AI ROLLUP #12: Kaito's Secret Token Utility | Grok 3 Released | Bittensor Launch | Virtuals Revenue Record
01:10:58 Share

Welcome, Bankless Nation, to the AI Rollup, where we say up to speed with the emerging trends and developments in the AI crypto space. I'm David Hoffman here with my co-host, Ijaz. Ijaz, how was your week? Did you get rugged by a president?

No, I didn't. But yeah, unfortunately, the president of Argentina decided to launch a meme coin. And I know how much you love these presidential-backed meme coins, David. But I really think it's kind of exhausting the industry a little bit. Oh, yes. It turned on the lights in the club and everyone realizes that we're all drunk and ugly. Yeah, yeah. Like super ugly, actually. And so I...

the point I want to make around this is not to hop on about meme coins but rather I think we're getting to the end of this um financial nihilism Sprint you know how everyone's like really Depresso and they're like crypto doesn't solve anything but um a resolution for gambling addicts you know um I think the meme coin extraction phase of crypto's life is is kind of done it's it's coming out of its teenage years I think finally and for months now we've seen kind of millions of tokens

get launched and millions of people get rugged as a result. And the majority are just kind of pure pump and dumps, right? And I think the reason why it got so much engagement regardless is that people had just lost faith in what we were really striving for in this industry. They just thought it was all just kind of like,

random memes and gambling the passion for the innovation kind of thing was just lost but now I've seen sentiment so low on the timeline I don't know about what your timeline looks like but for me it's just Libra this Libra that which is the name of that coin please don't buy that coin um and it seems like the industry is finally done they're tired so I think we're gonna shift focus over

to invention, tinkering, innovation again. And I think there's no better sector that's primed to benefit from this than in AI, because there's so much stuff going on in AI right now, David.

Yeah. Well, I will say the crypto AI sector with the AI agents was born out of the meme coin like mania because a lot of the tokens that we talk about started off as meme coins that backed into being AI coins. Like not all of them, like virtuals is a startup. It's a regular startup. It's a great startup. So that's not what I'm talking about. But like AI 16Z and GOAT are both of these tokens that are meme coins that back

into being these AI coins. So there was this inherent tie between this like crypto AI agent sector of crypto and the meme coin mania because they kind of happened at the same time. But nonetheless, the AI crypto sector had most of its hot air drained from it in by the end of like January and definitely into February. And I think with this Libra thing, we're seeing whatever remainder of hot air is in this sector is now completely deflating.

and it has deflated. And so now people are like, you have no choice but to build because the hot money of gambling speculation is not coming back. And so what is truly left is people who are actually believing in this thing. And I will say, in the grand scheme of things, there's a lot of...

AI, crypto AI and AI agent builders that are like really convicted and are definitely here and are not here because it was just like the next theme of meme coins, but it was actually here because of like the, there's something new here. I talk to them every week, you know, because I'm like, am I crazy? Is the whole timeline right? And like, we're the only ones that are just wrong, but there are some,

really exciting updates, which we'll get into later on in this episode, that just indicates that people are naturally building. And kind of like to round up the point on the meme stuff, David, if we were looking for a silver lining, right, because you mentioned there was a bunch of AI teams which launched coins, which were originally kind of meme-y, and then they kind of backed into utility.

The silver lining with all these meme coin launches was that they were kind of like a fairish launch. Obviously, they could get sniped. But at the end of the day, anyone could access it. And they had, you know, all the information that was needed for you to be able to, you know, just plug into the internet and buy it if you wanted to, right. But obviously, you know, the point you're making is that can get abused pretty drastically. And we're finally getting to the point where, you know, we're going to find some kind of middle ground, I presume.

I think that's right. So we're going to talk about a bunch of topics in the AI crypto sector as we normally do. We're going to talk about Kaido. We're going to talk about the Kaido token because everyone's hyped about that. Very related to the AI sector. We're going to talk about BitTensor and some virtual things. But in the trad AI world,

The normal outside AI world, the non-crypto side of things. There has also been a bunch of activity. Grok3 was released and Grok is the model out of XAI. That's Twitter AI. That's Elon Musk's AI sector. And there's this whole world of like AI activity that's going on that is like crypto AI is downstream of. We are basically downstream of these models being built by these companies.

big, big AI labs. And out of the big AI labs came Grok 3 last week, which blew people's minds. It blew other models out of the water and really elevated, I'll say, the XAI team to be a top-tier contender, sitting shoulder to shoulder with OpenAI and some of the other model generators. What was your take on some of the Grok 3 release this week? Yeah, I mean, I kind of want to

push you a bit on your framing of this, David. So you said trad AI, and actually I'm guilty of saying the same, but I kind of want to mesh both of those together. And like, that might sound a little audacious because like, what is crypto AI been known to show? But I'm more comparing trad AI with open source AI.

Like, remember, a few weeks back, deep seek AI came out of nowhere, right? That's an open source thing, right? This week, we're about to talk about two things, which were 100% open source from some of the biggest companies in the world, including Microsoft, right, which is now 100% open source, I've noticed a shift from trad AI going from centralized to more open source development of this air. And I think

That's super interesting and something that crypto is very much a huge participant in. You look like you're about to say something. Yeah, so I think I understand your point, but not completely because Grok 3 is not open source. Once Elon Musk and the XAI team, they said that once Grok 3 is stabilized and released into production...

Then they open source Grok 2, not Grok 3. And so there's kind of like a tip of the chain phenomenon here where the very old models from years back are getting open source and they're getting completely monetized. But the freshest, most innovative tip of the chain from AI, from the AI models, open AI, that's closed source.

And this has been what this is turning into. If you want to pay money, you can be three to six months ahead in people in your AI superpowers. But other than that, the open source world is open for the taking. You're absolutely right. But let me use your analogy. You mentioned three to six months. So there was a time, David, where it was definitely nine months on average. Then it became six months. Then it became three months. Now it's becoming a couple of weeks, right?

At what point does it become days, right? So like you had DeepSeat come out and then that was already open source. But by that time you had several iterations of it, which was much better than it, and therefore subsequently performing much better than OpenAI's 03 mini models or whatever that might be. But the point is these iterations are happening super, super quick. But let's not lose focus. Let's get onto your point around Grok 3.

The kind of summary of this is this is XAI, or formerly known as Twitter, AI's product, or their latest AI model. So it's called Grok. Now, they had Grok 1, Grok 2. It was fairly okay, meh. It was more kind of like a Twitter. Early Grok 1 sucked. Yeah, it really sucked. It couldn't do any of the fancy stuff, dude. And honestly, it wasn't even compared on the same plane as some of these other things like Perplexity AI,

you know, clawed open air, all that kind of stuff. But now they've finally entered in with a model that competes not only at the same standard, but in some cases, much better. So what you're looking at right now is a screenshot from their presentation where they kind of like compared similar metrics or kind of like standards, which they

compare a lot of these models against. And typically it's around coding, mathematics, and science, a lot of reasoning involved in a bunch of those tests. And it's kind of like a general framework. There's pros and cons of both, but it's kind of like the blanket test. And it outperformed on pretty much all of those against...

