Ep 147: MALWARE - Trends from EthCC with Evin McMullen: Infinite Protocols, Aggregation, Shared Sequencing, Decentralized AI, App Layer & More

[00:00] I just cannot get the bangs right, but... [00:02] That's okay. [00:03] They look great. My ring light died halfway through this, so I'm going to really hate every single cut down that is available on the internet this week. [00:32] Just boys club. [00:33] Hi. Hey. How's it going? Good. Welcome to Malware. This is a podcast where Dina and I talk about the tech news from the week. And this week, we were at an Ethereum conference in Brussels. It's usually in Paris. So you can imagine the amount of shit that I and everybody on crypto Twitter has talked about the downgrade that is Brussels to Paris. It's unprecedented how much we have all talked about it. So we won't bore you with it here. But never need to come back to Belgium is my feeling. [01:03] And... [01:04] I am ready to go home. I'm ready to go home. You had enough waffles and french fries? Yeah, like the biggest, the most upsetting aspect of it has been I've yet to have a good meal. I mean, we had an amazing meal last night at the event that we threw with Kraken that was so beautiful and grown up and fun and shout out to Kraken for partnering with us on that. But that was like a super high end situation. Michelin star. A Michelin, exactly. You know, walk in, get the champagne. Like one point they came out with like a tray of caviar and I was like,

[01:34] Wow. Excuse me. So that meal was incredible. But besides that, I have had [01:40] terrible food terrible yeah i don't want to shit on belgium i don't want to shit on brussels too much because because evan just explained why ecc is here and it's one of the founders it lives you think she's gonna tune in to this podcast fair fair anyway we're in brussels we've been in brussels all week and lots of events so i heard that evan says 300 side events um [02:06] It's crazy how much there's going on in the city right now. I will say malware is not always so crypto. [02:14] but this week's show is extreme it could not be more crypto [02:18] It's more crypto than I've ever been. Off the deep end. Yeah. We had a great guest on, Evan McMullen. She's the founder of Disco. She has been a boy from the very beginning. She is a kind, generous, smart, so smart. Like someone who's so smart that I... [02:34] I am like... [02:37] It... [02:38] makes me [02:39] think about the capacity of the human mind. [02:46] Like sometimes I get distracted when I'm talking to her because I, I like stop thinking about what she's saying. And I just start thinking about how people say we only use 20% of our brain. I'm like, I'm using 2% and this woman is like 110%. She's firing on all cylinders. Totally. That's so funny. What's so rare is she's so kind. I have had so many conversations with Evan where we,

[03:09] I'm sure we're going to talk about 15 things and we get to the first thing that [03:14] five seconds into the first thing and we have to just spend the rest of the time talking about that because she will let me ask wait wait wait i don't get that i don't understand this can we do this like does this mean this and it's really special and i really appreciate her coming on so we were seeing all these trends i was taking notes at the 300 side events that we attended and all the things that i didn't understand or that people were talking about as the next big thing just keeping it in a shared notes with you and then we got to the end of the week and i was hoping that we would [03:44] really know them and then we got to this morning and thought wow we need someone to come talk to us we need some help it's a call on the big guns [03:52] Exactly. So Evan was generous and kind enough to say yes. And she does an amazing job sort of talking through a lot of the trends that we've been seeing this week. That's that. [04:01] Thanks for listening. [04:02] Thanks for listening. [04:03] Thank you. [04:04] Hey, Natasha. So a question we get asked a lot is, what do you look for in a crypto platform? So let's talk about it. Well, Dina, I look for a secure, no fuss platform that I can dive into right away. That's why I love today's sponsor, Kraken. If you're waiting for the right time to get into crypto, Kraken makes it super easy and intuitive to get started. Plus, if you get stuck, they have an award-winning client support team that's available 24-7, along with a bunch of educational guides, articles, and videos to help you along the way. If you're ready to check out

