AI ROLLUP #7: DeFAI Mindshare | Arc’s Multi-Chain Moves | ai16z: Sink or Swim? | Virtuals $40M Burn
01:04:56 Share

Welcome Bankless Nation to the AI Roll-Up, where we cover the recent news, developments, and drama in the intersection of crypto and AI. I'm David Hoffman here with my co-host and AI expert, Ejaz, and we are here to help make sure that you are up to speed with the rapidly evolving AI agent space out there. Ejaz, how you doing, my man? How was your week last week? Anything happen? Nope, not much. Pretty quiet. Um,

Okay, so there's going to be some listeners who do have context about what we're about to talk about. And there's going to be other listeners who have zero context about what we're about to talk about. So maybe we can just start flagging that.

Yeah. So, so I think it's important to provide context and also highlight that this is not a kind of like a jokey part of this segment. Um, this is kind of like, I want to take a moment to kind of like address something that is, uh, serious to me and meaningful to me and a lot of other people. And, um, I think it's worth kind of just taking a moment. So for context here, um,

Last week, myself and two other folks called contributors Marcus and Rope launched a DAO that will be fully dedicated to its mission of helping to grow this wonderful and wacky crypto AI space. And our founding principles from the start has been to use this DAO to help discover and support the best builders, creators and enthusiasts, the leaders of tomorrow, basically.

um and i want to start by saying that the core team and i took some steps with the launch process of which we truly regret and take full accountability for and if i'm being completely honest we were just so excited to launch and get this thing going um and in hindsight it's so obvious that we should have installed things like vesting for the core team and advisors and in hindsight we should have taken the time to slow down and make sure more of our community was involved and that's definitely like a fault on on our side and we take full accountability but

We just didn't. And that's on us. Ever since Mark Rope and I started researching, building and nerding out on this space, we've had builders reach out to us and ask them to help them. And whilst we did the best that we could, there was only so much time in the day. And we thought,

Well, we didn't want these builders to not get the support that they needed, right? So we thought like, you know, we'd rush to assemble a team across all the blockchain ecosystems, which we thought would be the most capable to help us and help these builders do what they wanted to do.

And that included you, David, and a host of other builders, investors, and thought leaders in the space that we believe can help us fulfill this mission of taking this small niche sector in crypto and grow it into the potential that we think it deserves. And unfortunately, that rushing cost us.

None of us even thought for a second, like literally, if you could hear the conversations we had before, none of us thought for a second it would get as big as it did. And we just thought, you know, we're a bunch of like nerds, to be honest, launching at a time when markets are down. So we'll have, you know, this small launch and then we can kind of build from here. But the opposite actually happened. And we learned, you know, some really hard lessons, which tie back to the points I made earlier around investing in community distribution.

um that being said i'm so grateful for the community that came through and showed their support it genuinely means the world to me um and we're taking steps which will be announced very shortly to repair that damage and to kind of create and fulfill on our vision and i just want to end this with saying that the core team advisors and i are fully committed to realizing our original mission and

I also want to say that this is the only time we'll be mentioning this DAO on this show. We want this series to focus on all the amazing teams building in this space and not focus on anything that we're doing, to be honest. The goal of this series is to amplify and build up others, not ourselves or whatever we're doing on the side, which is why we've never talked about AICC on the show prior to this too.

David, I know you faced some heat. Maybe you can shine some light on that. Yeah, yes, yes, we did. Many times throughout Bankless's history, we've been mobbed and dogpiled on crypto Twitter due to sets of facts or conclusions that were simply wrong or heavily manipulated to look as bad as possible. This is the first time that we actually did do something that we consider wrong, that we regret.

that we immediately tried to do as soon as Ryan and I both realized what had happened. So Ryan and I, advisors, we were able to contribute our share of five SOL alongside many of the other advisors. And that also included Bankless Ventures as well, got to contribute two SOL, along with our other third GP, Ben Lakoff, who also contributed one SOL. Upon the launch of the DAO, the price action of the token, the AICC token was like

pretty insane. If you go look at CoinGecko or some of the charts, it touched a billion dollars in market cap right on launch just because people were playing games, people wanted to buy. It was very exciting.

No one really expected that or like accounted for that. There was a decision made by the Bankless Ventures team to sell 8% of its supply of tokens on launch day. Which, can I just say, which like shocked me, to be honest. I think like, I just want to come clean and say like, I mean, you have the DMs. I hit you up and was like, you know, what's going on here, man? Like, this is...

like unacceptable and i yeah anyway i just you know you you were understanding of that i think you were shocked as well i was also shocked because i was not aware of the fact that bankless ventures made the decision to sell eight percent of the tokens um that amounted to sixty five thousand dollars uh just to put that into perspective yeah the decision to sell the eight percent of tokens was made without my or ryan's like input or knowledge something that we would have never

approved of had we been made aware uh once we found out we were like sent a tweet by a friend about somebody saying like oh bankless ventures is like dumping the token on day one and i was like well that's not true uh that's that's got to be a mistake uh turns out it was actually real uh once we found out it was real we immediately just stepped in and re-bought the aicc token with all the proceeds from the sale um just a few hours later from the moment of the sale so like

Like because selling tokens so soon into a project's lifespan, not something that we want to do, that we support or we condone. There's extra context for how and why this sale happened. There's a link to a tweet that kind of gives a play by play in the show notes.

It's just kind of like a dominoes of events. Unfortunately, nonetheless, it was done. Then there are also parts of crypto Twitter who love to drag bank lists through the mud. And we had just basically giving them the material that they needed to make this seem as bad as possible.

The narrative for some people on crypto Twitter became that Bankless had built this whole thing up all along, that Bankless sold millions upon millions of dollars onto our followers, that Bankless conspired to create the whole AICC down in the first place in order to do this.

that we started doing the AI Rollup with E-Jazz for the sole purpose of creating hype around the AICC DAO, which to be clear, the Accelerate DAO was never mentioned or promoted on any AI Rollup or any other Bankless podcast prior to the segment that you're hearing right now. There's also some alternative media organizations out there who took advantage of this very bad look and helped promote a twisted version of the truth.

Perhaps one of the more, I'll say, like valid bits of critiques is that Bankless had a role in elevating the prominence and clout of EJAS here, which helped by proxy to generate hype and significance to the launch of the Accelerate DAO. To be clear, wasn't our intent. Ryan and I found out about the existence of the DAO the Tuesday before it launched, five days before it launched.