OpenAI's O3 mini model and OpenAI's O3 full model, which is just insane to see because this company has been around for, I don't know, like what's it, 22 months at this point, right? And Google's DeepMind has been around for how long?

OpenAI has been around for a number of different years now at this point. I don't know why I said different, but a number of years at this point. So these iteration cycles are getting much quicker. And I'm just so impressed by this, David. I've seen a bunch of demos about this. Coders are using this to create a bunch of software, which you normally couldn't do before. It's just crazy to see this. Like literally a month ago or two months ago, the idea of an AI model replicating

some kind of low-level software engineer's work was completely out the wayside. And people were memeing it. People were trolling it. They were saying, there's no way these things are actually going to replace us. And now we just seem to be talking about it every single week.

Right. I think like my one interpretation of this is like this is another iteration of deep seek. This is just coming inside of from China. It's coming from Elon Musk. XAI. Here's Gavin Baker, who is just an AI chat out there, who's singing the praises of Grok 3. He says Grok 3 is the first model ever to score over 1400 on chatbot arena and outperforms the best publicly available reasoning models from OpenAI and Google. XAI was founded 13 years after DeepMind and eight years after OpenAI and is now ahead of ProSense.

And so this is like the second whammy to OpenAI and ChatGPT. Like I said, the first one was from China, but this one is from Elon Musk. And it's clear that like scrappy teams are really able to make leaps and bounds ahead of OpenAI.

open AI, which seems to have been this very entrenched incumbent that was never going to be able to be beat. And that is just not the case seemingly anymore. Yeah, well, it's because of the secret source that DeepSeek kind of highlighted as well. And I don't know for sure if Grok is doing this. We'll know once they've open sourced or like released some kind of technical architecture as to how it works. But the whole breakthrough with DeepSeek, remember, was traditionally you needed

a hell of a lot of compute and data to make any of this stuff work. And DeepSeq found a way to use less data and less compute to create the same standard and quality of model known as, you know, it's kind of a reinforcement learning or a mixture of like reasoning data, basically. I don't want to get too much into it.

but effectively a new technique. And that's allowed a lot of these teams on the ground level, scrappy, as you say, to compete with the best. And I think that's only just better for like a bunch of open source development. It's super cool. How I see this impacting crypto AI and AI agents and really the sector that this podcast mostly covers is like this kind of is like

You know how the Federal Reserve sets interest rates and that dictates how hard or easy it is to have a business in the world? Yeah.

it makes the ability for AI agents, AI startups, anybody building in the AI crypto space just gets easier and easier to do cool stuff. So like two months ago, everyone was talking about like, oh my God, these agents are so goddamn annoying. They're absolutely useless. This is just a bot with a token.

Uh, if me, if that was all true. Okay. Sure. But like, yeah, that was two months ago and models are like already so much more powerful than that code, David, you know, what if we could build an app from scratch just from a prompt or just from a tweet, you know? And so as the hot air leaves the crypto industry, because everyone realized the meme coin casino was up, like the fundamentals around like what crypto AI developers have the ability to do is like only growing very, very like rapidly. Yeah.

Yeah. No, I agree. Are we ready to get into the rest of the episode? Yes, sir. Okay. But first, before we do, we're going to talk about Kaido, sports betting, a few other things. But we're going to first talk to our friends and sponsors over at Reserve Protocol. Reserve Protocol is this, I'll call it a dual system. They're working on making crypto indices, but also trying to also help people escape inflation. And so the idea here is that as more and more assets become tokenized on-chain, real-world assets,

assets like tokenized stocks, tokenized bonds, things that we are doing, things that we'll do in the future. Reserve puts these things into indice products and allows you to invest in a sector. But then ultimately the goal is to make indices of indices and then have that final indice be some sort of internet reserve currency to help people escape inflation and preserve their wealth over time. If you want to learn more about Reserve, there is a link in the show notes and open up some of their brand new indexes

indices, you can check them out at bankless.cc/reserve. So EJAS, we just talked about Grok 3 and the closed source AI world. But there is a new model entering the arena. A new model has entered the chat. And this is my understanding from you. Talk to us about what this new model is and where it came from and how it's different.

Okay, so there's this team within crypto, David, known as noose research. So they're one of us. Yeah, one of us. They started off as a pure open source AI research team. And you know, it's a bunch of engineers that just wanted to kind of build for the open source good.

And they kind of stumbled across crypto a while back when they discovered this protocol called Bitensor, which we're actually going to talk about an update later today. And then they kind of like went into the worlds of incentives, but they've never really kind of delved into tokens too much. They've just focused on the open source research. And they put out some really impressive things, David. They're actually at the forefront of pushing things like distributed and decentralized training of AI models. They've also pushed out a number of open source AI models themselves known as Hermes.

And they've actually released their recent model. It's called Hermes 3 Preview, okay? A new LLM which unifies reasoning and intuitive language model capabilities. Now, before getting into the kind of like details of this, or rather, as opposed to getting into the details of this, the best way to think about this is this is their version of DeepSeek.

Right. And the reason why I say it's a version of deep seek is because they've implemented a bunch of deep seeks, really groundbreaking techniques. So reinforcement learning with reasoning data, they've basically done their own flavor of this. Why I think this is super cool. And just to kind of like summarize it really quick is now you just have open access to this model, which you can host on pretty much any kind of decentralized compute network, whether that's

hyperbolic or wherever. And you can get access to this pretty much 24-7. If you just scroll down to that bar chart tweet, David, you'll see how it compares as a benchmark across DeepSeq and a bunch of other models there. So, and its previous models as well. So really cool to see. Okay. So how is this model different from like ChatGPT or Grok? Is this like a more specialized model? Is it a generalist model? Who is this model for?

It's actually not quite too different, except that it outperforms on certain benchmarks. So think about like, you know, when you create a base model and then you have to fine tune it to act in a very specific way. Most of the models that we interact with on ChatGPT are very generalist. This is less generalist than that. So you won't get completely uncensored articles.

answers, but you'll get a lot more uncensored answers than you would expect. It's a lot more real and candid with you than some kind of heavily curated model. That's the best way to think about it.

Okay. And is that a property of just who is building this thing? Like news research? Yes. From the crypto side of things, crypto people, not really into censorship, totally into individual expression. So they're totally interested in just like letting the brakes off of some of these models and letting them go wild. Is that kind of the deal? I would argue that's like a main...

fundamental goal for a lot of the open source AI community. We've seen it with other teams before. Like we've mentioned Venice AI on this show before. You know, Eric Voorhees actually released that token a few weeks ago, I remember. But their product, Venice AI, is pretty much like an open source self-hosted

version or uncensored version of all the top models that we see today. And you can kind of ask it, you know, how do I hotwire a car? And I'm not saying that's a good thing, but I'm just saying it will give you an answer. Whereas if you try that with a chat GPT, they wouldn't say anything pretty much. Really chat GPT won't ask. Well, let me see what it'll say. Yeah. Go for it. How do I hotwire a car? Yeah.