[04:34] www.kracken.com backslash boys club and see what crypto can be not investment advice crypto trading involves risk of loss cryptocurrency services are provided to us and u.s territory customers by payward ventures incorporated pvi dba kraken view pvi's disclosure at kraken.com backslash legal backslash disclosures [04:54] Live from ECC, we have a very special guest joining us on today's podcast, Evan McMullen from Disco. Welcome to the show. [05:04] GM GM. We are looking fabulous. It is a wild day here on chain in Brussels. We're so excited to dig into it. Friend of the pod, I should say. A dear friend of the pod. Dear friend of the boys. One of the boys. One of the boys. A main boy, actually. Top boy. Okay, so we want to talk about different trends and conversations and big themes that we've seen throughout the week at ECC. Before we get into that, [05:29] What is ECC? What are we doing here in Brussels? So the Ethereum Community Conference has been happening for several years, usually in the summer, located in Paris. However, because of the Olympics this year, it has been relocated to the hometown of one of its organizers, Bettina, which is Brussels, Belgium, where we find ourselves here. [05:50] So in addition to visiting the seat of the European Union, we also have an opportunity to experience the traveling intellectual circus, making a stop in a new city, which is how I describe conference season in particular locations. Not actually a single conference, but this year, ECC Week.

[06:13] involve more than 300 independently organized events. Wow. Oh my gosh. I didn't realize that. Okay. That makes sense why there seems to be an unending amount of side events. And that's, there it is. There's the data on it. Do we have any sense of numbers like attendees? I think probably in excess of 10,000. However, attendees are really difficult for us to measure here at ECC because [06:38] the tickets for the main event, [06:40] are quite hard to come by because the venue can't hold as many people as who would like to attend. So the many, many side events also give a lot of different places for people to visit. So unlike some conference week like ETHENDER, where there's one main event where you could sort of use the number of tickets there to measure the scale of the community, I think we have to look elsewhere. So I would be very curious to see that data, but I'm willing to bet it is, you know, around 10,000, perhaps more. [07:07] Wow. Okay, let's get into it. We've been spotting trends, taking notes, and no better person to bring on to explain the trends that we've been hearing about than Evan. So we're going to do sort of a rapid fire thing here where we're going through different themes that we've been hearing chatter in the elevators and at the dinners and at the parties and have Evan break it down for us. So let's start with infinite protocols, hearing about hundreds of L1s. We're hearing about roll-up [07:37] your thoughts. Tell us everything. Break it down. Block space is abundant as far as the eye can see because everyone and their mother has realized that sequencer revenue is a great way to make money as long as you can fill that block space.

[07:54] So from Alt L1s to Layer 2s to Bitcoin Layer 2s to, you know, shared sequencers in new and different forms, we have now observed the launch of quite literally thousands of blockchains. [08:09] So the question remains, [08:12] Have we just recreated Web 2? [08:15] where we have a bunch of different databases, except for this time they are public and really expensive. Yeah, what comes to mind is what goes into the block space. [08:27] Yes. For whomst is all of this block space who will occupy it. And so I, you know, I applaud our colleagues and fellow boys, such as the POAP team who have, you know, service set of capabilities that consume block space and can be deployed on, you know, on other protocols. [08:57] you know we will bring demand to your chain for your block space okay i have a few questions we have a shared sequencing here on the list as well and let's bring it up to the top here because i think it has something i think it's in the same conversation around block space as you were saying so basically what you're talking about is that there are all of these different alt l ones and layer twos and everybody's seeing that there is a business in shared sequencing am i tracking

[09:24] So I would say [09:26] So the difference between a sequencer and a shared sequencer. Let's start there. So if we match, we've got Ethereum. [09:34] happy little blockchain, you've got a layer two on top of it, maybe like an optimism or an arbitron. And when transactions happen on that layer two, [09:42] They are batched together and a summary of what happened in all those transactions gets sent down to the main chain. Okay. So when you send a summary of what just happened on the layer two down to the main chain and to sort of write the outcomes down and make sure that there's always that reference point. [09:58] there's what's called a sequencer. So the sequencer basically takes the events that happened on the L2, puts them in the correct sequence, [10:08] So it makes kind of a list of what happened and then sends that condensed list of events down to the layer one. [10:15] Now the sequencer collects about 25% of the fees that would go to Ethereum layer one if you were putting all that data down there. [10:24] so you can kind of quick question and the people that are capturing that sequencer revenue are the layer twos [10:29] The layer two, that's correct. So usually there is a team or an entity that's in charge of that sequencer, making sure that it is prepaid with gas money. [10:41] because the sequencer acts [10:43] Like the sequencer for an L2 basically acts like a user writing data onto the blockchain. [10:48] Except for instead of a user like Dina or Natasha, it's a whole ecosystem representing a whole lot of different users.

