Portal VC Founder: How To Invest In A Post-AI World (Huge Changes)

Transcript

All righty, guys. Welcome back to the tokenization tower, episode number 39 of our AI super cycle show. Our premiere weekly AI show powered by our friends at Near. Visit near.com to learn more about what they're doing in the AI sphere. I'm here with Evan Fisher tokenization tower. Evan, great to have you, man. Welcome to the tower New York City Empire State Building. Happy Friday. >> Good to be here. >> This is this is great. I mean, so I I was at Goldman out of undergrad and so I can I can see the office all the way downtown. um which is is wonderful and like I mean we'll get into it but so many people have talked about how excited the suits are. Yeah.

>> about crypto this week. I mean I I got lunch with the Goldman digital assets team on on Tuesday and and the suits are exciting. So >> that they are it's an institutional bull run. It's a token bear run almost. It's like a token bull uh institutional bull token bear. You are an investor at portal the co-founder of portal ventures. you guys have done, you guys have a massive portfolio across uh kind of crypto, fintech, adjacent industries. Today, I kind of wanted to have this conversation about what it means to be a venture investor, an investor in general in this world of AI. Just thinking about how it affects your thesises, how it affects the way you're uh assessing founders, assessing industries, sectors, and more. Before we get into it, just

want to give you a chance, you know, what is Portal? What's your long-term thesis here? Um, and just kind of give us the quick and dirty there. >> Yeah. Yeah. the the quick and dirty. Um well, so all we do is preede first check-in blockchain investing. Um it's awesome to be in this office. We were the first investors in Flume. >> Beautiful. >> Uh back in it's two and a half years ago at this point. >> Um I started Portal uh four and a half years ago. I was at Insight Partners before that. Spent three and a half years there. Helped invest 600 million across global fintech, internet um and software. And I was looking in 2021 at protocols. I said, "Wow, everyone thinks these are just random tokens, complete speculation." And and they were at the

time, but I said, I think you can build interesting income statements on these things. So in I wrote an article for Banklist in 2021 that was titled how makerdow makes money and we built an income statement for it. So I built the fund on that thesis. We said how do we invest in protocols that are just going to be profitable and be interesting businesses. You know, fast forward four and a half years, we've built a reputation as one of the best preede investors in crypto. We manage 120 million across the two funds. Um, great mix of investors, institutionals, and large single family offices. Uh, and you know, we're we're we're doing more of it. We're still excited about the category.

>> Yeah. Awesome. Yeah, man. I mean 2021 was you know pre AI it was pre-Chat GBTE was also pre pre a lot of like this like revenue fundamentals outlook that the space currently has in terms of like where we're at in that kind of view and in that sector. Um when you're looking now into the investing landscape preede first check in again kind of like the thematic here being investing in the era of AI we've seen different iterations of uh crypto companies kind of come and go and now there's this kind of like big narrative overhang and investing overhang in the AI category over crypto but these two are kind of coming together and so you know talk us through

just like very broad and then we'll get into some of the of the nuance responses of like what does it mean to be a venture investor in the post AI world? What changes do you investors have to make to their thesises to their understanding of their of the world to their world views to adapt to this kind of AI first world? >> Yeah, a lot changes. I think a lot of people talk about what changes. I mean, for instance, you can build code faster. A codebase isn't as defensible. You can ask hard questions. You can be an expert on anything. Um, I'm my job is to figure out what doesn't change, frankly. And so for us, what what doesn't change? Well, we think there are huge problems to solve specifically across finance and

other categories with blockchain. One of my takes is I don't think crypto is a financial technology. I think that's wrong. Um, so huge problems to solve. Um, and the founders that go really deep with an unfair advantage on understanding that problem. It's unique insight. They're using an inflection point technology. That doesn't change. that's that's still the same. And so >> part of being a crypto investor and being in this AI bubble that we're in right now is we're familiar with it. >> We've seen it before. And what it is is everyone gets excited. Eventually we'll enter a trough of disillusionment. And you know a lot doesn't change. And so even in the software world I don't think that much changes. You know is it easier to build companies? Yeah. But that just