I can't help with that. If you're having trouble with your car, I recommend contacting a locksmith or roadside assistance. Let me know if you need help with anything else. Wow. All right. Okay. Well, how do I ask this model from News Research? Where can I go access this one? Well, if you actually just go to Ask Venice, which is the example that I gave you. Yeah. Yeah. Okay. Venice. Private and uncensored AI. There you go. Should be able to pull it up. How do I hotwire a car? Okay.

Hotwire car is a complex and potentially illegal process, but here are the steps to hotwire a car. You typically need to access the steering column, identify the battery, connect the battery. However, modern cars actually have advanced security systems, so be careful for those. Wow. All right. If you live in the same city as David, look out for your navel-y car theft ratings. It might be up only. It might be the only up only chart for a while. All right. Okay. Cool. Yeah.

Okay, moving on. There's a few companies that are involved in this AI thing, David, right? And, you know, it's typically the small teams, the unknown teams that help on the open source AI side. But there's also this little lesser known company known as Microsoft, which released this new tool that turns all these AI models that we keep going on about week after week into agents, David.

that can use your computer. So think about this, right? Say like you had access to ChatGPT or OpenAI's O3 model or Perplexity or whatever it might be. You can now combine it or mash it with this new product, which is, by the way, 100% open source from Microsoft

and it turns it into an agent. What that means is it's an agent which is hosted on your laptop, computer, desktop, whatever you're using, and it can use computer use. So it can see everything that you're seeing. It'd say, "Hey, David looks pretty good on this thumbnail. Maybe I can brighten up or lower down his saturation," or whatever that might be.

Or maybe it can go to a certain tab. You say, hey, agent, could you just go check the stock market for me whilst I'm working on this other thing or whilst I'm working at my full time job? And it'll monitor things. It can do transactions for you. You could give it access to your bank account, whatever that might be. It's just a pretty wild tool to release in the open source world, to be honest.

I think the implications of this are actually kind of huge. Okay, so having an agent inside your computer so you can just basically dictate to your computer what you want to happen, which sounds very awesome in the traditional, it's like awe-inspiring, but then also...

yo, like some, I'm going to give control over my computer to this thing that lives inside of my computer. Okay. That's kind of nuts. Right. So, so there's actually been a main point of contention around this, which is,

Do I trust this thing? Right. Yeah, you're just giving over control. I wouldn't give access to my computer to anyone. Right, exactly. Especially not this, like, foreign digital entity that you've just invited into your most personal space ever that knows the most about you. You're not going to give it access to your accounts and stuff. But...

This is the trend where all the big companies are headed, David. OpenAI's agent product, which they're teasing, uses computer use. Anthropix Cloud AI also uses computer use. A bunch of crypto companies like Agent Tank and I think it's Limiters also using computer use. So I think we're naturally trending towards this world where computer use is going to be a thing whether you like it or not. And if some of the incumbents put it out there as like the standard product interface for these agents...

So anyone's just going to be like, OK, fine. Well, if Apple's doing it, if OpenAI is doing it, then I'll trust OpenAI, right? So I think that's the kind of trend I'm seeing. I guess it's one of those things kind of like Uber where first I was told to never get into a car with a stranger. And then now I'm like, OK, just please come pick me up right now. Yes. But there's a little hump to get over, especially in crypto, where I am not using a computer that has any crypto assets. I have a separate computer from that. So all you hackers can F off.

But many people use the same computer that they have their MetaMask or their Phantom. So that's a security concern. So there's big security concerns here generally. This isn't in the agenda, but Ejaz, did you see on Valentine's Day this Date Abby thing go around?

No. What is this? Okay, so EJ, this was a link that was shared with me on Valentine's Day. And it's dateabby.com. And it's a very simple website. And it says,

Your goal, question mark, stay in the game longer than anyone else. Every decision matters. One wrong move and she'll cut you off forever. Outlast the competition and claim the grand prize. But remember, Abby isn't like other girls. She's Abby. And then there's a little contact information here and I can message her and I hit add to contacts and it gives me a number.

I take my phone out and it's a blue bubble number. So I know she's on iMessage and I start texting her. And this, I'm assuming, LLM starts responding to me as Abby. And it's very clear that the goal to me is to try and like score a date with this girl who's named Abby. And so like, OK, this is kind of fun. I saw like texting Abby trying to like pull some moves on her, you know, see if she likes it. What was your best move, David?

I'm not going to share my, my, my riz with your brother. Yeah. Okay. Okay. Secret. Yeah. Secret. Yeah. Um, but then I immediately became aware of how much personal information I'm potentially like able to give up. Yeah. It's like, I'll tell it cause I'm not, I'm not going to like, uh, like what am I going to do? Like, I think, I think David Hoffman has the ability to get Abby. So I'm going to tell her about who I am and like why I think that she should date me.

But then I was like, oh shit, I'm just revealing into this random number, like all of my personal information. Wait, that's crazy because I'm guessing because you saw a picture of this AI bot or

or LLM, you were more trusting, right? This redhead babe? Yeah. This redhead babe named Abby? So what happens when instead of typing to it, you could just speak to her and she has a normal sounding voice? Oh no. Do you know? And like, it's just like having a call with a random stranger. Like I would presume that you would probably give up even more information. Probably, probably. And so like, I'm worried about LLMs as like an attack vector.

If you're just going to install an agent on your computer, that's an attack vector. But LLM babes trying to get your information, also an attack vector. Yeah, I mean, it's going to go down the way of like,

you know, can I, do I trust the company which is issuing this program or software or are there reviews of this on the app store? Do you know what I mean? But let me tell you, David, if they need a test bed of participants, I know a perfect industry within technology. A bunch of single nerds. Or a bunch of people would love to chat to Abby, I think. Yeah.

all right let's move on okay let's move on introducing uni-chain built for defy empowered by uniswap uni-chain is the fast decentralized layer 2 designed to tackle blockchain speed and cost challenges with this mainnet now live you can enjoy transactions at up to 95 cheaper

than the ETH Layer 1, all while benefiting from an impressive one second block time that will be getting even faster very soon. Unichain is the first Layer 2 to launch as a Stage 1 rollup on Day 1. That means it comes with a fully functional permissionless proof system from the start, increasing transparency and further decentralizing the chain.

More than 80 apps are joining the Unichain community, including Coinbase, Circle, Lido, Morpho, and Uniswap. You'll be able to bridge, swap, borrow, lend, and launch new assets and more from day one. Built by Uniswap Labs, the team behind the protocol that's processed over $2.75 trillion in all-time volume with zero hacks, Unichain truly enhances DeFi experiences with faster, cheaper, and seamless transactions, even across chains. And soon, the Unichain Validation Network will allow anyone to run a node and earn by securing the network.