Also, Ejaz, you and the other two core team members had already assembled a pretty stacked organization without any help from us. We didn't do any help with assembling that. I think many people out there perhaps first heard of you, Ejaz, by us bringing you on to the Bankless podcast as our AI roll-up host. And that's because...

You are on the crypto AI frontier where you've already established the relationships that you needed to make the accelerate DAO a reality without, again, without any help for us. I'm just a nerd in the trenches that just want to geek out and learn about this. Literally, like if you look at most of my tweets, it's like, can you please teach me this or that? So yeah, you kind of picked me up out of that. And it's been a roller coaster since, David. Yeah. So a lot of people don't have all the context.

Overall, call it a pretty dramatic start to the Accelerate DAO with a lot of lessons learned in hindsight. Any final comments before we move on? No, I just want to reiterate that

We're incredibly grateful for our community and the support that they've shown. And we are fully committed to realizing our mission. Despite this bumpy start, we're going to steer the ship and correct course and kind of forge ahead. And we've got some really exciting announcements, you know, that is going to kind of fulfill some of the things that we've heard from the community. But let's leave it at that, David. I don't want to talk about this. I don't want us to speak about this first and last time. Accelerate DAO has been mentioned on the rollout.

Yeah. So now I have to ask you. Yes. The theme of the week in AI crypto. What would you say? Update us on the last seven days of action, seven days of activity, seven days of fundamentals. What was the last seven days like?

Well, let me start with something which isn't fundamental but is important to cover, which is the markets were down, David. And I have this saying, which is by no means my own. Someone much wiser said this before me. But the narrative follows price. So what that means is if the price go up, all these narratives are like, oh, this is the reason why. And this is like amazing. Check out this new project. When the prices are down, David...

The pendulum swings back with ferocity the other way. When the prices are up, these are autonomous sovereign AI agents. When prices are down, these are just glorified bots. These are glorified bots. So we experienced some of the latter over the last seven days. Markets, in my opinion, did a healthy correction, David. And this was not just within crypto AI. This was within the macro perspective at all. And actually, maybe a spicy take here is

Overall, if you look at like the price performance of the crypto AI agent sector over the last, say, three months or two months, let's call it one and a half months, actually, because this sector, I keep forgetting how young it is, has actually outperformed macro, like to a crazy amount of times. Yeah, significantly, right? But obviously people like look at their tokens, they see a wild price accrual and they're like, oh my God, I'm used to this portfolio number that I'm looking at on my screen. And then when it corrects,

healthily, you know, after like a whatever, a 5X or something, people are like, oh, what's happened and all this kind of stuff. So that's what we experienced over the last week, a kind of like macro correction in majors. So like Bitcoin and stuff, and then that affecting crypto AI specifically. Especially, yeah. Maybe just to put some numbers on this, like Bitcoin got up to $107,000 and then bottomed at $91,000. You know, Ether touched below $3,000.

And then AI crypto tokens got hit especially hard. AI16Z, like one of the premier tokens, topped out at $2.25, bottomed out at a dollar, so lost 60% of its value. It's back up to $1.50 virtuals.

which topped out at $5, fell all the way down to $2.30, and is now back up to $3.60. This is what? A very skitterish...

bullish bull market looks like, I would say, is like things go up and down 30, 40% in a week because people's sentiments can change on a dime. Yep, yep, 100%. And if I kind of like could add a few other figures to this is, well, probably the main one, the total AI agent market cap, which I think peaked at

at around $17 billion about a week and a half ago, retraced to, I think, around $11 to $12 billion, right? And as you can see, markets have... I think this cookie data might be maybe slightly lagging or something, but markets are up.

today as we're recording this conversation on Wednesday. AIXBT is up 82% on the week. Yes. So these things are very, very reflexive, David. And like, you know, it's just interesting that like when the markets are down, people aren't very convicted in their thesis in general. But some of these things still stand and the market shows that when markets flip back to the positive, right? So we've seen kind of a healthy correction

within the AI agents market.

And what's happened this last week, if I were to summarize it, was a mixture of teams still shipping stuff and then a bunch of critics out there that were saying, you know, I think this is the end of the crypto AI agent sector. Now, obviously, like us on this show and having this whole segment around AI agents, we're kind of biased, but I do not think that's the case. I think we're actually furthest from that case. I put out a tweet a while ago, which basically said,

If you want to know in times of uncertainty or even certainty what on earth is going on, talk to the builders. Put your heads down and get in the weeds with things and you'll see that there are people that truly care about this space that are building some awesome stuff. So yeah, I think we should kind of like get into things, David. I wanted to start off with this really interesting chart that I saw on Kaito, which is kind of like a social sentiment tracker where...

Typically, they've had this metric called AI Mindshare, David. And AI Mindshare started off like this. If you imagine a little square, it looked like this. Tiny little blip at the corner. Tiny little baby little thing. And then halfway through last year, that just grew.

bigger and bigger and bigger. And then when agents came about, well, it looks like what you're looking like, what it looks like on this screen, right? Where it takes up like half the page. And for people who are confused or aren't looking at the screen right now, it's kind of like blocks of...

It's squares that each represent a specific segment. So it could be DeFi, it could be NFTs, and each square is proportional. Its size is proportional to how much it's being talked about within the ecosphere, within Twitter, within social media sites, within the crypto realm. And it's...

If you're looking at the screen right now, the AI square is taking up half the screen. 48%. 48%. It used to be 70. It used to be 70. Well, it used to be 70, but there's a reason. And this is why I actually brought up this chart. It wasn't to show that AI was taking up...

50%, but it's to point to a square that sits adjacent to it, David. Do you know which one I'm talking about? Yeah, it's the one that's extremely green, which implies extreme growth called DeFi, which we talked about last week. It was a big segment last week.

And I think I'm going to beat you to the punchline, Ejaz. You're going to talk about how AI used to have a single block of mindshare. And now there are two AI sectors that are different. You have AI and AI agents, probably. And now there's DeFi, DeFi with AI in it. So AI-enabled DeFi that has its own independent amount of mindshare coming in at 6.3%. Number three...

After memes, so memes coming in at 10%, AI coming in at 48%, and then third is DeFi. DeFi, we need a better word. AI finance, AI decentralized finance coming in at 6.3%.

