AI ROLLUP #9: $6M DeepSeek Shocker | $500B AI Push | Venice’s Billion-Dollar Airdrop | Solana DEX Record
01:10:37 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 Ejaz. Ejaz, man, big week in AI this week. Big week. Big week. How are you doing, man? Good to see you. China, David, China. China's house going. Dude, the CCP is coming for our AI models and I don't know what to do about it. But no, okay, so let's set the stage for a second here, right? Mm-hmm.

because a completely unheard of AI model has taken away the limelight, David. Forget about OpenAI, forget about Anthropix Clawed. It's all about this model called DeepSeek, which is an open source AI model developed by a team in China. And it may surprise you to hear that it costs next to nothing compared to OpenAI, and it beats OpenAI's top model by...

a decent chunk or at least matches its standard. And the thing that's blowing people's mind is, aside from the fact that it's from China, USA's number one rival, is the fact that it costs next to nothing. And they pioneered two groundbreaking research projects

techniques, which seem to catch everyone else who has spent billions and billions of dollars into this off guard. Did you catch any of this, David? Yeah, so I've actually been going down this rabbit hole pretty deeply over the last few days. We did an episode with this one individual who made this article that apparently went around both Wall Street investors, and then he was watching in Google Analytics where this article was getting read, and it made its way to Silicon Valley, where NVIDIA's headquarters are.

And so people think that the market got tanked because of DeepSeek. He thinks, and many other people think it's actually because of his article that he wrote, which is not just DeepSeek, but many other companies unbundling Nvidia from the periphery. And then DeepSeek was kind of the thing that lit the match, if you will.

I'll distill everything that I know about DeepSeek and I can get you to check it. DeepSeek, the model, is very legit. It does...

in a very legit way pass benchmarks and performance tests that is very competitive with OpenAI. The cheapness of how it was trained, the $6 million in training costs, people are saying that that is much less legit. That's where the CCCP, China, and DeepSeek has the most incentive to lie. And then OpenAI is saying, well, there's extremely strong evidence that they actually just used ChatGPT to train people

So nonetheless, there were some very real engineering breakthroughs that changed how LLMs work in order to produce an output that's much more efficient, that uses much less resource costs to produce a very strong output that is just cheaper to run. And so the fact that

DeepSeek, the model, is charging 95% less than OpenAI for their API calls. It's just evidence that this is very, very real. But they also just rode on the backs of ChatGPT in order to produce their open source. Nonetheless...

This thing is going to make waves in American AI models. And now everyone at Facebook, at Meta, and OpenAI are now understanding the tricks and bells and whistles that was built into DeepSeek to incorporate that into their models. And I think really people are realizing that the arms race between the United States and China is fully on.

And any innovation that we make in America is just going to be copied and improved upon by China. But that's a positive feedback loop between development in the West versus the East. That's kind of how I understand it. Yeah, really good summary, David. I would just add one more thing, which I don't know if you mentioned it, but just in case you didn't.

they open sourced the whole thing. They open sourced it. They open sourced it. Which is just a crazy move. That's like, you know, um, creating a, a potion that allows you to live forever and then just telling everyone what the recipe is instead of monetizing it, you know, which is the American way. Um,

but yeah. Did you see Archival's tweet? I didn't. Where he says, it's ironic that we got AI that costs $200 a month from a nonprofit and then we got the open source AI from a Chinese hedge fund. Yeah. Yeah. That was,

That was hilarious. We should pull that up. Okay, so I want to touch on a few things that you mentioned, David, right? So you mentioned that it's legit and it made these groundbreaking achievements. What are these groundbreaking achievements? Well, I want to kind of touch upon a few things here and kind of try and distill it for the audience, right? So the one pioneering thing that they figured out was...

how to use data that they train their model with more effectively. So typically, pre-DeepSeek, what everyone thought was the case was you had this model, so you kind of designed it to do like super cool things, but then you need to train the model. So you need to combine data with it. You know, all this like raw data, like, hey, this is David Hoffman, by the way, and these are humans and the sky is blue, all this kind of stuff, right? You need to run it through the model.