means it's easier to build new companies that you want to invest in. So from my perspective, it's a great time to be a venture investor. There's never been more opportunity. So we what keeps us up at night honestly is where should we focus because there are so many interesting thesis to run out today. >> Yeah. Yeah. So in the crypto market specifically then uh when you're looking at the different sectors, the the different categories, we're very much looking at revenues and monetiz uh monetizable modes uh differentiators. Now, we've kind of ran all the iterations of the different perexes or the different exchange or lending or whatever it is. Um, what is what is the what does the moat look like in blockchain in a in in maybe in

specific sectors or categories in this post AI world? And then what becomes far more commoditized now that intelligence becomes very very kind of abundant. what becomes far more commoditized and what becomes more rare and differentiating that you will look for in a founder, in an idea or in a company. >> There's a ton in there. So, I I'll try to address all of it, but let me know what I missed. >> Um I think uh you think where to start on this there honestly so so many. >> Um >> so what is interesting to look at today? Well, I think network effects are still interesting. I think liquidity is still an interesting moat at certain scales of course. Um I think trust is an

interesting moat. AI doesn't touch any of those. So you know when I think about categories that are actually safe and protected from AI blockchain and fintech are one of them because building code faster isn't what ships a fintech or a blockchain product. It's the trust. It's the incentives. Um, so most of what we do we actually think is quite protected and just gets a tailwind because our founders are building faster. I was at lunch with a founder in our portfolio yesterday and he said his CTO over the post using uh, you know, claude code to build things is doing 20 PRs a weekend.

That's insane to think about before. And so it's just a tailwind to him. But how's he going to win? It's not going to be based off the code base. It's going to be based off the insights and the markets. Um we can go a ton of different places from here. So I'll pause before I I give you a monologue on on all those questions. >> Yeah. And so liquidity um distribution network effects these become these are very powerful non-replicatable by AI like like >> dollars are not commoditizable because you can't just print more dollars. You can create more intelligence but you can't create more money.

>> I could build the Bitcoin codebase really quickly now but that doesn't make it valuable like Bitcoin. Yeah. Yeah. Interesting. Okay. So, distribution, liquidity, network effects, taste, even branding, even maybe um it, you know, we think about this acceleration of intelligence as like all right, we're going to see the rapid growth of and the rapiding cheapening of the cost of intelligence, meaning that we can we can build systems and products that took much much much longer, much quicker. Um, which also creates more rapid competition. Like I can spin up competitive products and protocols to things that previously took a long time

or a lot of money to build much faster. We saw this with token launches and with blockchains already, right? We had back in 2020, 2019, you know, you had to raise 10, 20, 30, $40 million to launch a blockchain. And then we just had L2s and roll-ups as a service which a lot of that didn't work out in the same way that that people thought but we drastically commoditized the cost to launch chains tokens pump fund came along same kind of content we had to raise all this money all these exchange listings no just do it on chain same and so is there's I'm just kind of like thinking what's coming here what is the where could we see that play out again in this market where where there was this big cost or big effort required Now we have you know far better

engineering, far less cost, far more pace. Where do you see that that like material tipping point or 0ero to one possibility? >> Yeah, it it happens across the board. That's what happens with some software businesses that you know friends of mine invest in where oh a company goes from founding to revenue to 5 million in a year's time and and like you don't flinch at that anymore because the barrier to entry has decreased so much >> in blockchain and crypto. We see so many you know new trading apps and trading platforms that just get built quickly and they give out quick incentives. They bootstrap a a trader base, but it's mercenary sometimes. And so, you know,

you end up having these really fast revenue spikes, but what's it good for? So, as an investor, what do we have to do when a business, you know, goes from one to five just like that? We sit there and we say, well, why isn't your competitor going to do that? Like, what's the unique insight? Right? again. Um, so I I think what it what it actually creates an issue for is Zerp created so many momentum investors >> where they just they looked at the outputs and they said that's enough to understand the inputs. They said oh business is doing these metrics that's enough for me to think it's a good business and it's going to have terminal value and that's not the case anymore.