Yeah, I think we need to either say DeFi AI or AI finance. AI DeFi is the easiest thing to say. AI DeFi. Yeah, and it doesn't sound ridiculous at all, David. But yes, you are smack bang on, you know, hammer on the nail. The AI category has split into AI and finance.

called DeFi AI, which kind of shows that there's a lot of attention, investment and focus on this particular subsector of crypto AI, which I think is super, super cool. And so we have to ask ourselves, well, why? Why is this whole sector even being spoken about at all? Well, we kind of spoke about this, I think two episodes ago, David, where we were saying, okay, well, what have we got so far in terms of like,

crypto AI agents. Well, there's some really smart kind of sassy Twitter bots, right? You know, people like to call them influencer bots, right? And AIXBT has kind of like embellished that quite a bit. I've been the leading example

But the next question was like, okay, I'm kind of getting bored of these Twitter. I didn't like influencer. Yeah. Well, there you go. There you go. And it's kind of like, well, what comes next? And the natural reaction to that was, which I think is actually kind of on the money here is, well, hang on a second. We've just built up an alternative financial stack over the last eight years.

Maybe they'll have some relevance there. And we've spoken about this before saying like, I think agents are going to be the connectors of all this complex blockchain stuff into a really neat UX where instead of having to spin up a wallet and sign transactions, you just ping an agent.

So basically, DeFi AI is the emergence of that. It's combining these AI agents with DeFi capabilities. So what does that mean? It means these agents will have their own wallets. It means these agents will be able to trade. It means these agents will be able to invest. It means these agents will be able to do all the things that you can do within DeFi today. And one might ask, well, why is that important? I mean, I think that's a good question to ask. There are many reasons. Number one,

I think a super smart, aligned AI agent that can work 24-7 and costs almost nothing to pay to do the work of 10 different humans is probably a valuable proposition. But of course, realizing this is going to take some time. But it's really cool to see the kind of focus emerge on this particular sector, David. This reminds me of something I said last week, which is like,

This is kind of just like we're talking about AI agents as like a focus point of this whole technological revolution that we're covering. But it's really the AI-ification of everything in crypto. Last week, we talked about how virtuals partnered with Illuvium, the game. So now we have AI innovating with gaming and gaming has its own sector. It's actually number fourth.

So you can kind of see AI starting to integrate with everything that makes up crypto in the same way that people are predicting that AI is integrating with everything about the rest of our normal lives.

I do want to bring up this tweet from Sammy talking about Franklin Templeton that just published a report on AI agents. This is triggering some memories of, I remember when, like back in 2020 and 2021, when we started to see institutional reporting about, you know, the emergence of DeFi. This is kind of the next step in the adoption of DeFi.

the mindshare of AI agents and what's going on. So it's not bearish, David? It's not bearish. It's not bearish that the suits are talking about? This is the next step on the bull market is the suits start making reports because their clients are like, hey, I heard about this agent thing. What's going on? And they felt compelled that they need to make a report. Did you read this report, EJAS?

Yeah, I did. I did. And what I will just say is, for those of you who have been plugged into the crypto AI agent space and all who've been watching these episodes, you're not going to see anything that is mind blowing. They are watching these roll ups in order to write the reports. They're watching this. Yeah.

Yeah, potentially. And what I will say is, the more prescient point that they make is, they think the sector is here to stay, which is really cool to, you know, see from an institutional practice, which, you know, are known for like taking long term bets, long term convicted bets, and their clients, you know, are expected to kind of be receiving the same type of thing. So for them to take the time to put out this report and highlight it for their clients means

to David's point, that they're getting some kind of demand or attention around this. The kind of like TLDR here is, you know, they expect it to probably grow to 100 billion plus. They expect these agents to kind of get a lot smarter and do a lot more than they do right now.

And they're kind of like watching the space very closely. I don't know if they're invested in any of this stuff. I sort of doubt it. But, you know, I'm always around to be surprised. But it's cool to start seeing some more of these ecosystem reports from some major players. And for people who are kind of like who's Franklin Templeton and stuff like that, they manage a huge amount of wealth. Billions upon billions.

I think it's like hundreds of billions in the TradFi sector. All their clients are, you know, people who aren't in the crypto trenches. They're people that are used to very traditional financial instruments, ETFs, stuff like that. So for these guys to put out this report, it's pretty insane to see. Yeah.

Yeah. Yeah. They start off the report talking about truth, terminal and goat, of course, and then fart coin as well. I mean, Franklin Templeton, as you know, trad fi as they come, they're pretty in the trenches as well. So you kind of got to tip your hat to them.

With over $1.5 billion in TVL, the METH protocol is home to METH, the fourth largest ETH liquid staking token.

offering one of the highest APRs among the top 10 LFTs. And now, CMEth takes things even further. This restaked version captures multiple yields across Kerak, EigenLayer, Symbiotic, and many more, making CMEth the most efficient and most composable LRT solution on the market. Metamorphosis season one dropped $7.7 million in Cook rewards to METH holders,

Season two is currently ongoing, allowing users to earn staking, restaking and AVS yields, plus rewards in Cook, METH protocol's governance token and more. Don't miss out on the opportunity to stake, restake and shape the future of METH protocol with Cook. Participate today at meeth.mantle.xyz. Celo is transitioning from a mobile first, EVM compatible layer one blockchain to a high performance Ethereum layer two, built on OP stack

with Eigen DA and one block finality, all happening soon with a hard fork. With over 600 million total transactions, 12 million weekly transactions and 750,000 daily active users, Celo's meteoric rise would place it among one of the top layer twos built for the real world and optimized for fast, low cost global payments. As the home of the stable coins, Celo hosts 13 native stable coins across seven different currencies.

including native USDT on Opera MiniPay and with over 4 million users in Africa alone. In November, stablecoin volumes hit $6.8 billion, made for seamless on-chain FX trading. Plus, users can pay gas with ERC-20 tokens like USDT and USDC and send crypto to phone numbers in seconds. But why should you care about Zello's transition to a Layer 2? Layer 2's Unify Ethereum, L1's FreeCash,

fragmented by becoming a layer two cello leads the way for other EVM compatible layer ones to follow follow cello on X and witness the great cello happening where cello cuts its inflation in half as it enters its layer two era and continuing its environmental leadership something we talked about last week which continued to make waves this week is uh arc arc dot fun um for those people who don't remember uh tee up arc for us and what happened in the last seven days of course so um arc

or Arc.Fun specifically, is a team who is building an AI agent protocol. So the way to think about this is your bog standard framework plus platform, but with some really interesting, you know, spices and sauce mixed in, right? And which is what does make them want to be- A competing AI framework to compete with virtual Eliza. Yeah.