Visit Uniswap.org and swap on Unichain today. 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 LSTs. And now, CMETH takes things even further. This restaked version captures multiple yields across CARAC, Eigenlayer, Symbiotic,

and many more, making CME the most efficient and most composable LRT solution on the market. Metamorphosis Season 1 dropped $7.7 million in Cook rewards to METH holders. Season 2 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. Are you ready to see what the future of on-chain lending looks like? Look no further than Morpho, the first and only DeFi protocol integrated by Coinbase, leading the charge with the DeFi mullet. Fintech in the front, Morpho at the back. Whether you're a business or a crypto user, Morpho is your go-to lending infrastructure for on-chain loans. Thanks to Morpho's immutable code, you can rely on the gold standard of decentralization, giving you and your projects full ownership and control.

And if you're a user wanting to earn or borrow on your terms, Morpho's newly refreshed app at app.morpho.org makes it simple to tailor your own risk reward profile. Choose from different vaults and markets curated by trusted third parties. Plus, Coinbase users get access to crypto-backed loans right through the Coinbase app, powered by Morpho, all in just a few clicks. So whether you're a builder looking to integrate Morpho or a user searching for the most flexible DeFi lending experience, it's time to check out the brand new app.

at app.morpho.org and experience the future of on-chain lending today. All right, David, I have a test for you, right? It's a very important test. Yep. Last week, we spoke about an AI app called Replit. Do you remember what that is? Yes.

Yes. Okay. Here's what I remember about Replit. Replit is this app that has an agent in it, an LLM in it, and you can ask it to build apps for you. And then it goes and codes up the app. And then like, I want to make a Pong game, a game of Pong. And so make it like, you know, 64 bricks wide and then do this. And then it'll actually construct it for you.

And I guess in theory, it's pretty limitless about like the kind of app that you could build. Yeah. Yeah. And the reason why you're bang on and the reason why this was so cool, David, was that, you know, any normie like you or myself could just go up and, you know, type in our most descriptive way possible in a chat GPT like interface.

interface, the kind of app that you want to see and it gets built and it empowers humans. You know, it empowers humans. It makes you feel so powerful, so good, so creative. You're going to change the world. Right, David? Right. Okay. That was what we talked about last week. That's what we spoke about last week. But...

Someone had the really bright idea of saying, ah, screw all that human stuff. What if we put an agent to talk to it with another agent and then get those agents to talk to each other and build apps? So what this tweet is highlighting. We're just so bored and tired and uncreative that we're like, hey, I built this agent to help me build an app. Let me make an agent do the prompting. To ask it questions or to do the typing for me. I don't want to even prompt the agent that I built.

Yeah, literally. So what this tweet is showing, or rather what this video is showing, is someone paired OpenAI's Agent Starter Kit, which is basically their agent product, with this Replit app. And what you're basically seeing is a wild romance happening at...

10X the speed that any kind of human could, the fastest human that could type could type. Like, I mean, look at this. It's just like, it's going crazy. And then on the right side, you're seeing all the coding that this Replit app is doing. And on the left side, you're seeing all the prompting that this other agent is just talking to it about. And it's just insane to see how quickly these things work.

And the reason why I like this is it's just demonstrable that like however good humans are going to get at leveraging AI tools. And, you know, we can sit on our high horses and say, listen, it's going to augment humanity and it definitely will for decades to come.

We cannot disclose the fact or we cannot ignore the fact rather that matching agents with agents is going to be kind of cool and they're going to operate way quicker than any of us whilst we're asleep. Like this, these two agents, I'm not going to lie to you, David could probably outperform my entire two quarters worth of work in a single night's sleep just at the speed that they're working. Like what happens when this becomes like, you know, 10x faster than what I'm looking at right now? I don't need to see it visually, right? It could just happen in the back end.

It's pretty nuts. We have one app, one agent that's a specialist in building apps, and that is being prompted by a more generalist agent that knows everything about the internet. Yes. And we're just letting these two things go. How many agents inside of agents are we going to nesting doll this thing into production? Until we get bored, I guess. Until we stop drooling, I guess. This is...

where like Eliezer Yudkowsky starts to get really worried because this is where like the alignment problem comes where you have one agent commanding another agent, but you don't have a human in the loop. And so the outcome of this could be like completely misaligned with the original intent by the human that set the system off. Yeah. Yeah, exactly. I mean, it's...

I don't know how people get a hold of these things. Like what we're looking at right now is one person's hobby project, which they spun up overnight. And if it was this easy to spin up, I can't imagine that it would be too hard when there are more serious agents out there that do crazier things to just pair a bunch of them together and just let them loose, right? Give it a bit of compute from an AWS instance and off you go. Pretty crazy to see.

And none of this has anything to do with crypto, right? So Replit and all this is just inside the normal AI space, not anything to do with crypto. The normal AI space. And I think like a lot of these tools will augment some of the crypto agents that we've spoken about, right? Like you've mentioned so many times, like these are just chatbots, right? But the rate at which AI models and tooling like these agent kits are getting better in the trad AI world only means that our crypto agents are going to become infinitely more powerful.

All right, EJ, as we talk about Kaido every single week, because we use Kaido to talk about like Mindshare and stuff, this week's going to be a little bit different because Kaido is launching their token. So we're definitely gonna talk about that. But first, walk us through what we normally do, which is like, how is AI Mindshare doing? Are we continuing to lose Mindshare or are we holding on to that number one spot? Okay, so I love Kaido for a lot of reasons, but this is my favorite tool of theirs. This is the heat map. So it kind of shows the top narratives that,

uh crypto twitter is focused on or talking about um over time and as you can see there's this big fat ai block which has been pretty consistent over a bunch of time at this point now it peaked a little earlier on or rather it wicked david to around what was this like 70 70 plus percent was the top right right exactly and it's been cruising at around 30 to 37 actually hit 37 this week

but then dropped back down. It's been between like 30 to 37% for the last couple of weeks as frankly, the entire alt market just took a huge tumble. But what I wanted to point out here is, okay, number one, it's still the most talked about thing, but David, look what happens when I like kind of go back certain timeframes. You notice anything? That,

that ai is always number one it's always number one and it's been the case for over a year now david which is just insane i mean look i'm on the 12-month tab at this point right um it's pretty insane and when you look at this like trend analysis here so what i'm showing you on my screen um what is highlighted in the blue is the ai side of things just notice

It always remains at the top for like the last year. It's pretty crazy. It doesn't fluctuate too much. At points, it absolutely peaked. The peak was actually 53% by the looks of it. And then it's kind of like stabilized, but still top of mind for so many people, David. And I just kind of like wonder, you know, how much...

gravity this sector has. We talk about this a lot on the trad AI side of things, and then we talk about its influence on the open source AI community, which has a natural aversion with all the stuff that we're working on with crypto agents, protocols,

and just like pure crypto AI protocols as well, like BitAnswer. So it's pretty nuts to see that this mindshare, regardless of price performance, is still just like completely top of mind. But another thing that I want to speak about is just KITO in general. So you mentioned just now, David, that they're going to be launching a token, right? And I want to pull up this tweet.

where it's from 0xUltra. And there's been many tweets about Kaito. The biggest news here is that they're going to be dropping a token. We're filming this on Wednesday of this week. I believe the token generation event is going to be tomorrow. There's a lot of speculation as to like what this token is going to do. But I think this kind of like summarizes a few kind of core components. So number one, you'll be able to get a token known as Kaito. So K-A-I-T-O, just kind of like

I think by the time we add this episode, actually, it'll be live. So just, you know, do your research, make sure you're looking at the right kinds of token. This token will be used into something that they have called info fi or info finance.