So you just have to think like a traditional venture investor now. So the boom and bust is going to be with the momentum investors. >> Interesting. So then who loses in that scenario right? Because sometimes when you think about who wins or how to kind of allocate for like you know in any scenario any kind of path dependent scenario it's sometimes it's easier just to look at who's going to lose right like in this kind of tokenization boom that's happening right now the losers are going to be the the analoges environments call the the call the broker to buy the stock the middlemen in between the the owner of the stock and where the stock certificate lives when you know when a lot of this comes on chain stable coins replacing remittances western etc. Um, so kind of thinking then in the

lens of network effects, liquidity, distribution, taste, branding, who who loses then if we're if if this is the winning characteristics of investments whether public or private, who loses? >> Um, it is the generalized horizontal platforms are probably most at risk. >> Explain that. Um, I was at dinner with a friend who was explaining this and we were talking about it and so I know how to apply it to software. We're still trying to figure out how to apply it to to blockchain.

>> Yeah. >> But let's say that you use Salesforce right now or you use some ERP system and it's good enough but you know he used the example you run a liquor store. Well there are lots of nuances to running a liquor store. You have to check IDs. You have certain licensing that you have to manage certain inventory rules. And so it's it's not perfect, but you say it's it's good enough. And so entrepreneurs are going to come along and they're going to say, "Oh, I could over the course of a few months build something that's actually like an ERP system or a supply chain management system for a liquor store."

>> And in the past, you'd say the development cost of that was too great. >> But technology is deflationary. And so cost to produce reduced. Therefore, we can have more specific products. Yeah. >> Um so you can have more consumer surplus is one good thing. Uh but I think you're just going to have a lot of smaller outcomes. So if you're a generalized platform, you need to figure out how do I build customized solutions all of a sudden? Is it worth it? >> Yeah. So then applying that framework to blockchain, there was this kind of like every app is going to have its own chain idea for a while, right? But that didn't really end up happening.

>> Yeah. I think that's ridiculous. >> Yeah. >> Well, and to be clear on that, like people depending on people's bags, they change their mind on that every six to 12 months, you know? >> Yeah. But then how do you apply that framework then to so so so like the generalizable horizontal platforms or what you know whatever that is in crypto will not be as favorable as the more specific use case specific ideas. Well, well, so what's interesting is for crypto, I think it's different again because it has the benefit of network effects and liquidity, >> right?

>> Um, so you know, where the liquor store example applies, it actually doesn't. This is back to my point of I think crypto is quite protected from all things AI. Like I I think it is a tailwind, not a headwind. It's helping teams build faster. As an investor, you have to be cautious because you see quick momentum. You say, "Is this durable?" But I don't see it really risking market disruption. Another thing people talk about is like, "Oh, are the founders, you know, are all the founders going to to AI?" >> It's like, well, >> you know, you have fintech founders entering, you have people from marketmaking firms, HFT firms entering, >> and you just have some people that come from financial institutions. And so they

were never in the candidate list of of building AI. People that are momentum have left. Um but it's it's like I I wish I could give you more. I I really just think it's a tailwind. >> Yeah, it's interesting kind of then more specifically on the convergence, right? We we we've seen a lot of traction around like decentralized training, decentralized inference. Um obviously like you know products like Venice have done extremely well with with their privacy aspect. nearest private inference. We've seen this convergence start to happen. I know you guys are kind of more focused on the financial services but also have done some kind of like convergence investments.

If we're thinking about so kind of like this conversation has has been like investing broadly in that world but then like what about more nuanced in the convergence of the two worlds it do do you do do does portal have a thesis on uh the value of blockchain in AI world? Yeah, we Katrina on our team wrote an article on this um three and a half years ago. Yeah, I think which was beyond hype here are the four use cases we think matter in AI times crypto. >> Cool.

>> And if you go back on that article, >> it's pretty much played out that way. Um one of the first is can you use incentives to uh you know create compute to acquire compute and so we were an early investor in exhibits as an example. they started off as a decentralized GPU marketplace have scaled massively on that thesis. Uh, another one is sovereignty, you know, like that I think is still underexplored actually. Um, or excuse me, not sovereignty, providence. Um, and so Worldcoin is a great example of that. How do you decide what's real? But I think there's tons of opportunity there.