Yes, exactly. Yeah, think of a protocol team which is competing with those bigger players. And so the context here is they take a slightly different approach in the way that they're building their platform and their framework. Well, firstly, it's written in the Rust code, which for those of you who have been following the Solana ecosystem, which is also written in Rust, you know, people are trying to

trying to compare the alignment between them. But people, well, it's actually interesting, David. People, or devs in general, would show that they don't really like Rust because it's a harder language to kind of learn. It's less accessible to the average developer, right? A software engineer that's coding in Python or TypeScript will go to a Zoribro or

an AI 16Z to kind of like experiment and launch agents on their frameworks. But there's a bigger learning gap to overcome in terms of engaging with like a Rust framework. But anyway, the Arc team specifically were convicted in their resolve of building out this Rust-based AI agent framework. And the reason why they've been getting a lot of excitement recently is two things.

Number one, they've been behind the scenes working tirelessly to...

support the initial teams that are building agents on that platform. So they're kind of taking a more curated approach where they're going to start announcing certain teams that are building agents. And what we're really hoping to see, David, kind of going back to the point of like, we've seen all the Twitter bots now, we've seen all the Twitter agents, we want to see something different here. We want to see some like interesting primitives, we want to see these agents do a little bit more than just tweet at us, right?

The second important thing here is kind of like this string of tweets, which honestly has come out over the last 20 years.

24 to 36 hours of partnerships that ARK is being made. I mean, there's some heavy hitters in here, like ARK announced a partnership with Abstract or Abstract Chain for EVM compatibility. For those of you who don't know what Abstract is, it's kind of birthed from the brainchild of Luca Netz, who is the founder of Pudgy Penguins. So ZK Sync Chain. ZK Sync Chain. And Saiga, who is one of the most cracked...

crypto AI agent devs that I've had the pleasure to interact with online and learn from. So the idea behind Abstract is, you know, it's going to be basically the best UX ever to interact with crypto and they'll have their own wallet and cross-chain compatibility.

I'll extend into that at another time. But the point is they've partnered with Arc to enable that capability for anyone that is building on the Arc framework. Other kinds of announcements are announcements with Eternals, announcements with this team called Send AI, which allows them to have a bunch of different...

integrations with solana so if you want to swap on salon if you want to trade on salon if you want to create a token on solana if you want to attribute um an agent with another token or if you want to send swaps or whatever you can do it through using you know these guys um this team sendai and their product um arc also announced with arbitrum which kind of like is another different chain separate integrating all the chains solana abstract arbitrage

Yeah. Okay. So let's kind of like condense all of this, right? Like, you know, we've just listed off a bunch of partnerships, but so what? Like, what does it mean? Well, there's a few reasons why this is good. Number one, for ARK, well, strategically good. Number one, for ARK,

they're going to increase developer adoption. And there are two ways that they're doing this, right? Number one, they're available on way more chains. So, you know, whereas their birthplace is Solana, they'll be available on Arbitrum and they will be available on any kind of EVM compatible chain through their integration with Abstract. The second important thing is it's going to make it easier to build on their framework, Rake. And as I mentioned earlier, David, it's coded in Rust,

It's less accessible to developers there. But if they make these integrations with an EVM compatible wallet or infrastructure like abstract, or if they partner with Arbitrum, it's going to make it easier for developers to come and be like, okay, well, maybe let's let me tinker around with this. Maybe there's like some kind of abstraction layer, which doesn't

me to code or interact in Rust, but it can kind of like communicate what I am trying to build into Rust so that I can create an agent. So a really smart and strategic move from their end in terms of like increasing developer adoption. The second thing is it leads to more experimentation, which leads to more innovation, David, right? So think about this, like a lot of developers

blockchains, if we were to use this analogy, that have launched in the past that, you know, we may have heard of back in the day in a previous cycle, but we've never heard of again. One of their main issues was acquiring a strong developer mode. You know, it's why Ethereum is so strong today. It's why Solana is so strong today. It's why base L2s and other

other L2s are so strong because they were able to acquire a bunch of developers that want to build stuff on them. Arc is taking a similar strategic approach where, whereas it might have happened more organically in leading protocols like virtuals and AI 16C, the truth of the matter is it's getting more competitive, David. Mm-hmm.

And so these new leading protocols that have something unique still need to do something to onboard new devs. So it's really cool to see this setup, which allows for more experimentation. And so developers can build stuff and it increases your service area of creating something cool. And finally...

Something that's really cool about these partnerships is that it enhances capabilities for any agents that are built on top of it. So we mentioned earlier, like, you know, the Send AI integration. Well, now any agent you spin up has like 30 Solana plugins, which means that you can do a bunch of different actions natively on Solana just by hooking up into certain APIs.

which I think is super cool. - I wanna ask about, so we're talking about how these AI agent platforms are kind of like, they've been considered the AI, the layer one trade of the AI agent cycle. - Yes. - And we just spent a lot of time talking about Arc. I wanna ask kind of why Arc specifically? How do you rank it? There's like virtuals, there's ELIZA, there's other platforms as well. Are we saying that this is like a third? Is that why we're covering, why are we covering this so much?

Yeah, so the reason why... Okay, there's two reasons. One reason why I'm covering it is it's just interesting. I like what the team is building and I think it's very unique. I have no idea what agent teams are building on them. I just like the approach that the founders are taking. For those of you that are active on Twitter and look at my tweets...

probably 30% of my tweets is, hey, nice to meet you. Can we jump on a call? I would love to learn. And that's what I did with the ARK team. And they were gracious enough to do a call with me. And they taught me about everything that I didn't know. And, you know, I was able to kind of then convey that knowledge to listeners on this show to kind of like share my excitement. I'm like, hey, like, check this out. This is this is super cool. So so I got that's one reason why we're talking about ARK. And to be honest, it should be the main reason because

things that are being built by cool teams, I think should be voiced here, right? Number two, I do think it's gaining enough traction. Now, David, it's been around for about a month and a half, at least publicly, where I think this has the potential to join some of the bigger league protocols out there. Now, to be clear, virtuals, AI16Z, these are the kingpins, right? They have established themselves, they are leading this sector forward, and they are some of the best teams ever, right? But

I still think there's chance for a few others to emerge, David. And that's not for me to be like, you know, I think, you know, it's because they're offering something so unique. I just think this space is so underexplored right now. I just think that there are so many things that we don't know from a design perspective, whether it makes sense to build these agents on, that we should be paying attention to some of the bigger teams that are coming. And I think ARK has the potential to, you know, take that third or fourth position or whatever, you know, it might be. Cool.