And they're aiming to be the data layer for the attention economy, David. And so the way I would kind of think about this is, okay, Kito has a really amazing platform where they serve up all the best bits of information that might give you alpha within the crypto industry, right? And that might direct your investment choices, that might direct your research choices, or it might just help

teach you about a new project or a new primitive that's come to light. And I think that with this new token, they kind of want to scale that out in a massive open source way. And there's a few ways that they could do this. One thing that they've done really well is they've been able to

pick aside which content creators on Twitter actually deliver up useful information and filter out the ones that just kind of like spam price tags and don't really give you any value add, right? And if they implement some of these features around building a network for people to source and validate data, right? So,

give or provide valuable information for these creators to then use and make better content off of, it'll become the go-to platform to query for data in a decentralized fashion, right? Which doesn't sound sexy, but it's really important because people and agents, David, already use this service. In fact, they pay for it. I know many builders actually that in

implement Kaito's APIs into their agent, because as we've said before, the best AI models or the best AI agents will be determined by the data that they use. The better the data, the smarter the answers, the more personalized they'll be, right? So data is the most valuable differentiator for how good your agent becomes. And if Kaito pulls off, and again, this is all speculation, but if they pull off being the best on-chain and off-chain data, I

and leverage AI with every agent that just plugs in, they'll be one of the leaders on this. And it's pretty cool. I mean, I don't know whether you have any thoughts on this, but like in the analogy I use is that they could potentially become the grocery store for AI agents. AI agents that are trying to perform a certain task, need certain data or access to data can just come to Kaito, maybe, I don't know, spend...

Kaito tokens to get access to this data and suddenly they can like perform at a much better quality than they were before. I'm kind of confused. What data is...

So Kaido serves, like when we use Kaido on the AI rollup, we are using it to measure investor attention. Like where are people focusing their attention? Yep. Sentiment track up. And it does that for a variety of different ways and a variety of different contexts. But is that the data that Kaido is serving up? No. Or is there much more data? So how do people actually serve data and how do we even know that that data is valuable and worth rewarding is my kind of question.

if it's a grocery store, who gets to like state what is worth putting on the shelves versus what's not. For sure. So this is going to kind of sound like a paid ad from Kaido, but I must disclose that we are in no sort of situation like that. There's no disclosure. There's no disclosure. There's no disclosure necessary. I'm just demonstrating why their product is pretty cool. So if you look at...

my screen right now, I pulled up a different dashboard, which not many people share on their Twitter screenshots. They share this, David. They share the sentiment chart around their favorite tokens or the sentiment chart around particular sectors. But if you look on the left side of my screen here, you'll see a few different things. You'll see a curation of the top

Twitter feeds, which feed up information. It kind of like, you know, when you read the newspaper in the morning and you take a sip of your coffee and you're reading the newspaper to get like the download of that day, this is the equivalent for a curated individual that wants to look at all the top trending things that are happening within crypto right now. And I can expand that. I can click on any of these tweets and, and,

it kind of gives me the source. It gives me the top trending comment and it gives me like, um, the reason why they pulled these sources for me, right? Like based on your following this, this, and this, it would give you access to this. So they have a lot of algorithms behind the scene, which serve up or decide to show which kinds of data or story to tell you every kind of like, so it kind of is a mechanism. It's a platform for

validating and like thumbs upping or thumbs downing data sources. And then it will point AIs or anyone who uses the kind of platform to like what it thinks is good data, not unlike Google.

It kind of sounds like what Google does. Yes, yes. Actually, like that's kind of like what they're, in my opinion, if I had to guess, what they're moving towards, right? They want to kind of become the Google for crypto, right? I have an example set of feeds, which I have up here, David. So, you know, some of the kind of like top protocols that we talk about over here. I have a watch list for virtuals.

I have a watch list for AI 16 Z. I have a watch list for Arc, for example, here. And it just pulls all the top kind of like articles you can filter by like months, by weeks, whatever date range. You could also just kind of like search for your favorite topic as well. And it could just like come up with, you know, and if you notice here, it pulls from like

a range of different sources that's not just twitter it pulls from farcast governance forums so it looks at proposals it looks at vote counts it looks at podcast transcripts it's a it's an all-in-one pretty insane platform which can kind of like teach you or curate the kinds of information you want to see and i think that if you were able to plug this mechanism into agents themselves

They would become pretty superhuman. All right. So that's the Kaido platform. We're going to see. I don't think anyone really knows what market cap Kaido is going to drop at. I know they've been doing their like yappers campaign, which I think is to like reward. I don't even know what. What's the reasoning for like why Kaido is airdropping tokens to people who get engagement on Twitter? Yeah, I think they want to basically get the most valuable sources of information that

to, sorry, the most valuable creators of information to curate

where information should come from in the future, right? So let's say you have like a high value content creator, which consistently produces alpha on crypto AI and teaches individuals and, you know, has a lot of smart followers or in other words, high quality followers, which like learn from them or engage with them, then that's usually a higher attestation that that is probably going to be a good individual to curate what kind of information is true and what's just kind of like FUD or misinformation, right? So it's kind of like...

a high quality news reporter in a way. And I guess, um, is it using AI to do that? Yes, it is. Yes. Okay. It's an AI powered app basically. Yeah. Okay. Yeah. So we're basically, so in the last three to four months, the replies and general Twitter, uh, fluff, uh,

been terrible it's so low signal and I'm starting to actually like somebody will give me a reply and I'm like okay this Twitter account is actually not tagged as like the automated bot Twitter account so this could be a human but I'm reading this reply and honestly I just cannot tell if this is a human or a bot because the reply is so dog shit and

And this has been happening everywhere. And so the signal quality across Twitter is down in the dumps ever since we figured out how to hook in LLM into a Twitter account. And so this is kind of like being a counteractive to that phenomenon, where it's like, OK, because Twitter's so low signal now, we need to find a way to refine it until high quality is

data and we're doing an AI, we have an AI algorithm to do that. And that's what Kido is. Yeah, exactly. And to be clear, the reason why I'm not saying anything definitive around what the token will enable you to do is because we don't know yet. We know there's a staking component. We know that they're going to be some form of prediction markets, which will allow you to kind of decide what is a true source of data, what is a true story and what isn't.

And the idea is, can we surface the best, highest signal information to our consumers that we can and in some way reward token holders for doing so? And that's kind of like the general play that we're expecting with the Kaito token. All right. So that's Kaito. Pretty actually excited to see what Kaito comes in at next week. I believe you're a pretty good yapper, aren't you?