In order to do that, you need to combine it with a bunch of compute power, right? That's why Nvidia rose to fame because it like produces the best chips, which allows you to run a bunch of compute, which allows you to train these super smart models. And then suddenly you end up with magic like ChatGPT. And so what that looked like was data,

and a ton of compute. And it was so expensive to do, David. It cost billions and billions of dollars to train these things, right? But then what these guys, the DeepSeek team figured out was like, okay, well, we don't have a lot of compute. That's pretty expensive. And, you know, there's a conspiracy theory saying that they actually did, didn't, it hasn't been proven yet. But,

The point is- We do know that China has less resources, compute resources available. They are constrained, correct. They are compute constrained because of the chip ban. Exactly. So even though they are getting able to get around the ban in finicky ways, nonetheless, that is a constraint that they have. Yes, exactly. So they're constrained, right? And so they have to figure out a smarter way to overcome this setup where they need a lot of compute to train a smart model. So what did they do? Well, they designed the model and then they thought,

hmm, I wonder if a smart way to train it is to feed it a bit of compute, let it come out with an output from the little bit of data that it's used,

and then get it to review the output and figure out why it's wrong or why it's correct or how it might get closer to the truth. And then, you know, rerun that compute, but just a little more smarter. So think about it, right? It's like learning to ride a bike for the first time. Imagine hypothetically, you jumped on a bike, you know, you sat on the seat, but then you put your feet on the handlebars and you...

fell off immediately because you lost your balance and obviously putting your feet on the handlebar doesn't make the bike move forward. It's like getting off and being like, hmm. Let's not do that again. Yeah, let's not do that again. I wonder if I put my feet on these little peddly things beneath me. Maybe it'll center my gravity and maybe

Maybe I'll figure something out. Like maybe if I push on it, it'll make the bike go forward. So they pioneered this method, David, which basically meant that the model had to take fewer steps, less compute, and less data to figure out what it needed to do. It's kind of like a human. You know, when you make a mistake, you don't just run straight back. Well, some humans do. They just run straight back at it. But it forces you to think and be like, hmm, okay, interesting. The way that I heard this is that...

LLMs and we all I think we all saw this in early models of chat GPT is like hallucinations where you would throw it a prompt and then it would spit you back out a very long answer and the end of that answer wouldn't really like line up with the first part of the answer because LLMs are very bad at backtracking and

correcting previous responses uh the metaphor that i heard is like it's kind of like when a child is just like having train of thought they're like they're three years old four years old and their logic isn't so great and they can't stop themselves and consider what they've said earlier and how does it square with what they're saying now so they kind of just sound schizophrenic and they go off into the wild yeah and so what this has done is it's broken up into chunks

It's broken up its process into chunks reviewed each chunk and make sure the chunks fit before creating a new answer and so it'll before actually spitting out an answer it is able to make sure that the chunks fit with each other and Coherently and then it produces an answer but this this actually my understanding was that this takes more compute This is more inference time because each chunk has

Chunk has inference, costs inference, and then summating all of them is also more inference. And so this boosted the valuation of Nvidia because they were like, "Wow, the value of inference is just up only." So much inference makes a better answer. Right. So there's some nuance there with your chunk theory, David. So you're right that you need more inference, and for those listening who don't know what

what inference is, think of it as like making a call to the model. Like, you know, a website will make a call to the model to be like, hey, by the way, this user asked this question. Can you give us the answer for it real quick? Okay, cool. Thank you. Right. When you are asked, when you are querying ChatGPT, you are making it do inference. That is the inference side. When you type in a sentence like, hey, I want to make my mama's banana bread recipe, it's making a call to the GPT model. Anyway, inference is cheaper than training.

full stop right so um the point of training is cheap inference correct exactly so whilst it's making a lot more calls inference calls um it's getting smarter by using less compute effectively but the point you were making just now david is people suddenly woke up and was like wait hang on a second

if we apply this inference model to like a bunch of other things, right? Like, hey, like what if I could just like keep prompting ChatGPT in my chat to give me a better recipe or something? Like maybe I'll end up with the ultimate recipe, right? And people are now like thinking of how they apply that model, you know, to a number of different things. But I just want to round this up, David. The second breakthrough that they made, which is super important is, you know, typically how you have like the GPT model, right? Let's take OpenAI, right? They have the model.