Um, we're investors in a business called Interval, which is using blockchain um, for private enterprise AI. And one of the value props is you know when you throw the data on chain you just know what's real what's not real compliance so on so forth. >> Um the the other two you know you have some governance dynamics which I think are turning into harnesses >> and the last one is how do you incentivize data collection which is absolutely playing up. >> So that's everything that we believed three and a half years ago and we we've been investing along those lines. Um you know the first check into plume but we do a lot. We're always saying what's the market missing that they're going to be on to in two to three years, right?

>> And so two to three years ago, FTX world, people were like, "Oh, >> you know, >> they're never DeFi is never going to come back. It's it's it's dead." And you know, one of our best DeFi investments ever. I I I wrote the check the week after FDX. I was like, "This is >> of course it's coming back. It's the best use case so far." >> Yeah. Yeah, man. It definitely I mean it pays to be like country in that regard when it comes to kind of thinking about um you know bringing these worlds together as you're saying um obviously there's this kind of growth of AI having like like this kind of tailwind as you're saying to uh all the use cases that blockchain and digital assets can serve you know but one thing that that I think is is clear is that people are pe

people are looking at this and wondering kind of when will the the growth of this kind of open-source decentralized you know insert whatever part of the stack here really ever like compete with that the kind of like frontier labs and the actual like model companies that are uh you know have raised billions and billions and billions of dollars and so my question to you is and again as part of this kind of like investing in the post AI world I think we're going to get to a spot where we're we're going to realize that we've gone too far we've given too much data we've allocated too much resources. This this thing is getting too fat, too good, too fast, you know. And I'm just curious if you think that a tipping point is required for people to kind of like turn over to the

more open- source open source, decentralized iterations of kind, you know, some of the more closed source centralized iterations of LM and AI that exists. just like broadly like we kind of reached that tipping point when the Bitcoin block was the Bitcoin Genesis block was created when it was like banks on the bank of brink of bailout 2008 it was like holy we need something different do you think we might get to a point like that in the AI world >> you know I think the consensus answer would be oh the US government's going to nationalize the AI labs it's a national security threat and that privacy is going to be an issue or oh Gavin Newsome's going to make it so you have to sign into your computer with

um an ID and you're you know you're going to get KYC to use your computer. Honestly, I don't think that would do it. Like I I think it's either going to be a better product and people are going to use it or it's not and the you know privacy use case like it doesn't people there's not much track record of people making decisions for the sake of privacy. >> Yeah. Um, even within crypto with people who, you know, are very investing heavily in everything privacy related. Whenever there's something easier, cheaper, faster, better, they use the product. Yeah.

>> Like we're you humans are lazy at at our core. So, >> it needs to be better. Now, what could make it better? If regulation actually starts making these products harder to access for certain questions >> or makes it harder to access the best models such that the best models publicly are from the labs are actually weaker than some of the open-source models all of a sudden then you start to see adoption. So >> it still needs to be a better problem. >> So it's not like an ethos game or like a privacy game or a data game. It's uh it's just humans are you know privacy has has this kind of stated belief issue where there's like this stated belief of of the desire of it but then the actions don't always kind of show and yeah

you're you're kind of saying it's just going to be a product thing. >> It's you could run a fun game with your crypto friends and play stated versus revealed preferences on privacy and just go through their tech stack. >> Yeah. Yeah. Yeah. Yeah. Interesting. kind of zooming out to the more like fintech financial services investing era that you guys have really really done a lot of deals in and then thinking about the acceleration of these kind of more macro events. So obviously AI kind of being the topic of conversation today. a lot of these investments that you do and I'm sure that you found this I found this you know through the investments that we've done plume being one of them as well which is there's all these like path dependencies externally and internally that affect the likelihood of success of an investment and like one of

the externalities right now is we're just in this mega acceleration spending capex on AI and so as it pertains to some of the financial services and fintech investments that you've done you know in 2020 20 2019 maybe you know you couldn't have predicted that we could write into an LLM to send stable coins globally and you know it would be done in fractions of a second for fractions of cost agentically but like these tailwinds have played in our favor such that AI is just very good at reading zeros and ones blockchain hashes transaction addresses IDs are zeros and ones and letters and numbers and the rise of stable coins obviously kind of like probably more more predictable so

when when it when you're looking at your portfolio of fintech companies you feel free to talk about some uh financial services companies. The proliferation of both AI and stable coins simultaneously from like a regulatory perspective and just a capex perspective are like very interesting headwinds for or sorry tailwinds for like this kind of path dependency in a positive way. Um talk to us about kind of what that that more macro broad view is that you have for fintech and financial services and then how that kind of dribbles down into your investing framework for some of your portfolio companies.