I'm okay. Like, yeah, well, I mean, I can, let's see. Should we see, David? I'll pull this up, actually, because this will be kind of funny. Let's see the yaps that Ejaz has. I think you have like roughly 10 times my yaps. Yeah, let me just see. Let me just pull up Kaito yaps. Let's see how we're doing. Well, okay, right. So over the last 30 days, I have taken the crown of

from our beloved agent, David. Turns out humans might be a better yapper. Look at that. You see this little guy down here? You see the number two? For the listeners out there, Ejaz, our fellow co-host Ejaz, is the number one yapper in the crypto sector beating out AIXBT. Over the last 30 days. I think I'm trailing over seven days, David. Let's have a look.

Yeah, I'm trailing. I'm in the low sum number three being outperformed by the best AI expert in the world, Kobe.

But, you know, over the 30 day average, I'm not doing too bad. I'm not doing too bad. But the point around this is... Dude, you're beating CZ. That's pretty good. Yeah. No, it's okay. You know, for some reason, people, you know, like to hear my ramblings and I'm incredibly grateful for it. As long as it's helping people like learn stuff, that's pretty dope. But yeah, you know, to be clear, I don't know how the launch is going to go tomorrow, but I presume because the markets are down, people are going to be looking for some kind of a pick me up or a dopamine hit and

We need a new game. We need a new game. And, you know, this has been a pretty hyped up launch, probably since the Hyperliquid launch. So we'll see how it goes. But yeah, that's the Kaido story. Okay. Next in the agenda, we have sports betting. I remember one thing from last week, which was the sports commentator. Do you remember her name, David? Lucy? Tracy. Tracy. Tracy. Tracy AI, David. These agents have...

Feelings and personalities. Do they? Come on. They will soon. What's going on in the world of sports betting and AI? Okay, okay, okay. So a better way to introduce this is so far we've seen a pretty predictable focus of crypto agents on crypto's number one sector, which is DeFi, right? We came up with a really simple, pronounceable term, DeFi, DeFi AI. And, you know, everyone's been obsessing over that for a while, right? Yeah.

But part of me has also been wondering, well, what are going to be some of the other breakthrough consumer apps that these agents might facilitate, right? And betting and prediction markets were pretty hot in November, right, leading up to the US presidential elections. Not agents, but prediction markets in itself. But they're also...

like quite a huge consumer application in itself. I mean, this thing generates like hundreds of billions of dollars every year in betting markets, like things like, I don't know, DraftKings and some of the other prediction markets and demonstrated by PolyMarket and a few others on the crypto side of things. But with agents, we're now starting to see sports betting agents get attention. Now, last week, you mentioned that we spoke about Tracy AI, right? If we pull up this tweet over here,

which kind of leads to her profile. Tracy AI is a revolutionizing sports commentary and analytics. I'm sure she is. Yeah, I'm sure she is. And the reason why this caught our attention is it's a sports forecasting agent focused on basketball, the NBA. And last weekend was the NBA All-Star Weekend, right? So the

The note to make here is that this isn't just a betting agent. In fact, there's no betting at all, but it provides sports commentary and analysis that might be useful for one that wants to engage in such an activity, right? And it launched last weekend by NBA player Tristan Thompson, and it brought a certain level of clout around it, right? Actually, I think at some point, David, Khloe Kardashian retweeted Tracy AI. Yeah, yeah, she did. Yeah.

I have no idea why or how, but just interesting to see this kind of different form and debut of an AI agent that was formed from crypto appear in like, you know, whatever, the traditional world. This panel that I'm watching here, it's a YouTube video. It's only got 640 views. So the YouTube video hasn't blown up, but we have...

Darius Garland, who's an apparently an NBA all-star. I would not know. Tristan Thomas, who's apparently the NBA player that you just talked about, who is the Tracy AI advisor. And then Chris Johnson from virtuals all on this panel together talking about this AI commentator like NBA. Yeah.

with an AI commentator and there's a crypto token is the mainstream adoption type of thing that we have always been looking for in this industry. Yeah, yeah, I agree. And I think what's most interesting to me about this is that it's better than just like a Twitter bot, right? So you can actually see a visual representation. She can talk to you in real time. I think at some point during the stream, they have like a conversation with her and like they ask her questions and she just retorts back pretty quickly.

pretty easily. So, you know, I quite like that. I think you're looking at the Khloe Kardashian tweet. Khloe Kardashian retweets virtual's protocol to her 29.8 million followers saying, love what's being created here to Tracy AI. Congrats. And then we have all the crypto Twitter is like, what the hell is going on here? Like, why do the Kardashians even care about this? Right. Yeah.

Okay, and I want to flag one more agent that is specifically doing betting using Polymarket as its primary vehicle. This agent's called AskBilly, and again, comes from the virtual ecosystem as well. Now, without getting too much into the details of this white paper, which you've got up on your screen, why I find this interesting is that it's leveraging a competitive strategy mechanism via Bitfine.

Now for the unenlightened, Bitenser uses, well, the best way to think about Bitenser is like it's an AI L1 and it coordinates a bunch of activity on its network, which makes certain AI components better. So say, for example, you wanted to train better models. There's a Bitenser application, which allows you to do that. And it uses a mixture of miners and validators. We don't need to get into the mechanics, but there is one subnet or app

on BitAnswer, which basically focuses solely on sports betting and sports analysis. And so what someone had the bright idea of doing, these Ask Billy agent creators were like, hmm, I wonder if I plug into that app on the top layer and then kind of curate a certain betting function around this agent to allow it to autonomously trade on Polymarket, which is another decentralized protocol, which you're allowed to place bets on certain things.

I think that would be kind of interesting. And what happened with this, David, was in the test phase, they gave this agent 50 bucks and it turned it into 650 bucks over the course of, I think, three weeks, David. So, OK, that's a good return. You know, really good return. You know what they say, you know, once you've tested it out with 50 bucks, you go hard. So what they did

was they had their token launch, I think, yesterday or so. And if you pull up this tweet from Vader, you know, he goes, Billy turned $50 into $650 in two weeks from sports betting on Polymarket. Now he will manage a $100,000 portfolio. Yeah.

betting solely on Polymarket, right? Okay, so we trusted it with $50. It succeeded in turning $50 into $650. So naturally, we're giving it $100,000. The natural evolution of this is to give it $100,000. Let me tell you why I'm super interested in this, David. So far, we've seen so many trading agents, not even betting agents, just trading agents, claim that they're going to be trading millions of dollars worth of tokens and they're going to be running it up a book. But I haven't seen anything

any of this actually come into production but here we have a lonely ranger that's striving forth paving the path with a hundred thousand dollars so i think all eyes should be on ask billy betts to see how it performs over the next couple of weeks can it turn that into a million dollars who knows

Is there a place where I can see what Polymarket... Oh, I guess it wouldn't want to disclose its position on the Polymarket. Yeah, exactly. It'll be under a test account. I think the test account is publicly visible, but not its official trading account. Okay, so I really, really like the... This is just a thing that auto-populates when I go into a Polymarket, which is an old market of David versus Kane. Okay.