All right, well, let's talk about AI16Z because that is maybe call it number one, number two, at least the ELISA framework has the most developed options. So getting into the world of AI16Z, give me up to speed. What are we looking at here? Okay, so as you mentioned earlier, David, AI16Z had...

Well, the entire market had a volatile week, but the AICZ token in particular dipped from a peak of, I think, $2.50 all the way down to a dollar, right? With a lot of, frankly, in my opinion, unjustified FUD on the timeline. You know, it seems when markets are down, the critics are out with their pitchforks in full force, right? I don't think the scrutiny is bad at all, actually, as long as it's well-reasoned. And what I wanted to spend a moment on

on here is digging into what AI16Z's ELIZA framework, which is their agent framework, their famed esteemed framework, which is number one trending on GitHub, is now and what it could be. So what I want to talk about is what I think it can do now and what it could eventually do, right? So I want to bring up this post from the guys at Reforge who, contacts here are like

basically, former engineers that are now investors. So what they did was they did this kind of like deep dive into Eliza, where they kind of went into the GitHub, and they wanted to see like, okay, what can this thing actually do? And there's some important lessons from here, right? And let me give you some of the takeaways. Essentially, Eliza is a great framework to use if you want to build something that they refer to as a personality agent.

This is an agent that can tweet, socialize with humans across multiple social mediums like Discord, Farcaster, TikTok, whatever it might be. Another advantage is it's really easy to spin these agents up. And they're fairly extensible in the sense that there's, you know, 40 plus plugins now available on Eliza that, you know, would allow these agents to do a number of different things, right? You know, they could post on TikTok, they can, you know,

trade or do something, they can like create their own wallet, they can, you know, have memory that kind of like matches across all different platforms, you know, stuff like that. But

They claim what it lacks. So this team claims that what AI16Z and the ELISA framework lacks is the stuff that's needed to build an agent that's truly autonomous. Things like self-learning or strategic planning. For example, they mentioned memory systems, right? And how an agent can store its memory in a database. Think of this like...

storing files on Google Drive, right? So you have like a memory or context of something that you've written down, a note or a document, you know, that's pretty much what Google Drive is. It's a compilation of things that we want to kind of like remember and revisit.

So imagine that this agent has this memory database, which it can on ELISA, but it doesn't yet have the capability to determine which memories or which parts of memories is good to use in different scenarios. So think about this, David. Imagine you have like this database of all this important knowledge

but you don't know which parts are important and you don't know when to pull it out. So you can imagine on paper, these agents look really cool, right? It's like, hey, you can spin up this agent. They're going to have a database stuff. Great. But the coordination layer, the orchestration layer is,

is a little lacking right now. And so what I expect to see is these frameworks kind of develop on that. Can I summarize what I think I heard? Just make sure I understand it right. These guys, these guys from, what is it, Reforge? They popped open the hood of the ELISA framework to kind of like,

Yes.

at the very beginning of this whole AI agent bull market, that like innovation cycle that we're doing, where people in crypto are going to like,

create a lot of attention and awareness around things. And that's kind of where the incentives lie. Let's create, first we're going to create the influencer agents before we create anything else. And what these guys are saying is like, yeah, the Eliza framework is good for making influencers. It's less good at maybe some of the more ambitious hopes and dreams that I personally have for this AI agent space. Like what I would like out of an AI agent, Ejaz is this agent that is truly autonomous and

That is kind of like out of Pandora's box where it is has some sort of arc to it that cannot be undone. It can't necessarily be turned off. It makes its own choices. It goes in its own direction. It has some notion of free will about itself or at least some like, you know, directional outcome. Whereas like a lot of these agents are just like,

they're bots that are designed to make creative tweets. And so what we're missing is some components to really make these things have a little bit more life. And they just looked at the Eliza framework and like, right now it's kind of good for making snapshot in time influencers. Is that a fair analysis?

Yeah, I think so. And what you're saying is we want like basically little baby human robots. Yeah, we want these things to be a real boy, not tied to strings. Yeah, exactly. One thing I will push back on slightly, David, is, and I was very specific in the way I phrased my kind of like framing of what the Reforge guys put out here, which is for now,

For now. In its inexistent form. For now. And I kind of want to bring up a tweet that was put

put in response to this from the guy himself, from Shor, the founder of AI16z, which basically gave his side or his opinion on what the framework can do right now. And rather than go through this entire tweet, because it gets quite technical, the TL;DR is it's more nuanced than we think. It's not just like, hey, you spin up a social agent and you can't do A, B, and C. There are certain contexts or environments or setups

that would enable the agent to do more things, but it's just not as accessible plug and play as we want right now. So the argument that Reforge was making was like, hey, you can spin up these personality agents, but you're going to have to do a lot of custom build work.

And Shaw is rebuttaling saying, you can kind of like do what you need to do on ELISA if you're aware of how to do it. We make it easier. So basically, if you're an AI ML engineer, you shouldn't have an issue doing the kind of like crazier things. But it's that knowledge gap, you know?

is a plug and play platform where you just click a few buttons and launch a really super smart agent better than having something that is a little more nuanced and technical right now that will eventually become something easier to do. So I emphasize for now in its existent form, Eliza may be limited in those regards, but they're supposedly meant to be shipping Eliza V2, which will have a bunch of upgrades that will enable this platform

way more autonomy, basically, for agents. But let's step back for a second, David, and kind of like ask the bigger question, because you kind of just spent some time there describing, you know, what these agents should be able to do. But the audience might be wondering, well, why is autonomy even important?

Well, firstly, if we don't progress autonomy, then all we end up with is these rigid agents, which appear pretty novel at the start, but kind of get boring real quick. They're inert. Autonomy, yeah, yeah. Autonomy, to your point earlier, will make these agents more human-like. You know, they'll be able to remember things, have context on situations and make decisions for themselves instead of some human needing to kind of like spoon feed them like a child every other step. And this ties to a wider point that I love.

to see this sort of friendly, transparent debate happening between two people making frameworks. Do you know what I mean? It's forcing us to think deeper and push this sector's tooling to its extremities, right? I put out this tweet yesterday which speaks to this, which is like, you know, catching up with everything that's going on. I've seen a flashback

flurry of teams launching these agent platforms and agent frameworks. David, if you remember, like on episode two, there were like three frameworks. There were like three platforms. Now I think we're at like 50 plus and counting. Like they're like two launching today. 50 platforms? 50 like quote unquote layer ones for agents. Yeah, yeah. Like loads of, you know, big teams and niche teams that are building and launching these frameworks and testing it out. And it's crazy to see. Now that tells me two things.