>> Yeah. I I view um I'm sure Kyle will get mad at me because I'm about to do an analogy. Yeah. Um but I I view uh blockchains as kind of the app store moment for finance. So what does that mean? Like the app store really created a ton of new software applications because it created global distribution. It managed a lot of your backend for you and the result was you could have stories like WhatsApp. You know that was a huge story back in the day. sub 20 person team building a you know a platform and an application that scaled to millions potentially even billions of of users I forget when they were acquired. Um so what does blockchain do?

It it opensources some of the hardest parts of financial services innovation. You have the custody, the distribution, the liquidity, it's all open source software now. And so you're lowering the barrier to entry for creating these new products. And that's where I think you're seeing this proliferation of, you know, teams building unique financial products. So one way that we've expressed this is we have this thesis we call latent capital markets. There are so many financial markets across the world where there is demand for products but the demand's not met because the cost of serving it has been too high >> or there's been a you know PR issue or a regulatory there's so many different reasons but it's it's just underserved demand. And so we specifically were looking into some of the AI spend and

the energy spend. We said wow uranium's interesting. So, we went super deep on all of these energy categories. Landed on uranium, realized uranium doesn't have a spot market. >> Yeah, >> it's it's a it's obviously a huge asset for the next coming dec like the coming decades, but there's no spot market. And if there's no spot market, there's no derivatives market. >> So, there's a huge prize to be done. So, we were the first check into a business called Uranium Digital. Um, they are tokenizing uranium to create the world's first spot market for uranium. >> Yeah. >> Physically settled, sells through brokers. You can access it through your traditional prime brokerage platforms if you're an institution. That's possible because of blockchains, right? Because a tiny team, team sub 10 can all of a sudden create a product that like it

would have taken a financial institution to actually get off the ground. >> Yeah. And so again, kind of coming back to where we started on the top here, this is like a second and third order effect of this AI capex boom. You get this high demand for commodities. copper, uranium have been absolutely on a tear the last 12 to 18 months. And so you're, you know, then then we start thinking about like, uh, the financialization of these assets. Uranium not having a spot market. There's like a couple ETFs you can buy right now, but they're mostly miners. Um, there's like a Tezos spot market right now on chain, but doesn't really exist globally. And then you have like the financialization of compute happening as well. There there's a company that just raised from A16Z I I believe yesterday it was announced and

they're basically just all in on how do we create a basically commodities like market for compute doesn't really exist right now. >> Yeah. And so that's happening too. And so like we're like >> there's all these little ripple effects but one of them is the financialization of these of the commodity sector that is powering this kind of AI capex boom. >> Yeah. And if we zoom out too, like some mental models I have for the space are um the world is getting faster, more complex and more AIdriven. And finance and financialization is actually really valuable for that. I think financialization has gotten a horrible rep because it's just been associated with like >> markets, gambling, whatever it is.

>> Financial markets are are like queriable. It's maybe it's like a Nick land idea like it's like this queriable intelligence all of a sudden. I mean if you want to understand a belief about the world study financial markets understand the revealed preferences. You can ask such interesting questions. So I think financializing things is actually great and I think as the world becomes more complex it becomes more valuable. So that's where things like financializing compute, financializing commodity markets, financializing new credit instruments, financializing intangible assets, financializing trust, reputation, all of this I I actually think it's good because it helps us query a very complex world and understand where value sits.

>> Well, it's, you know, it's just about creating the best price discovery as well for things that people are interested in. Yeah. >> And that and I mean I think the thesis here is that whether it's finan uh financializing compute, uranium, etc. 247 global market access via blockchain rails is the best way to create price discovery for global investors. >> Yeah. Yeah. Correct. And so we spend a lot of time asking what are other nonobvious things to financialize. We're investors in the company that's creating interval is this company I mentioned earlier creating a data marketplace. >> So how do you do secure compliant trades of enterprise data?