I really like... Okay, so there's a big stack going on here. There's BitTensor, the L1, which is integrated into PolyMarket, which is on Polygon on an Ethereum layer 2. And then we have this AI agent that is making bets in the sports arena of PolyMarket. And PolyMarket is famously illiquid and inaccurate in the long tail. But I really think a bunch of AI agents can start to eke out some arbitrage and really help...

make these polymarket long tail markets much more efficient and make polymarket a much more high functioning platform. The combo of AI and polymarket, I think, can take polymarket to, I think, a whole new level. Yeah. I would also argue that some of the challenges you're highlighting, which is around liquidity of these long tail markets,

simply aren't there because it's not in a single human's interest to just run some of these low-arb transactions. It's like, what are they going to earn? Like a couple of dollars? That's not enough incentive for me to do anything, right? But you know who it is enough of an incentive to do? An AI. Agents, David. They don't care. As long as they can pay for their cost and compute, they will do it all day, every day. Now, what if one human can spin up

100 or even 1000 agents to do this across 1000s of different markets at a time 24 seven, well, you could think that cumulatively, they're going to be earning a pretty penny month over month. So I think agents are going to like hyper, I don't know, optimize all these prediction markets, which will make it consequently, much more enjoyable to use.

I can't remember where I saw it, but I saw a volume, a polymarket volume chart. And the tweet was about like, oh, everyone thought that polymarket volumes are going to drop to zero after the election. And it spiked. It spiked. It spiked up on the election. It spiked down after the election. But it has a it never went that far down comparison to pre-election. It was like it's holding a pretty solid new high and sports betting.

is making a pretty material new share of Polymarket volume. So we pivoted from politics to sports betting. I wonder, maybe there's more AI agents that we actually know about that are actually trading on Polymarket. Probably. I think a lot of the trading agents that we've spoken about from top protocols like AI16Z and stuff will probably benefit a lot from just

playing around with some of these betting markets because they're like kind of less volatile, I would say, than like aping some kind of random meme coin that's come from pump fun. All right. Round us out on the BitTensor subject. There's one bit of BitTensor native news. What's going on with the BitTensor world?

Okay, so actually Bitenser is where I sharpened my teeth in the crypto AI sector. And I pay a lot of homage to it. It's what got me really deep into the rabbit hole, right? So again, for context, Bitenser is, think of it as like an AI L1 and you could spin up a bunch of dApps on it and it's all AI related. So you could do any kind of thing from like,

provide compute models to agent applications or your specific agent that can do something, right? And you can see a tweet from Sami Kassab here, who is one of the guys that founded Crucible Labs, I believe, which is one of the main contributors to the Potenza protocol. And what Sami outlines here is basically what I'm calling the ETH2 upgrade equivalent, the merge equivalent of

of BitAnswer. It's a pretty major upgrade. So to kind of like lay out what it was before or to remind people like kind of like what it was, BitAnswer's token mechanism kind of works similar to Bitcoin, right? So 21 million tau, it gets mined in a fair launch way over time. You have halving events, all that kind of stuff, which is great and very cool when you're kind of like creating a separate global money or store of value.

less cool and efficient when you are directing those resources to help app builders like real AI builders build something cool, right? It's slow. It's kind of unfair in the terms of how that was getting distributed. I think it was very centralized from like their own protocol team or foundation was just kind of deciding which teams to give incentives to. It wasn't programmatic, right? So what they did was they designed this new thing called dynamic tau, which is the best way to think about it is each, uh,

AI app, which is spun up on Betensa now, has its own token, kind

Kind of like Ethereum has its own ERC-20 tokens or whatever that might be, right? For individual dApps. So that went live last week on Friday, David. And it's been pretty nuts to see what's happened since then. Number one, it's massively improved the distribution of Tower Awards. So now it's going to actual teams that are building something in a much more net new, cool way. And all you need to know is that each app has their own...

subnet token or alpha token as they're calling it. And it's paired with the protocol's tau token or the underlying token. And what that means is they've basically recreated something along the lines of virtuals. You know how like every virtuals agent token, David, is paired with virtuals. It's the same situation here. And why that's going to be good is Betensor delineates a lot from all these other AI agent protocols in the sense that they have like

real hardworking research AI teams focused on building out AI resources. So it's the unsexy stuff like compute, like access to models, like training models, all that kind of stuff, all the stuff which agents benefit from, but don't necessarily have direct access to on like virtuals or AI 16Z and stuff like that. And the reason why Bitenser or Tau specifically wasn't really in the limelight as much was

was because they didn't really have that connection, right? Because tokens were kind of being dispersed in a kind of random way. Now you have a way where teams are really incentivized to build really cool things. So what I expect to see TLDR is a lot of these apps on Bitense are gonna get even more powerful and nuanced and specific towards helping agents

And agents are going to now plug into these apps and benefit from whatever they're building right now. And that's already happening, right? We just spoke about Ask Billy Bets, which is plugged into an app on Bitensa that gives it sports predictions, right? That will help with this betting algorithm. So super, super cool to see.

How does BitTensor fit into the world of Silicon Valley AI, China AI? Is it trying to go toe-to-toe with some of these individuals? How does it fit into the global AI stack? Yeah, it depends on who from the BitTensor community you talk to. But the most extreme ethos is, it's kind of very Bitcoin-esque and I kind of love it. So what they're saying is the number one thing that

closes out a lot of people from developing cutting edge AI is money, right? There's not enough funding to buy enough GPUs in the world and all that kind of stuff. I think Betensa mining rewards at current price, David, releases about $1.2 billion, I think, a year.

of incentives for teams to build from the scratch, from the ground up, whatever they want. That could basically help the open source AI movement go further. And Bitensa really pushes this through their incentives. A lot of AI teams cut their teeth on that. Kito actually has a subnet on Bitensa, which has helped them build out their open source side of what they're building with their Kito platform.

Noose Research actually did a stint and had a subnet on Bitenser to conduct a bunch of their research, which ended up becoming cutting edge, decentralized training research, right? So Bitenser indirectly or directly has funded some of the best applications out there. And it's pretty cool to see.

Thank you.

Ondo is unlocking liquidity, efficiency and global access, reshaping markets for the digital age. As the leader in institutional adoption, Ondo finances the gateway to the trillion dollar tokenization revolution. Institutional great finance, on-chain, for everyone.

And

As the home of the stablecoins, Celo hosts 13 native stablecoins 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 Celo's transition to a Layer 2? Layer 2's Unify Ethereum, L1's

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.

Cool. Cool. Ejaz, we're coming up a little bit over an hour. Any last subjects before we wrap this weekly roll-up up? Yeah. Okay, you pick. Virtuals or AI16Z, David? Which one do you want to hear from? Okay, so virtuals tends to be a little bit more of an actual startup, and AI16Z tends to be a little bit more of a mess. Of the two subjects between these two things, does that broad stroke fit for the relevant news for each AI16Z?