Firstly, there's a lot of excitement about this. And I think that there's a lot of potential that could come from it. Like we're onto something. We don't quite know what it looks like, but we're onto something. And we're trying to figure out what these platforms should do and how they should be designed to enable what they should do. Right? The second thing this tells me is we haven't quite figured it out. Mm-hmm.

So it's equally like we've figured something out, but we haven't quite figured the thing out, right? And the reason behind that is if we have figured the thing out, I think there'd be fewer platforms that would be gaining mindshare. There would be five or less, not 50. There would be five or less. Also, one big reason as to why there's 50 is that there's money in it as well. There's money involved. Yes. There's money involved. There's money in sentence. Yes.

Yes. Yeah. Loads of incentives like that. So I think it is a mixture or cacophony of money that people could potentially make from these things by launching one of these things. Innovation that I think a few hard devs are like working hard to kind of like figure out. And then just a bunch of curiosity, you know, like we go back to that mindshare, David, where 50%, if you want to add AI and DeFi AI, that's like over 60%.

of Mindshare, which is just insane to see. So there's a lot of eyes on this space and we owe it to ourselves to try and build something cool here. So really excited to see that kind of stuff. Anyway, moving on with this kind of like general thought around AI16Z, I wanna kind of like spend a moment on their white paper. So instead of like getting into the weeds of this and speaking of future iterations of what this framework might look like, the team put out this white paper

kind of talking about the next version of their framework and what they could potentially achieve. Now, saving everyone the kind of like time of, and honestly, boredom of me kind of like running through this in its entirety. That diagram you have pulled up, David, is pretty helpful there. Basically, these agents are going to get access to a ton of different things. And it'll all be kind of like harnessed through the framework, through the platform that they're building, if I'm understanding this correctly.

Which means that the whole logic, memory system, reasoning, strategic planning that we spoke about earlier will become embedded so that developers will be able to do something with it. So I just wanted to pause momentarily and mention that.

I also wanted to bring up another tweet, which was kind of like from a week ago, David, but it's on how Eliza actually works. And I think, yeah, our favorite dev, Saiga, who's been like in the weeds, kind of like explained this really, really simply. And like, so...

without getting into the kind of details of this, he basically describes the AI model itself, you know, so your chat GPTs and stuff as being the brain, right? So think about it as kind of like the brain setup where it's thinking of different things, right? And it's like processing data and it's trying to figure out

okay, what can this do? How will this work, et cetera? Next, we have like the kind of like database section, right? This is where he says the agent stores relevant information for generating responses, including previous tweets, interactions, stuff like that.

And then you have something called runtime. And this is the final thing that I'll make, which is it's kind of like the processing environment. So if you think about the LLM being the brain, if you think about the database also being included in that brain, then runtime is your body. You know, it's your nervous system. It's everything composed together.

The Arbitrum Portal is your one-stop hub to entering the Ethereum ecosystem. With over 800 apps, Arbitrum offers something for everyone.

Dive into the epicenter of DeFi, where advanced trading, lending, and staking platforms are redefining how we interact with money. Explore Arbitrum's rapidly growing gaming hub, from immersed role-playing games, fast-paced fantasy MMOs, to casual luck battle mobile games.

Move assets effortlessly between chains and access the ecosystem with ease via Arbitrum's expansive network of bridges and onrifts. Step into Arbitrum's flourishing NFT and creator space where artists, collectors, and social converge and support your favorite streamers all on-chain. Find new and trending apps and learn how to earn rewards across the Arbitrum ecosystem with limited time campaigns from your favorite projects. Empower your future with Arbitrum. Visit portal.arbitrum.io to find out what's next on your Web3 journey.

Uniswap Labs is making history with the largest bug bounty ever. $15.5 million for critical bugs found in Uniswap v4. This isn't just any update. Uniswap v4 is built with hundreds of contributions from community developers and has already undergone nine independent audits.

making it one of the most rigorously reviewed codebases to be deployed on-chain. And with 2.4 trillion in cumulative volume processed across Uniswap v2 and v3 without a single hack, the commitment to security and transparency is rock solid. Now Uniswap Labs is taking an extra step to make v4 as secure as possible with a $15.5 million bug bounty.

Head to the link in the show notes to dive in and participate in the Uniswap v4 bug bounty. All the details from eligibility and scope to the rewards are there. All right, so that's a wrap on our section on Eliza and AI16Z. Let's get into virtuals, the other very large platform. But I want to actually start with AIXBT, which is still to this day the number one agent out there, again, born out of the virtuals platform. On virtuals. Yeah, from virtuals, yeah. And so there's been some speculation about CryptoPunks lately.

because there was big price movements. CryptoPunks are up like 20% on the week. And in ETH terms, there's rumors that CryptoPunks are going to, the IP of CryptoPunks is going to be sold by Yuga Labs. People do not know where to. So this one punk just asks this question and tags it.

AIXBT, who's of course everyone's favorite degen agent on crypto Twitter, giving out facts and figures and metrics about crypto things. And the AIXBT responds to the requests about any knowledge, any rumors, what do you know? And it goes, AIXBT just responds completely confidently because that's his only mode of operating.

Yuga's selling the IP, Kevin Rose is buying it for $65 million in ETH and stock. Wild times and markets don't know what to do with this info. Wait, wait, wait, hang on a second. You're telling me the reason why I saw a bunch of Kevin Rose buying punks memes on my timeline was birthed by this agent? By this rogue agent spitting out absolute falsities. What?

Wait, I have no idea. Absolute falsities, yeah. I didn't know it was cracked by this agent. I didn't know the rumor was started by this agent. Yeah. Wait, what? So Zeneca, who's a known NFT market participant, NFT influencer, I don't know if he likes the influencer label or not. He just goes satirically, breaking Kevin Rose to buy CryptoPunks IP from Yuga for $65 million. Source, Zeneca.

not at all a hallucinating AIXBT. So AIXBT just dropping just false facts, fake news on the timeline.