>> Really interesting. I think in the world of AI well if your data is public it's obviously commoditized. everyone's trained on it and so what are all the foundational models trying to do? Get private data. That's where the alpha is. Yeah. So you need infrastructure for that. Um >> so the world's going to be more financialized. >> Yeah. Yeah. Interesting. And so when you're thinking about like like the parts of this kind of like uh AI and finance stack, you've got kind of parts that are financialized as a result of the growth of the sector and then you've got kind of like the actual stable coin sending the actual trading or whatever it is. Where do you see like from the investor's perspective, right? Where do you see the

I I suppose like the value capture that will happen for fintex for financial service companies for these kind of like like as they collide with the agentic kind of finance flows and and workflows where are you placing your chips for here's where I think the these types of companies or these types of parts of this kind of broader picture are going to be the most valuable versus like the most commoditizable or kind of like basic. >> Yeah. Um what's really interesting about financial services innovation is the innovator's dilemma is stronger than what we explain to us. Yeah.

>> Um you know the innovator's dilemma is the the concept of uh the person who built the big technology in the past has so many assumptions about how they view the world. They're locked into certain worldviews because the products they've created. The incentives make it so they're not as incentivized to innovate. And that's where people often said, "Oh, you know, the big companies aren't going to compete with the small companies." Um, people like Google came along and have somewhat disproved that. Google, Amazon, like it's it's incredible the their ability to fight the innovative innovators dilemma. But and a lot of a lot of crypto investors go wrong when they view crypto as software >> because it's not. If it's if you're serving financial services, point to

financial services examples that have handled platform shifts well, you know, like Stripe >> is a new company. Robin Hood is a rel like both these are relatively new companies. So I I look at that and I say, >> well, I don't think the incumbents are going to capture this. >> Yeah. Um, so we're for the most part there are rare exceptions where I think that there's actually an innovator's dilemma like um where there's actually an incumbency risk, but for the most part we think it goes to new players. >> Interesting. And so that is that like is that like the banks that the incumbents, the brokerages, the traditional rails?

Yeah, it it depends on the category you're you're looking for too, you know, like for a lot of these credit products, the banks are are the incumbents for obviously for aentic payments. And that's one where like I'm I'm curious to see how it plays out because there there is a benefit to adding the rails and owning the distribution. Um but there it's like the Visa and the Mastercards are >> right. But like in that specific example, Visa, Mastercard, there's kind of conflicting views here of they know that their card networks are not meant to support agentic micro payments at scale, but they're also going to kind of build out stable coin rails. But then you have people arguing like, oh, like the credit cards are going to be better or you know, these card programs are better than stable coins. That one's been a kind of a big point of conflict,

but I think both I think both will work. >> That that's usually the answer. Is it >> um like financial services is the second largest TAM in the world, >> right? >> There are huge huge outcomes. When people were looking at payments companies over the last decade in the traditional venture world, >> the biggest mistake people made every round was undersizing what they thought the TAM was >> because they they would look at numbers and they'd say, "That's insane. There's no world in which you could have that much payments volume." There's sticker shock to it. And that's the case when you look at tokenization, you're like, "How big could tokenization get?" You see numbers that start with a T.

>> Yeah. >> You're like, "No, that that that could that could never be true." >> Yeah. >> Um but it's just such a big category. So the answer is like it's going to be all of it. Which is why we say sometimes people are like, "Oh, is is crypto just finance? If so, has it compressed? Is it not interesting enough?" I'm like, >> if that's our bare case? >> Yeah, >> that's that's interesting. And I don't think that's true. Like I think crypto is um is like a coordination technology. it's its competitor is the legal system in my mind, not >> it's it's not a financial technology. So, >> yeah. >> All right. So, just kind of zooming out here as as we kind of come to a wrap in a bit. Just thinking from the investors perspective, we've seen iterations of the ramifications of this AI boom. We

saw a we've seen a massive runup in memory stocks. We've seen a massive runup in commodities. We've seen a massive run up in uh storage and in compute and in chip and semiconductor stocks. >> Yeah. >> Looking out now with where we're currently at, what is what is Evan Fischer's Portal Ventures view of the world of investing, you know, either tangentially or in this kind of broad category, crypto, AI, global macro going forward. Obviously, if if if you didn't, you know, get into the early rounds of the big frontier labs or if you weren't buying semiconductors or commodities over the last 18 months, you kind of