When you query it or when you inference it, it queries the entire model. So think of this model as the design. It hits the entire brain, David. Every single neuron is stimulated. It's queried. Exactly. Now, if we take the DeepSeq model, it gets a request and it only routes it to the section of the model brain that it needs to hit.

Which is just so much more efficient. What do they call this? MOE? There's something of experts. Mixture of experts. Mixture of experts, right. So think of this new model as literally a mixture of experts. And when you query it, it identifies, all right, I need to, I'm going to add...

ask these three experts of my 27 experts and 24 of them are going to shut down and not run inference and three of them are going to answer and that's going to produce 97% of the quality of the output by saving 97% of the energy required. Exactly. And David, to be honest, we've just like

splurged a bunch of stats and a story to folks. Can you pull up this chart that I just sent you, please? So it basically displays what we're talking about here. So what are we looking at on this website? So artificialanalysis.ai, by the way, just basically tracks all the top models, whether it's closed source or open source, and ranks them against each other. Now, if you look on the left, we look at overall quality of these models. You'll notice that O1 is there, right at the stop, right at the front. It's got...

all the bells and whistles, but right next to it is suddenly DeepSeek, R1, out of nowhere. And, you know, it's just a single point below it. Okay, okay, but what about the speed? You know, we're like, come on, I heard this thing isn't very quick. Well, actually, if you look at the speed tab, it is the slowest. But if you pay attention to what's right next to it,

There's a familiar brand there. It's O1. So it's actually not too indifferent from O1 when it comes to like reasoning and delivery. And finally, David, I mean, the point that we're making is this thing is like so much cheaper. If you look at the price, right? O1 is so expensive. Yeah. I mean, DeepSeek is at a four and you're looking at O1, which is like over six times as expensive or around six times as expensive, which is just nuts. So clearly like this just captures the breakthrough here.

And maybe to get to why this rattled the markets so far, it's like so hard on Monday, is that the idea here is that through creative use of this medium of experts and a few other efficiency gains, this DeepSeek R1 model is measured to be 45 times more efficient than its competitors, its United States-based competitors. And so what that means is you're using 45 times less data

compute resources in order to get the same response in about the same amount of time. And as a result, people are like, well, realizing like, well, this was a major, in the tug of war between software and hardware, between the AI outputs, software just got a big victory. And it's diminishing the value of hardware because we're realizing we can eke more out of our models by doing things other than just throwing brute force resources

at it. There's like this ancient, not even ancient, I've just used this metaphor on Bankless probably like 20 times over the years to talk about scaling, how scaling works. And I use this in the context of blockchains. There's two different ways to get scale out of any system whatsoever. You can write more efficient software

And you can just build stronger, more performant hardware. And I think maybe a metaphor that people can relate to is like early Xbox 360 games versus late Xbox 360 games. Grand Theft Auto San Andreas, or maybe it's Grand Theft Auto 4, came out on the Xbox 360 and so did Grand Theft Auto 5. Same hardware, same hardware.

different games and you can just look on screen as to like how you can probably count the polygons on Grand Theft Auto San Andreas and in Grand Theft Auto 5 you might be able to mistake that for like a real photo and so this was just an example of these Grand Theft Auto teams writing better software to use the same amount of the hardware better to produce a better product and

And then you can also scale hardware and just now we're on Xbox One, so the hardware is even better. And so this is how blockchains work, right? We can either scale a blockchain by writing better software

And we can also scale blockchain by using better hardware. Honestly, the answer is always going to be both. And this is what we're going to see out of LLM models is we're going to write models that use compute resources more efficiently. And with the creation of DeepSeq R1, it's 45 times more efficient than previous models. And then also, we're going to be able to use more hardware as hardware improves.