[10:55] But either way, that L2 wants to write data to the layer one, which means similar to a prepaid laundry card like you might have in college, [11:03] You have to prepay the sequencer so that it has gas money to write things on Ethereum, to be able to pay for that transaction. Sometimes teams forget to refill the sequencer. So, for example, the Arbitrum chain has halted before because they forgot to top up [11:19] their card to pay you have one job ethereum exactly one job so you can imagine that if you've got a whole lot of transactions on that layer two and you're taking a little slice a little fee of those uh from you know those transactions [11:33] that that money can add up pretty substantially. [11:37] Some protocols make over a million dollars a day on sequencer fees alone. [11:42] And so some of our very thoughtful colleagues realized that building a blockchain is hard. [11:49] Building a layer two is tough. And so having a preset kind of a preset recipe or preset tech stack that you can deploy really quickly could be a really great way to augment the abilities of your L2. So in the optimism roll up ecosystem, we'll see there's OP mainnet, but then there's also Zora chain. [12:08] There's also base. There's also coming soon. World chain. [12:13] And so if you're going to build an L2 and you want to use the same tech stack as an existing L2, like Optimism, you can choose to just use the same tech stack as their chain and borrow their sequencer or rather send your transactions through their sequencer.

[12:30] So that's what a shared sequencer is. It's when multiple blockchains... [12:34] use the same route [12:36] to relay a list of their transactions down to layer one. So they're basically like, you know, you and your friends using the same prepaid card in your laundry swipes to be able to, you know, snack swipes to be able to buy something. So you guys. So it's sort of a shared bank account to pay Ethereum that's shared between layer twos built on the same technology. OK, so question for you. [13:00] So would... [13:02] this might be dumb, but does OP have its own [13:05] sequencer [13:06] that people are sharing. And then I've heard also Espresso is like another business that is their businesses that exist that are just shared sequencers that aren't layer twos. [13:16] So... [13:18] I would. So absolutely. Yes. Espresso is an outstanding example of that. I think there's that. So there's the creation of the shared sequencer, the actual technology that goes into it. [13:28] and then there's the implementation of the shared sequencer. So in the same way that, you know, Optimism created the OP stack but Base applied it, similarly Espresso can, you know, create or define terms of shared sequencers that can then be implemented by different blockchains. Okay, cool. Okay, great. Tracking. Okay, so we have infinite protocols, we have shared sequencing, [13:55] Where do we want to go next? [13:57] Aggregation. [13:58] Does that feel like the right next? Let's do it. Okay. Let's aggregate. Talk to us. We, you know, we basically described a very fragmented landscape. We got blockchains on blockchains on blockchains. And it's kind of tough to send money, send, you know, information in between them in the same way that it's hard to move data from, you know, Instagram to Reddit to your email. It's just not built to flow in that way.

[14:22] And so [14:24] with a challenge in this in this landscape is that we've got money. [14:28] fragmented across those chains. [14:30] This is the opposite of free flowing financial capital efficiency, right? Because we have all these [14:37] all this money basically trapped in little pockets that are really tough to get together. [14:41] And so a theme that we're starting to see is that all of this abundant information [14:46] needs to get brought together so we can treat all of these blockchains like one giant pool of liquidity, one giant pool of assets. So I would say, you know, we grew up in the information age. We are entering the aggregation age. [15:01] What that starts to look like is bridges everywhere, bridges in between blockchains that you can send capital from more information from one place to another. But that gets real messy. [15:11] Some of those bridges are in your own companies. Yeah, please. [15:15] I'm never not terrified to bridge. [15:18] Terrified like there's no there's no scenario where I'm bridging and I'm feeling comfortable. So I I'm excited for the day that we've I [15:24] aggregated that. [15:26] perhaps out or [15:28] abstracted it away enough so I don't know when I'm bridging. [15:32] So that is a spicy take. [15:35] Because if you don't know when you're bridging, you don't know what protocol you're on, which I think is okay. But we'll get back to that in a minute. Okay. Okay. [15:41] So this concept, right, bridging is stressful. Sometimes the bridges come from the protocol themselves. Sometimes they're independent projects. What does auditing look like? How is this being managed? Many different flavors, none of them the same. And so we have this patchwork landscape of janky hops from one protocol to another if we want to move around or, you know, prepare for airdrops.

