute markets. Um being able to build derivatives on compute which today doesn't really exist. It's not an existing market. Um, it's exactly the kind of thing that crypto and the the understanding we have of how to build markets from scratch actually makes you better prepared to do than an investor who maybe knows a lot more about AI but doesn't really understand markets. Mh. So u these are the you know if you think about prediction markets for example we're obviously early investors in Polymarket. We understand prediction markets really well. Today, Polymarket's US domestic regulated off-chain venue is bigger than their overseas DeFi venue. Meaning that Polymarket is more of an off-chain business now than an on-chain business. And Kalshi, you know, Kalshi also has a lot of crypto investors on its cap table. Kalshi has no crypto at all. Kalshi is entirely a domestic US regulated exchange. M. So I think the reality is that crypto is kind of you know the ideas behind crypto of understanding the way in which these small changes in technology can have reverberating effects through how we do almost everything. Mhm. That is the core of I think what we are good at and what we know how to underwrite and how we can help founders. So for example we're investors in Venice.ai. Venice is a privacy preserving unrestricted AI inference platform. So it's think of it like private ChatGPT and it's growing like gangbusters. They're now over 100 million revenue run rate. Um this product you know they have a token but most of the inference on Venice just happens on you know normal GPUs. It's just happening out out there on like a normal platform. Most their customers aren't crypto users. They're you know paying with Stripe accounts but one, Venice has a token. Two, they use NEAR and Phala in order to get their confidential compute to allow users to be able to do inference end-to-end encrypted on a GPU without anyone being able to see what they're doing. But third, if you want to use Venice totally anonymously, you don't have to auth in. You can go in with just a crypto crypto address and you can pay with stablecoins. So they literally don't know who you are. They don't have to know who you are. If you really want to be completely private in your AI usage, you can be, but you don't have to be. This kind of thing where I think you see more and more of these platforms where crypto is a part of the story but it's not the be-all end-all of what they do. Um more and more you know what adoption looks like is that you are one part of many people's stacks but the core of what they do is not crypto in the same way you know we don't look at all the companies being built today and say oh they're all internet companies. Every company uses the internet but they're not an internet company. That's what it looks like to win as a technology and that's where crypto is today.
So when it comes to Venice for example, right, you found a unicorn. Um, take us through the thought process of finding Venice.
So we actually knew about Venice for quite a while. Um, we first saw them I think back when they first launched on Base. They did a fair launch. So they launched the token and um, you know, they they they airdropped to a bunch of the users of the platform very early on and early in the very early days they were kind of wandering through the desert. They didn't really know what they were going to do that and it was mostly bootstrapped at the time because so uh Eric, the founder of Venice, he's a successful entrepreneur from the early days of Bitcoin. He did SatoshiDice, he did ShapeShift, so successful founder in his own right. He was able to self-fund the first generation of Venice with his own money. Um it was eventually after they really started to scale and had much higher compute costs that we first met Venice and um you know we made the mistake that we first met them I think in December of last year and at that time they were doing roughly 16 million revenue run rate and we passed and huge mistake. We could have gotten in at a much better price. Um but uh we were just a little bit concerned about um you know was this growth sustainable? How long was this going to you know they kind of had this growth spurt and we didn't know if it was really going to uh sustain compared to all their competitors they had and there were a number of people who were doing this kind of private inference stuff at the time. Um, fast forward just a couple of months later, their revenue was growing crazy fast. We could see that they were clobbering all their competition. They were outperforming everybody. They were gathering more compute and being able to serve a lot of the demand that was growing just as fast as their compute was growing. Um, and we realized, oh [__] we made a mistake. We got to get back on the train. And so we got back in touch with the team. We proposed uh an investment, got to know them, built some trust, added some value in order to justify our keep, and we ended up leading their round. And uh now you can see since we invested the company's revenue has more than doubled in the span of just a few months.
Take us through the investment structure in terms of like the way the company is structured from finding a hot deal to investing.