feel like you've missed out. You feel like there's been a huge boom and I missed it. And honestly, I missed a lot of it. I I had a good bit of gold and silver and somehow they were kind of bucketed into that somehow. So, that's nice. But I didn't, you know, I didn't get any seed checks into uh any Frontier Labs and uh passed on all the SPVS because I just was like, I'm never going to get my money out of this. And uh looking at like semiconductor storage, etc. It's just been a crazy mania. Um where do we go from here from the investor who's looking at this market now global macro wise and wanting to allocate you know bitcoin and crypto AI based kind of investments uh commodities give us like your portfolio man put your

portfolio manager hat on how do we allocate in today's world >> well well I'll start with if you it's it's not too late to trade the memory stocks and if you believe that you should go on vest markets and trade the memory stocks the only place you can do that right now if you want me to plug more more portfolio >> companies >> also a portfolio company of of our of our fund. >> Wonderful. Yeah. >> Um uh so where do we go from here? >> Again, you have to kind of zoom out and ask the question like what do you believe? And I I'll focus on on early stage investing because like I'm a venture investor by training. So >> you're you're thinking 10 plus years out. Yeah. What's the world going to look like in 10 years?

>> Yeah. Uh the world is getting faster, more complex, and more AIdriven. And if you think about like the coordination technologies that we have for that, it's legal technology is is one of them. So it's like our legal system is starting to be outdated in terms of the mechanisms it uses to enforce things. >> Think about law. What does it do? It creates trust because there's a verifiable commitment. I see I'm going to do something. It's in a contract. And then there's enforcement. So you can trust me as a counterparty. Well, if you start to think about AI and the pace at which everything is going to run, it's like are are we going to are we going to do lawsuits from AI to AI for a

transaction that happened, you know, that happens every minute? Like there there's so many things that are going to play out that are just too fast and too complex for our existing legal infrastructure. Yeah. >> So, I think the most interesting blockchain applications into AI are coordination technology. M >> and I think people keep pointing to the payments and the finance of it because if you compare it uh to a legal technology again what do you use legal technology for in the first place was like accounting ledgers it was financial use cases were the first use case of legal technology. So that's where everyone's getting mistaken and they're saying oh blockchain's a financial technology. No, it's it's like a legal technology >> where the first use case is finance and the rest of them are going to become so

obvious to the general public given AI over the next 5 to 10 years. And so that's what we're most focused on right now as an investor. I don't think other people are playing here. I think people are too focused everything finance. >> Um so we're we're we're doing some deals there. The founders building there. We we I took a pitch just before this in Williamsburg. Um, that's what we're most excited about. >> So, focusing not on finance, not on kind of like the more financialized primitives, that's still a part of the vision, the fintech and financial service and the stable coin proliferation and just the kind of the growth of the total global crypto total market cap, etc. But it's really just this coordination focus, this next stage. You know, it kind of reminds me

of the of the blockchain, not Bitcoin era, right? A >> Exactly. And um don't get me wrong, we're still doing financial investing. >> Yeah. >> We're just we're in the business of saying what are people not paying attention to? >> Of course. >> And so, you know, when FTX crashed, it was DeFi. It was like when we invested in Plume, people were saying, "Oh, crypto is never going to be regulated in the US. Everyone's going to have to move offshore. It's going to be this like, you know, renegade technology." And I was like, >> "Yeah, sure. >> That's dumb. It's obviously going to be regulated. And whoever does that first wins, >> right?

>> Um and so for us right now, the coordination stuff is just completely missed and and we think potentially a bigger opportunity than some of these finance applications. >> Love it. Ladies and gentlemen, this has been Evan Fischer live from the tokenization tower on AI super episode number 39 powered by our friends at Near covering the investment framework and mental models for investing in a post AI future. thinking about commodities, equities, crypto, the second and third order consequences of what is happening across financial services, fintech, and aentic finance. Evan, it's been an absolute pleasure. Thanks for coming to the tower. It's been a pleasure. I'm sure you're busy.