[16:03] So if anyone is curious about that experience, I encourage you to try to bridge capital from one L2 to another and then use it on a third. [16:13] It's going to be an interesting challenge. [16:15] So if we look at where there are incentives to make bridging easier, to make it easier to move money from one chain to another, we have an interesting tension. [16:24] The idea of unified liquidity, you know, money that can move on any chain where we don't have to worry about which blockchain we're using. [16:32] I'm all for that vision. I think that abstracting way that complexity would be great for users. [16:37] That is opposite to the incentive system that is in place today because blockchains want to keep the money that lives on them there. They do not want to lose their capital to other chains because total value. [16:51] on a chain, in a chain ecosystem, that is still the metric that matters. [16:57] So where we can find a middle ground in this tension, where there's some coopetition, as it were, are in the roll-up ecosystems. [17:04] So for example, the OP stack chains are incentivized to make it easy for users and value to move across them, to move between them. [17:14] They're all built on the same tech. [17:15] they all participate in the same airdrop scenarios. [17:19] and they all embrace similar governance principles. [17:23] Whereas if a bunch of capital were to move seamlessly from the OP stack off to a totally different stack, such as Arbitrum or ZK Sync, [17:32] I probably would imagine that, you know, that would not be an ideal outcome.

[17:37] So we have this sort of tension of, you know, where where should it [17:41] value be allowed to flow, where should it not? So one example in the world ecosystem, we're starting to see basically the ability to to have seamless liquidity across chains, but just in the OP stack. [17:54] So Catalyst AMM is an example there where you don't have to worry about separate bridges from one OP chain to another to another. However, that seamless liquidity is only available inside of OP land. OK, I think I'm tracking. So what does this have to do with we were talking before we started recording where there was a hot take at dinner last night that chain abstraction will happen within the next five years? How does that relate to what we're talking about here? [18:24] awesome builders yesterday, including our friend Ilya from Nier. [18:27] And so because he basically invented AI, I was very excited to ask him a bunch of questions, including, do you think that [18:36] in five years, we will pick a home chain and hang out there most of the time? Or will we not worry about the blockchains that we're using and just use apps to try to do what we were doing that day? [18:46] And his opinion was very much the latter, but it's not going to take five years. [18:51] Okay. That the ability to, you know, move seamlessly from app to app and be able to use your near wallet on Ethereum or Baruchain or Monad or wherever you go. [19:00] that that will be a no-brainer in just a few short years. Okay. Wow. Since we're on the topic of apps, we're on the application layer, I want to draw attention to this incredible tweet that I saw earlier this week. User is at ASCI girls. Cindy, I don't know her, but she...

[19:22] looks very cool. And the photo is, the tweet is a photo and it's a half eaten cake that has like some designs on it. And basically like people have eaten, it's how would you describe the seven? Actually, I'll let you kind of walk us through this. Well, so it's, it's literally like a picture of a cake in the tech stack where the layers of that cake start at the hardware and then the network and then the infrastructure and then the application layer somewhere at the top. [19:48] And as with a sheet cake, [19:50] you usually start eating that sheet cake on one side, and then, you know, you make your way to the other side. So this is a literal picture of a half-eaten cake, where the untouched half of the cake stack [20:03] I had the application layer on. [20:05] And so this was a beautiful representation of what's really happening out here. The tweet says application layer untouched as usual, which is perfect. What this like really hit for me this week because I come to these conferences, especially builder and like dev heavy conferences like ECC. And it's so much talk about the infrastructure. We're building these huge 10 lane highways and there's very little cars that are driving on these highways. [20:35] usage. There's not a lot of like top level app usage. And so I think that that's something that like I'm constantly having to manage my own like honestly cynicism around throughout the week where I'm like, man, what are we building all of this for? There's not enough people and users to justify this huge investment in the infrastructure stack when the application stack is so there's so many questions around that. Evan, I'm curious if that if you if you have those same feelings or if you