Yeah. So we try to see as much as we can um within the deals that are within our purview. We're probably seeing call it 80 to 90% of all the deals that get that get funded in our in our sector. M. So step one is making sure that you have high deal flow coverage and people on your team are seeing all the most important stuff that's happening. A lot of that comes from founders finding us but some of it is us going out and finding the founders. So once we see a deal then the question is uh is this something that's worth going deeper on diligence?
How do people where do people can find Dragonfly? Do you guys go to conferences to find this if there's founders want to find you?
It's a little bit of everything. So of course you know I'm here at the Avalanche Summit so I come to conferences and I'm making sure that we're in the places that matter. Um, you know, I'm going to Token249 in a few weeks. So, it's, you know, one of the big conferences in Asia. Um, there's that. There's also, of course, that we're very active online, so I am very public figure in terms of my thoughts when I tweet and and write in public. Uh, we also do podcasts. So, me and Tom uh from Dragonfly, we have a podcast together called The Chopping Block. Um, Rob from our team also has a podcast called Empire. So, we're very out there. People know what we think. They they hear our voices. A lot of our clips sometimes go viral when we have something to say that people uh either agree with or disagree with. Um so that's a lot of how we end up on people's radars. At this point, if you're in crypto, you know Dragonfly. You might not uh have us be your number one choice, but you know who we are. And if you like us, you like us. If you don't like us, you don't like us. So, um our philosophy has always been we want to put ourselves out there so people know what they're getting. Nobody's going to be confused if they end up partnering with me about who I am and what I'm about and what I believe. And if I'm not the right person for you, it's better that you know that upfront and decide for yourself whether you like me. Thankfully, more often than not, founders are like, "Hey, I like how you think. I see very much from the stuff that you put out there what kind of person you are. I want to work with you." So, assuming they want to work with me, then the question is, is this person the kind of person we want to invest into? And a lot of that is diligence on the founder, but it's also diligence on the market, diligence on if it's an existing business, they have revenue, they have customers, they have uh a growth rate that we can actually dig into and underwrite. And especially if they have technology then we can go in and examine and really understand how much innovation is there, what's the quality of the underlying product. So once we get all that then we take it to our investment committee and the investment committee is the partners. We sit down together and it's you know majority rules. So we chat with each other about this deal that deal. We see you know what the evidence is that this company is worth investing into. And if we believe in it then it pass our investment committee and we go commit to them and try to win the deal. And we don't win every deal that we do. We win most of the deals that we end up committing to. Um, but you know, winning the deal is kind of the hardest part very often when you're talking about the some of the best companies because if you're a good company, the balance of power is in your favor is that you've got all the great investors coming around the table wanting a piece of your round. Uh, and so a lot of the skill of VC as you get more experienced is learning how to win the deal. And that's not just in, okay, what pricing do you offer in a competitive round, but it's also, you know, sitting across from somebody, how do you convince them that you're the partner that's going to take them over the line versus somebody else who's just going to be a check?
Actually, talking about that, um, are you guys an operational VC? You know, do you guys bring in external third parties to help with some of your portfolio companies as as soon as you can? For example, um, recruitment agencies, marketing agencies, PR agencies. Is uh do you bring in you know a network of distribution channels straight away to them available when they come to you with problems on the first day?
We do. It's a big part of how we end up winning a lot of these deals and why people recommend us so highly when they talk to other founders and get references on us. So uh platform in VC is this called the platform and platform team that we have spans uh recruiting. I think we have probably the strongest recruiting team of any of the crypto funds. Um we also help out with PR. We also help out with security, which is a big thing for a lot of crypto startups. Um, we help out with data. So, we have a little bit of everything. And of course, we have relationships with almost all of the big players, exchanges, uh, platforms, financial institutions. And so, when people need their first customer, their first relationship, their first introduction, uh, or they just need help with vendors, right? They need help access to compute, they need access to credits, they need access to this or that, um, accounting firms, etc. All of those things are things that we centralize in our value add when we work with a portfolio company and try to bring them from zero to one.
What are some of like the the trends that you're looking at at the moment? What what's hot for you? If someone had to meet you right now, Token249 in a few weeks and want to speak to Dragonfly, what do they need to do? What what sector do they have to be in?