But also look at the influence it has. That's crazy. A younger version of me would have seen that. Would have believed that. Everyone's posting it. And the fact that this kind of has this butterfly effect is crazy. Oh my God. Okay, so Kevin Rose is not buying CryptoPunks. I mean, if Kevin Rose is buying CryptoPunks, if that is true, that will be one of the most insane calls from an agent that will have ever been made. Yeah.

That's insane. Okay. Just to be clear, we do not know who is buying CryptoPunks. Yes. Or if that's even being up for sale at all. I think the rumors are legit, but the facts are just still waiting. Yes. Yes, sir. Yes, sir. All right. Get us into the rest of virtuals. What's going on in the virtuals land? Okay. So first, I want to start with $350 million in assets.

annualized revenue for the virtuals protocol and this is according to uh the masari deepin report um deepin stands for decentralized physical infrastructure network i'm not sure why it's included in that because i think wrong category this is specifically a wrong category but nonetheless cool to cool to show um

it just dwarfs any other protocol that isn't an L1 at this point. Hint, hint, it could be an L1. Hint, hint, another reason why it should be priced as an L1. But just astounding. David, if you remember, I think, actually, I think maybe it was the roll-up that I did with...

Ryan when you were out and Ryan filled in, it was $35 million of revenue at that point. And I think that that's like skyrocketed. And again, to emphasize this is annualized, but the fact that this kind of figure keeps growing is just insane and is another reason why like just virtuals is just one of my favorite protocols on here. Another thing that I want to kind of flag is

It's not just about fees. It's about what they're doing with these fees, David. And it was just announced literally yesterday, or maybe actually under 24 hours, they're doing a $40 million buyback and burn accrued from pure trading fees. Now, for some context here, when you create an agent on virtuals, a token is made.

And that token is paired with their native token, virtual. So let's say the agent that you create for liquidity. Kind of like how ETH is paired with USDC or any other token in the Uniswap AMM. Virtual has replaced ETH in the AMM. So it's virtual, the token, mixed, paired with the agent's token. Kind of bumping out ETH as the money of the virtual's ecosystem. Pretty good move. Yeah.

Yeah, so virtuals is money. Virtuals is money in the virtuals ecosystem, basically. So you're right, David. And why did they do this? So people can buy, trade, or interact with this token. They can acquire this token. So really smart, really interesting. But one of the understated things about this model is whenever you trade, there's a fee that's taken. Mm-hmm.

And a portion of this fee accrues back to the virtuals token itself. So it's kind of used to acquire virtuals and it puts a lot of buy pressure. It's a really smart kind of like tokenomic flywheel and system that they've set up. But then the question becomes, okay, what are you doing with these fees, right? They're generating revenue in the treasury at a rate of $500 million a year annually.

$350 million. Excuse me, $350 million. Yes, exactly. And so the question is, what are they doing with it? Well, they've just announced that they're going to do a buyback and burn with $40 million worth of their token virtuals, right? And

Just to kind of put things into context here, that means that they are going to purchase $40 million worth of virtuals and burn that, if I'm understanding it correctly, which means that it puts a deflationary pressure on their virtuals token, which means that everyone that is kind of holding the token, theoretically or economically,

price should, you know, accrue to the tokens that you're holding because there are less of those tokens available in the supply. If it's not already, of course, priced into the market already. I think one update to that, I think it's not just the virtuals token. They're also buying back

agent tokens as well. I think that's true because I also saw that the agent category, the virtual's agent category yesterday just went through the absolute roof. Let's see if I can find the category. But I think this is why AIXBT is basically at all-time highs right now. So all virtual's agent tokens got a very strong premium boost to their valuation yesterday.

Yes. Yes. So actually, thank you for correcting me, David. You're right. So this $40 million buy pressure isn't just to buy back virtuals tokens specifically. It's going to be used to buy back agent tokens, agent specific tokens. So look at the charts right here on the right. Like it's a normal chart. And then yesterday was announced and everything. It has like a very vertical line up. So virtuals is up 7% or excuse me, 10% in seven days, but 30% in 24 hours.

AIXBT is up 90% in seven days. Vader AI up 50%. So this announcement was, you know, if you had a bunch of virtual agent tokens, you're pretty happy right now.

Yeah, and just to kind of explain why that's the case, it's like, okay, let's say that AIXBT token on virtuals was traded the most. They're going to have the most fees, right? And those fees are going to be used to buy back AIXBT and burn it. So think about the deflationary effect happening on all of these virtuals agent tokens, which should accrue value to those who are holding those tokens itself. So just to...

crazy smart move, in my opinion, by the team in terms of like, seeding liquidity and value back to their token holders and to their community, which I think ultimately leads into a very loyal kind of community, which pushes and wants to see your team kind of like win. And I think, you know, the team being led by Ethermage and

His co-founder is just so, so impressive. And I'm excited to see kind of like where they take it. One thing I do want to bring up, David, is another thing, which is they're also starting to accrue fees in CBBTC. So CBBTC is Coinbase BTC. So yeah, they're announcing it. They're buying Bitcoin. And so one might ask, well, why?

Why aren't they buying virtuals, right? Well, think about it, right? If they acquire these fees and they want to use it to kind of like supplement other teams, if they just sell virtuals to support those teams, it's not going to look great, right? So, you know, it's going to put a downward selling pressure on virtuals. But if they were able to buy a much more liquid asset like BTC, then they won't have that issue anymore. So it's kind of cool.

Another thing I want to point out is what are they doing with the CBTC? Well, it's going to get split amongst creators and agents themselves. So, you know, think of this as like an economy where people who live and thrive in that economy have to pay taxes. So the same way that humans pay taxes for the work that we do, agents in the virtual economy have to do the same. And these taxes get kind of like bundled into like a treasury, a tax treasury, and

And typically a government will decide where and how it gets distributed. The way it works in virtuals is they have like a high level framework, which says 30% goes to agents, 40% to the creators, and it just gets distributed automatically into their wallets. Really kind of like efficient system. And, you know, that's what we're calling out with this CBBTC thing. Yeah.

And then I guess the final thing that I want to point out is less around the tokenomics, but more around just like Jansen's vision for this entire thing. I don't know if you have the post, David, but EtherMage kind of posted this about

I don't know, like maybe less than a week ago or so, talking about like how agents are, virtual agents specifically, are taking their next big step into gaming. And like, if we just read through this, right, look at this. Agents in gaming were how many of us researchers began exploring autonomous agents. Open world environments, diverse action spaces, and multi-agent systems served as the perfect testbed for testing true autonomy.