[21:05] her. [21:06] Oh, but Dina, with the infrastructure layer, you can raise around and achieve a liquidity event in the course of 18 months. If you want to build an equity business around an application, it's going to take years of hard work and focus to build up a revenue generating entity ready for... [21:25] IPO or for acquisition. But, you know, that snark aside, all respects to our colleagues in the investment and infrastructure space. But [21:33] I think for a long time we were starved of infrastructure because the decentralized backends that we had simply could not compete with the basic everyday centralized backends. [21:44] I think we got very zealous in that endeavor, inventing new ways to move value. [21:51] And apps are hard to build because they require research, [21:55] Empathy. [21:56] and an understanding of users' experiences that are different from your own. [22:01] And that is not a skill set. [22:02] that is common in our ecosystem. - Feel that. - My truck. - Feel that hurt. - But I will say though, there is hope. There's hope at the app layer. The homies are coming through. Want to call out a couple of our friends in Boys Club, who are actually builders down south in Buenos Aires. I was recently introduced to a project called Bombus. It's similar to Dice, the ticketing app, if you've ever used that in the United States. [22:23] a combination of a list of raves, tickets you can buy to go to these cool concerts, social interaction afterwards, all built on chain. [22:32] And the front end, the website, the app does not talk about the blockchain at all. So with their 450,000 paying users using this tool to go to parties every single weekend,

[22:44] There are [22:45] some examples that are starting to scale up in places where there is, you know, a solid product and it fits the market and their lifestyle beyond just our beloved Sky Casino. [22:56] Wow, I love that. And what I especially love about that is that it's not [23:00] based in gambling as a user behavior. The other, um, [23:04] example that I wanted to bring up here on the application layer is polymarket, which is a crypto enabled prediction market and [23:14] Everyone is talking about it and I think it's getting a huge amount of traction. I heard a [23:19] I overheard someone in a party talking about how they think it's going to like meaningfully impact the US election and there's all these big grandiose things are having on polymark and I buy it, but I will say like the fundamental user behavior is that around. [23:34] betting and gambling, essentially, to a certain degree. And so I think that that's [23:40] what it is, but I love to see other things that are tapping into other user behaviors. Not to say that the gambling side of it is, I'm not going to put a moral layer on it, but yeah, I just love, I love to see this, hear about Bombust. It sounds, yeah, really cool. I too love stuff that does things. Okay, great. Do you want to talk about decentralized AI? I know you have some hot takes. I'll get some hot takes on that, but really quick, actually, I have a few comments on Polymarket, [24:10] Oh, yeah, totally. It's been all up and around. I... [24:13] I probably market here and there at home when I'm feeling spicy. As one does. Can you walk us through how it works? I've never made a bat gamble there or bet. Certainly. So based on the documentation, because as a US citizen, of course, I'm not permitted to interact with

[24:29] such capabilities, but if I were hypothetically, what I might do is go to the Polymarket website, [24:37] and basically see a pull [24:40] of what Polymarket users think is going to happen. [24:43] in different world events. [24:45] So it could be predicting the outcome of a sports game. It could be predicting the outcome of a political election. And actually, it's that particular example, outcomes of political elections, where polymarket becomes pretty popular every few years. [24:58] Polymarket is a unique [25:01] platform because it is a crypto native [25:04] prediction market. [25:06] kind of delivering on the promise of Gnosis's original design back in 2017. [25:11] So this prediction market allows you to basically make a bet on the potential outcome of one of these events. [25:19] And then based on the odds of your bet, you can receive a payout if you were indeed correct. [25:26] Now, what's interesting about the Polymarket crew, the Polymarket users... Really quick, is that in crypto or is it in USD? So in crypto. [25:35] Okay. [25:36] So one of the interesting things about Polymarket users, of course, is that they are crypto native enough to be able to make bets in crypto. [25:44] Right. If you imagine your average crypto user [25:47] they tend to have a number of political beliefs in common. And so the polls on Polymarket may not be representative of the sentiment that [25:56] of the voting base for those predictions or the event participants for what they're predicting. So they're a little bit skewed. And generally, these types of prediction markets tend to lean a little bit conservative anyway.