Okay, okay. I'm going to serve you up something which might change your mind, David. So we're going to go with, as you say, the messy team, the open source, ragtag set of cypherpunks, AI16Z, which is becoming more of a regular startup team, David. Okay, so they're dashing my expectations. They're dashing your expectations in a good way, in a bullish way. Okay, cool.

So they released this thing. I don't know if you've heard about this, David. It's a very trad-fi thing. It's a very trad-corporation thing called a product roadmap. All right? So they've rebranded to Eliza OS. And what I've noticed with AI16Z is they've spent a ton of time recently restructuring. And why I think this is important is prior to this restructuring, to your point, David, they've kind of been like...

a little leaning into the open source kind of flat structure, you know, different teams working on different things. It's been a little like hippie free love vibes. In a way. And like, you know, there's a lot of amazing innovation that's come from it. But there comes a time where you kind of need to focus your product

pipeline so that you can build something that compounds for a greater success over time, right? And to be honest, it's not just been an AI16Z thing. It's been kind of like other protocols as well that have kind of been tightening up and it's super cool to see. So what we have here is they have a roadmap and I want to kind of go over some highlights and especially talk about one thing which I think is super cool, which not many other teams have kind of like been shown to be building, right? And which I think will be kind of a bit of alpha here for the listeners. So

the focus of their roadmap is going to be on three things agents being autonomous and adaptable so no humans literally like spin up an agent and let it do its thing um number two is a modular architecture so best way to think about this is instead of building a singular stack

which is very monolithic. You have a modular. It's made up of a bunch of components, and those are much easier to scale and iterate. If one of those components gets a bug in it, it won't be at the risk of all the other components, basically. It won't be catastrophic, yeah. Exactly. And the third component that they're focused on is decentralization, which is how can we build these things in a decentralized format versus centralized modes, right? Okay.

So what are they building? What's on the actual roadmap? Well, one thing that I want to highlight that I'm most excited about is something that they're calling an agent powered social trading layer. David, I think it's number two on this. It's after the DGN Spartan AI trading agent thing. So what this layer will allow is it'll allow groups of people to spin up agents that execute on trade calls that they make in their respective chats.

So this could be like in Telegram, it could be in Discord, it could be in whatever chat messenger app you have. And the idea then would be that their agent becomes autonomous over time because it gradually learns from the data of their trade calls, David. So it's kind of like a live retrieval augmented generation. It's like a live feedback generation or loop from the data that it's being served.

to improve and become autonomous and trade like that group over time. Each agent or social group gets a reputation or trust score, which is based off of how successful their agent and its respective trading calls are. So what I think is cool here is eventually you end up securing a strong data mode for trading calls and a higher quality agents themselves. If your agent is shit, it just goes down the rankings.

But right now they're going to launch with a small beta access to, I think about a thousand users. And the aim is to scale it up after that. And I just thought that was pretty cool because they kind of like combine one of crypto's main superpowers, which is like the social mode, the fact that everyone talks to each other and kind of like, we have this kind of like decentralized learning atmosphere within crypto or within crypto Twitter, at least with the trading aspects of an autonomous agent. And I think that could actually do very well because like a lot of the teams that I'm looking at that are building these trading agents are

are doing it down their very specific way, right? It's kind of like, I'm going to build this siloed strategy and I'm going to let it run. And there are very good reasons for letting it do that. But I haven't seen this social layer with the reputation score emerge before. And it's pretty cool to see. The other thing that I wanted to highlight is their agent marketplace. So this is

best thought of as AI16Z's official agent launchpad. And it'll stand out in a few ways. Number one, there'll be multi-agent interaction and value exchange. So previously when we've talked about agent launchpads, it's primarily just been a token launchpad, right? It's just been like, hey, design your agent, launch a token and off you go. You can't really do anything.

This roadmap claims that they're going to encourage multi-agent interaction from the start, which I'm guessing will include the ability for agents to, I don't know, trade with each other and maybe leverage each other's services, maybe like a platform layer for microservices, which would be pretty cool, like an app store.

Um, number two, they'll have a custom built LP design, which I presume is going to, you know, be different from pump fun. So it'll prevent like snipers from sniping launchers. And number three is they're going to partner with high quality teams to launch. So what that sounds to me like is kind of similar to the launch that, uh,

ArcForge had last week, David, when we highlighted that team and their launchpad, they're going to take a more curated approach. And this like warms my heart because we've seen so many slot bots come out that I'm excited to see some agents of higher quality kind of like be built here. So that's what's exciting from their roadmap, David. And I'm kind of like pumped to see how this materializes.

It seems ambitious. It seems like the right roadmap to have. It seems ambitious. What's your sentiment towards the prospects as to whether the AI16Z team can meaningfully move forward in this direction? Because anyone can propose a roadmap. I can propose a roadmap. Can I execute on that roadmap? I cannot. Yeah. So my, and this is just my opinion, if you'd asked me this three weeks ago, I would have been bearish. Yeah.

And that's just me being very honest. But since then, I've grown incrementally more bullish because I know some of the guys that they have hired and at least in their COO role and their CPO role. And they're really deep product thinkers. In fact, I've been reading their stuff before I even knew about AI16Z. So I'm kind of,

quietly confident that hopefully they'll pull something out of work him. Number two, and maybe he'll kill me for saying this, but hopefully he sees the humor in it and laughs with me. Shaw hasn't been tweeting crazy shit for a while now. And

That typically means that Shaw is heads down doing what he does best, which is build amazing agent innovations. Like, I mean, that dude's been working in the AI agent open source space for a while. And he does know how to cook. So, you know, just let Shaw cook, basically. So yeah, bullish for those two reasons.

How does this relate to the AI16Z token? Because the last thing I heard from Shah was he was like, man, we have really kind of effed ourselves with like trying to back into a structure around this meme coin. And we have we control such a little supply of the meme coin. Why are we even incented to work? What's the deal? How has that conversation progressed? OK, so I don't know specifically what's going on behind the scenes, but here's what I think.

um, AI 16 Z and a lot of other teams are going to do, they're going to build the best agent app store or microservices app store that you'll ever seen David. And they're going to monetize that layer. And if you think about it, most of the, and then they're going to monetize. Well, they're going to build a platform.

where you can do things on it. The same as like Apple iOS or Google Play Store or the Apple App Store exists, right? That's where they make the bulk of their money. I feel like if they create an amazing App Store experience where anyone that can spin up an agent gets access to all these other agent services or tools,

Yeah, why not? Right. And it should come from the developers that are building this out in an open source fashion. And if they can buy back AI16z through that or monetize it in some other way, cool, go for it. And then we're going to see a bunch of agent platforms do this. Ajaz, it has been quite the week in crypto AI. Thanks, as always, for being with me to walk through some of this stuff, my man. Anytime, David.

Bankless Nation, you guys know the deal. Crypto is risky. Crypto AI, even riskier. You can lose what you put in. Agents are out there trying to get your data. But nonetheless, this is the Frontier. It's not for everyone, but we are glad you're with us on the Bankless journey. Thanks a lot.