If your agents can pick up a gun in the game and make a conscious decision whether to fire it or not, use it as a warning shot or even gift it to their spouse, it is not just a bot. It's an agent. And he goes on to kind of explain how and why

virtual specifically is the best setup for these agents eventually becoming, you know, these high powered gaming agents, right? As like one example use case. And they're partnered with some of the biggest names. We spoke about Illuvium last week and a few others. But I think the wider point here is a testament to the team. If anyone was around kind of like studying the virtuals team before they became virtuals, they were a gaming DAO.

That kind of like did well. Interesting. Yeah, they were a gaming DAO, David. And then they had to come back and figure out, you know, what can we keep doing? Now they could have packed it in. And we're talking about like a multi-year bear. But they decided, you know, we have something here. We're going to keep hacking and build something cool. And that's what ended up becoming virtuals. You know, they saw the vision of these kind of like gaming characters that would have autonomy and realized that the infrastructure that they were building were for agents. So it's like super cool to see.

I think there's a very strong potential of like a mutual growth in the sector of AI agents and gaming. Crypto gaming has been a subject for years, and I think nothing really has taken off.

There's more narrative and imagination about what crypto gaming could look like than what's actually been shipped. Maybe a catalyst could be the introduction of agents into the world of crypto gaming, where crypto and gaming and AI agents combined really create something powerful. I know developing in the gaming world is slower than it's developing in crypto or slower than it's developing in AI agents because gamers are very...

needy. They have a very high bar of quality in order to consider a game fun. But you can easily imagine something there between the intersection of crypto gaming and AI.

Yes, yes, absolutely. And I think like, you know, we spoke about this last week when we wanted to kind of focus on more emerging trends. Teams like, I'm forgetting what the name, HyperFi, I think in the AI16Z world was, you know, is building like this, this like really easy platform to kind of like launch virtual worlds. We saw that, I think that there was a Roblox partnership between, I think, I might be confused here, maybe virtuals and Roblox, or maybe it was AI16Z and Roblox. But, you know, the,

there's a huge amount of interest of these gamers or game developers to spin up things with agents. And I think that's a natural overlap, David, you know? And I kind of want to hold us accountable here because like,

i'm sure like uh bankless said similar things um back in the nft cycle right but i have we had many gaming episodes we had many nft episodes some things came into fruition most gaming statements did not come into fruition right so i want to hold ourselves accountable and not like hype this up into something that you know might be something similar but i do think that this time is different in the sense that these agents are way more dynamic you know last cycle we had just static images and now we have something much more dynamic um

that we can interface over this. So super cool. I think you and I both see potential here. And I will also say, don't hold your breath. This will take a lot of time to get right. This will be a multi-weekend. What we are going to do on the AI Rollup is we are going to watch the development. And then when the final products are at all ready, it's not going to be this year. It's going to be next year at the earliest, at the very earliest.

Yes, yes, agree. Like these are in their very early form and you should be very cautious around research and investments that you do based on any of this, to be honest. Yeah. Okay, Ejaz, we are at the one hour mark, which is where we want to keep things. But before we end things, just like rush us through the last little bits of development over the week in case anyone wants to go and do some more homework. Yes, okay. So homework for the crew this week. There are two...

teams or protocols that just popped up and dominated the timeline this week. And, you know, that's either for good or for bad, but they were taking Mindshare. And I want to kind of like shine a little bit of a light as to like what they are and like, you know, how things have manifested so far. So the first team is, or like the first protocol is called HoloLens or HoloWorld, which has a token called AVA. And

The point around this ecosystem is it's another agent platform geared around kind of like more virtual agents. Now, we just spoke about gaming, David. This is less gaming to start off with, but it's more gamer-esque. It just means that instead of having agents that interface with Twitter or

or write written statements, these agents will be able to speak. These agents will have a virtual body that can maybe appear on a TikTok video or on Instagram as a picture or whatever that might be. So this team has been kind of like focused for quite a while, supposedly building out the infrastructure to enable a bunch of these agents. So it's a different kind of medium to explore and definitely want to kind of keep an eye out. I personally need to dig in more

before I can speak more to like what they're building, but super, super cool. The second thing I want to dig into is something called Pippin. Now there's a lot of controversy around this. So I'm going to kind of like give my quick take and we can move on. But Pippin,

is another agent platform or framework that was birthed from this guy called Yohei. And I might be mistaking his name, so forgive me if I am. But Yohei is basically this part-time VC. And I think it's untapped VC. And also at night, you know how the theme is...

I'm Batman at night. Yo-hai is building agents at night. And he actually built this experiment called Baby AGI back in the day, which got a lot of traction. People were like, oh my God, this guy knows what he's doing and he should build something super cool. And so there was a lot of hype as Crypto Twitter does.

hyping around this platform, hyping around this framework that he was building. And what happened was, you know, it was being pitched as like this novel thing. It was going to outdo AI16Z, all these like major claims. And then there was a demo day, which, you know, Yohei and the team, I'm sure like,

did their best to kind of deliver what they wanted as a V1. But because the market had hyped this stuff up so much, you know, it led to a lot of volatility in the token. So I, to be clear, I don't know the kind of credentials and benefits of the framework. I need to get more involved and look into the Git hubs of all of these things. But it just shows it's such a reflexive environment right now, David. And it's important to keep an eye on these things, but not to take it, you know, to heart so seriously. Yeah.

Yeah. Yeah. I think as we get further into this, as we get into further into all bull markets, you see the quality starts to decay and it gets harder to identify signal from noise. And that kind of sounds like that's kind of where we are at with the current currently being introduced, like new frameworks where there is. And it's also very easy for crypto Twitter to overhype something, as we have recently learned.

And so that kind of sounds like that's still in that gray area. But I also do appreciate you bringing things here to the AI Rollup. That is truly on the frontier. You know, this is not for everyone. Some things that we talk about, I think, on the AI Rollup are going to blow up. And that's just going to be how it works because we're on the frontier. That's what we're here to do. Yep.

It's been super fun, David. Cool. Awesome. Thank you, Ejaz, for guiding us through once again to another week of the AI rollup. Bankless Nation, you guys know the deal. Crypto is risky. Crypto AI is even riskier. We are specifically looking for things on the frontier. That's what we're doing here on the podcast. It's not for everyone. All disclosures can be found at bankless.com slash disclosures. We are glad you are with us on the bankless journey. Thanks a lot.

you