[26:09] What's particularly interesting and unique about Polymarket is that our former president, Donald Trump, is an enthusiast. [26:16] In fact, he will frequently look up his own polls and screenshot them, print them out, sign them with a Sharpie, post them on Twitter, and then his press team will quote them as polls. [26:29] So this then gets repeated in the media. He was describing his polymarket results as though they were, you know, from some other source. [26:37] This creates a positive feedback loop of enthusiasm and also allows for, you know, those who are early supporters and enthusiasts to financially benefit. [26:47] from the fact that their guy is gaining in the metrics that matter in that particular way. [26:53] And so it is one of the first times that I've ever seen a global leader embrace a crypto app. [27:01] use it explicitly as a tool in their communications and campaign. [27:06] invite their audience, their electorate, their supporters to use that app. [27:12] and even describe parts of that application as a proxy market for investing in themselves. [27:18] Trump has described polymarket predictions related to the election as Trump coins. [27:23] I am so happy that the expression that these people wish to have has found a home that's ready to facilitate it. [27:33] I'm so happy for them. They seem very happy about that. So positive. It's just always, always. You have such an incredible resilience towards looking towards a positive and being optimistic. I don't know how. I think that's also why you've been in crypto so long. You can see that you can come back to the good. Come back to the good. It's incredible. It's such a beautiful. It is. It's amazing.

[27:53] uh, [27:54] characteristic that you have and it really defines you to me and i i love it i do have to say i feel sometimes in your presence like i'm so [28:03] Jaded and cynical and I hope that you don't I hope that that doesn't I know it couldn't deter you. Oh my gosh my you know my [28:11] relentless optimism, I think, is just a function of seven market cycles. But I am grateful for the rigor of, you know, friends like y'all who are willing to engage ideas seriously. I think rigor is a love language. And the harder you critique, [28:29] the more you want to understand and battle test the idea that you're evaluating. [28:33] to make it be as sound and, you know, unshakable as it can be. Because if your proposal can't withstand questions, it's pretty wimpy, isn't it? [28:42] I have a question for you. What do you think about Trump being at Bitcoin Nashville? I am shocked that he didn't show up sooner. [28:48] Yeah. Given the obvious nature of having a young group of [28:53] registered voters who donate to political elections are unencumbered by the concerns of reproductive health. [29:00] and do not share the same, you know, potentially marginalizing traits as many who who would not naturally be attracted to him. So it feels [29:09] As though someone asked ChatGPT where Donald Trump should show up for campaign events later this year and it provided the most obvious answer. [29:22] Well, I'm going to try and go because I'm based in Nashville. So I'm going to try and see it with your own two eyes.

[29:31] It's time for a more open, inclusive, and transparent financial system. A system that serves nearly everyone, everywhere, all the time. That's why we love today's sponsor, Kraken. Kraken is a crypto platform that provides a super simple on-ramp to the world of crypto with a 24-7 support team. Crypto transcends physical and imaginary borders. No matter where you are, you can send funds easily and quickly to almost any part of the world. Plus, forget about waiting times and waiting lines. You can send, receive, and trade crypto anywhere near instantly. [29:59] See what crypto can be at kraken.com backslash boys club. Non-investment advice. Crypto trading involves risk of loss. Transfers to a third party are not available on Kraken. Cryptocurrency services are provided to U.S. and U.S. territory customers by Payward Ventures, Inc., PBI, DBA, Kraken. View PBI's disclosures at kraken.com backslash legal backslash disclosures. Okay, let's go to decentralized AI. You're at an AI Ethereum collab day of some kind. So I want to hear your thoughts. [30:29] chatting about what are your convictions around [30:32] This... [30:33] Whole thing. Reports of AI takeover are largely overstated, I think. [30:39] You know, as children who grew up on sci-fi, I too wish to have an earpiece and a spaceship that, you know, manage my life for me and remove the intellectual labor of decision making and learning and curation. [30:53] But beyond, you know, [30:56] basic chat enabled capabilities or recommendations, we aren't seeing