There's no single sector that you have to be in. The reality is that there are things where there's more momentum at any given time, but we're oftentimes investing in things before they have momentum. And I think that's kind of part and parcel of being a good VC is that you're not looking backwards, you're looking forwards. And looking forwards often means things that aren't yet hot are the places where you want to be putting your money. So when we did the seed round of Athena, for example, um you know, we led their seed round at I think it was uh a few months after uh Terra had collapsed. Yeah. And at that time people thought DeFi is dead. Stable coins are dead. Definitely can't be doing a DeFi based stable coin. And so most people thought this is totally uninvestable. How can you touch this? And our thought was that this moment will pass. DeFi is okay. Terra was unique. There was there was fundamental problems with Terra, but they're not things that you should generalize over the entire industry with. So, we thought it was going to be great. Um, you know, the time that we led the seed round of Monad, um, a lot of people thought that like, hey, L1s are over. There's no more room for any new players. And we thought, no, I think what they're doing is really compelling and there's room for more players. Yeah. So as a VC it's very important to not be too prescriptive about trends, themes or you know these kind of cookie cutter startups. Um now at this present moment where am I seeing the most heat? Mhm. I'd say right now there's a lot happening with prediction markets. The intersection of crypto and AI is obviously very full of stuff. There's stable coins and and uh tokenization. Uh, a lot of the payments tradfi stuff is very very hot and of course that's where Avalanche plays in large part. Um, but if somebody comes to me with some idea that really breaks the mold and pushes something forward that hasn't been figured out yet, I'm always my ears are open, my mind is open to see what might be the thing that comes around the corner.
Let's break that down because one of the things you just mentioned was prediction markets, right? Uh, prediction market volume um, collapsed right after the World Cup. Um there's a lot of volume obviously during that time because there's a lot of hype. Um from that there's a lot of segmentation of other prediction markets that obviously were not similar to Kalshi and Polymarket but we're just segmenting in for example uh you know a prediction market for sports, prediction market for horse racing, different segmentations of prediction market. Have you guys invested into any of those and what are your thoughts upon that?
We've done a couple seed bets in people who are doing these kind of verticalized prediction markets. I think it's tough if you are being shadowed by Kalshi, Polymarket obviously right now we're in a duopoly and if you are saying okay you know I'm going to take one subset of the markets like let's say I'm going to do up down markets right and okay you know this five minute bitcoin goes up, bitcoin goes down uh these markets it's kind of like a subset of their business that's not enough for you to really have a unique go to market and some unique value add that's going to build you something big. If on the other hand you say look uh I think the problem with these guys are not tackling is internationalization is you know there's markets that are going to be specific to Southeast Asia or specific to LATAM or specific to you know some other region that these guys are just not targeting, they're not doing the regional specific go to market, they're not doing the localization, they're not doing the ground game and I'm going to go do those things that I think actually is more differentiated.
It's more unique. It's saying these guys are so big they can't focus on these other smaller markets. That gives me the right to win if I just really focus and run at one thing. That to me is more of a shape that I find attractive rather than, you know, I'm going to go carve off one little piece of this and say, great, I'm going to do prediction markets for just sports. Well, you know, Polymarket and Kalshi are already doing sports, so competing with them toe for toe is probably not going to end well.
Have you guys invested into any specific localized companies? I'm not talking about prediction markets, but any sort of localized exchanges that you guys...
We can't disclose that yet, just because some of these companies are in stealth and we want to respect their low profile. But the answer is probably yes.
Okay, good to hear that. Yeah. In February you raised $650 million, right? In July you said crypto venture capital may be dead by 2030. Which one of those is the marketing?
So, neither is marketing. I would say that we have a reputation for saying what we really think, and that is very much what we really think. Now, when I say crypto VC will be over by 2030, I don't mean that Dragonfly will be over. What I do mean is that what you think of as a crypto VC — a fund that only does crypto deals, that mostly exists to buy tokens and kind of drip them out in the market and blah blah blah — the things that only have expertise in this one thing of like on-chain layer 1, DeFi degeneracy, blah blah blah blah. That thing, of which there were many, many, many crypto funds that emerged over the last few years, they will not exist anymore.
Like what was it, 87% of crypto investors have died in 2022. So...