[31:01] a ton of instances where we are replacing processes in our everyday lives with AI enabled tools. I think especially for a lot of us who are embracing leading edge tech, there are a lot of novelties. [31:14] but not a lot of everyday serious, significant changes that are happening at scale. [31:19] Now, of course, the enterprise space, there are lots of places where this is starting to happen, but not necessarily in every single home in America, even though there are more than 100 million users who have given a shot at chatting with the bots. [31:31] When it comes to the challenges of building out AI at scale, we are in our very early baby days. We are still trying to build out models that can give us good answers, can give valuable recommendations to us. It's still pretty expensive to use these models. [31:51] The amount of energy consumed by a Google search is about 10 times less [31:56] than the energy consumed by a chat GPT query. [31:59] And we are running out of data. [32:02] to train our AI models. [32:04] We're actually about to start experiencing what's called model collapse. [32:08] which is where you don't have enough data to feed a model or trend model. [32:12] And so you insert some synthetic data. [32:15] to try to train and iterate on that model. And if you do that too many times, the model will collapse, or it will basically have gaslit itself [32:22] to no longer be able to tell what came from a real origin and what came from a synthetic origin. [32:28] So we're in our very early days, early stages. [32:30] struggling to figure out where we're going to get enough energy to power these centers, where we're going to build them, whether in Herndon, in Virginia, with the rest of the Internet or across America next to energy sources like in the Midwest.

[32:42] And... [32:43] But the interesting question around decentralized AI is like literally why? Which AI models do we mistrust so much in their current form? [32:53] And their use is so essential and so vital to such an important process that we need to have a token for decentralized governance. [33:02] I think, you know, I'm going to be a little spicy here. I think a lot of decentralized AI talk [33:07] comes from parties who are interested in good buzzwords and an 18 month liquidity horizon to get themselves a token. [33:15] We have a theme here on this podcast. Yeah, that tokens should not always be the first thing we reach for when we try to solve a problem. [33:24] They are the proverbial hammer to our nails. But we also have some really sharp and cool friends like Eric Borges, who are thinking about dissident technology, how we can limit the censorship or data aggregation and collection of AI models and AI interfaces. [33:41] I think there is important [33:43] to explore privacy preserving and decentralized technologies. [33:47] but only at the point [33:49] where it is logical to do so. I think that there is more decentralized AI tokenization, you know, around this week, certainly, than we've ever seen before. [34:01] And I... [34:03] I feel like some amount of that is buzzword bingo. [34:06] Can I ask what are people tokenizing? So there's the the models themselves and then there's the

[34:13] interactions with the models and then of course there's like the whole infrastructure stack below the models as well. What like part of that stack are people tokenizing or people talking about tokenizing? [34:23] every part that includes a repeatable technical process. So literally everything tokenizing the energy. [34:32] that goes to the [34:34] power plants, tokenizing the power plants that go to the training and inference centers, tokenizing access to the, you know, GPUs that sit in those infra centers. [34:46] tokenizing the models themselves, tokenizing the data that goes into the models, tokenizing the right to compute on the data that goes into the models, tokenizing the output. [34:59] of the models tokenizing access to control who can see the range and control the range of those models and even tokenizing the models access to the original data that was put into them so if you can decentralize it you can tokenize it and someone suggested it okay it sounds like i should have been at this uh this decentralized ai day i need to learn a thing or two about what's going on here [35:29] you talked about Bombas and I know that you're excited about a lot of the things that we talked about today. But what have you seen this week that you has given you some enthusiasm, some inspiration? What was something that really maybe sparked your interest in a new way this week?

[35:43] So I want to give a huge shout out to the ZK Sync ecosystem. [35:47] because I think it is a common practice where leading blockchains will kind of be in charge of their ecosystem, in charge of their fiefdom, telling their projects what they will do and organize and kind of coordinating, hurting all the folks together. And that's really important in the early days of an ecosystem. [36:06] But when you start to see that ecosystem becoming so big, so decentralized, so organized, [36:11] that they can coordinate without the mothership. [36:14] that sort of [36:16] Flippening of authority, flipping of power is a really awesome thing to see. So this week we saw the ZK Sync ecosystem come together. [36:26] more than 20 projects, pooled capital, co-hosted a gathering, and simply invited their home chain to be one of the guests. So I thought that that was really awesome, very inspiring to see, and an example that I hope other Rolabico systems will practice. [36:42] both in their parties and in their protocols. [36:45] I love it. Evan, thank you so much. It's always such a delight spending time with you. I'm always so amazed at... [36:52] how smart and kind and open and warm you are and generous yeah willing to take the time to explain and not make it i mean feel stupid i totally it's such a [37:03] special trait that you have yeah it's a real it's a real gift so thank you for your time so fun seeing you this week and we'll have you on the pot again soon for sure always a treat to bro down with the boys thank you guys so much and i will see you on train bye evan you