People have been talking about Jevons paradox in this like NVIDIA market crash, tech crash that's already recovering. But Jevons paradox, I think, is something like useful to understand. Jevons paradox states that as technological advancements improve the efficiency of a resource, its overall consumption can paradoxically increase dramatically.

rather than decrease. This occurs because increased efficiency lowers costs and expands potential use cases, driving greater demand. AI is now a major economic driver, and Jevons' paradox suggests that as compute gets cheaper, AI adoption will spread out faster, expanding demand rather than reducing it. So, in short...

as NVIDIA and DeepSeq improve AI efficiency, they don't reduce overall compute consumption. They make AI more accessible, leading to greater global demand for GPUs, energy, and data infrastructure. And so this is why- So if I were to left curve thesis what you've just said, David, because I think it's very important and I want people to understand. DeepSeq made models that are groundbreaking much cheaper to make. Yes. Which means that there's this massive supply

surplus of compute and everyone's panicking. They're like, oh my God, we have all this compute. What are we going to use it on? Maybe we've been totally over-indexing on compute. Maybe NVIDIA is super overvalued. Yes. And so what you've just said is, hang on a second. No.

No, there's this paradox which kind of explains that this compute will simply just get utilized kind of at the app player. It's induced demand. It's induced demand. It's induced demand. So this groundbreaking discovery has basically meant that now more models are going to exist, which means more apps and more cool things are going to get built, which will consume that same compute that is the surplus. So eventually we'll still have over-demand for the same thing.

Totally. And I think listeners are probably looking at the crypto AI token prices over the last two weeks and they're like, oh, uh-oh. And then maybe they're thinking, all right, well, how does DeepSeek and NVIDIA and all of this impact my bags? What about my virtual tokens? And so I kind of want to just attack that conversation head on. DeepSeek R1, the introduction of R1 is inherently, fundamentally, massively bullish for

for the consumer. The AI sector is going to become more useful, more efficient, more just like we're gonna be able to build better products faster because of something like DeepSeq R1. Nonetheless, Nvidia is still, has still like sold off. It started to pick back up. Here's the chart. I'm looking at the chart right now. It started to pick back up a little bit yesterday, but it's still down

14% since last Friday, which is hundreds of billions of dollars. And so I kind of want to just talk about the stars that need to align before our AI tokens go up in price. Because AI tokens, AI crypto, AI agents, it's a niche within a niche. This is a single sector of cryptocurrency.

crypto and crypto is a niche inside of broader finance. So we need broader, we need macro to go well, actual macro. And that's the Nvidia price. That is the performance of the stock market. That's also federal reserve monetary policy and Trump action. Nvidia is macro, right David? Nvidia is macro at this point because of how large it is. It's the biggest company.

And so we need macro. We need the traditional stock market to at least hold on rather than go down. We need the Federal Reserve to signal that it's okay to invest in risk on assets. We need...

And then Trump can also kind of do that same thing by just spending a bunch of money. That's actual macro. And then internal to crypto, we need AI agent tokens to be the meta. And I think people are going back and forth on whether or not AI agents and AI tokens are actually going to be the meta of 2025. I think if you ask people a month ago or two months ago, they're like, oh yeah, this is going to totally cause a bull market. I think now today people are not so sure. And so there's like a stacking of

contingencies that we need in order to have an AI agent driven bull market. And while nonetheless, like this DeepSeek R1 is fundamentally bullish, and I think Ejaz and I are going to say that this makes the AI agent meta, the AI agent part of crypto stronger because we now have another more efficient model to use these agents on. Nonetheless, we still need like a series of contingencies in order to like be bullish about crypto AI.

So I kind of just wanted to lay that conversation out because I think we're both bullish on the fundamentals here, but does that actually translate into AI agent tokens? Unsure. Yeah. It's a worthy reminder for the audience, David, and for us actually, that crypto is still such a young industry. It's still burgeoning and blossoming. So it's...

It is very dependent on macro. The macro people, by the way, are still looking at crypto and they've seen these ETFs and stuff. They're seeing Larry Fink talking about tokenizing everything. And only now, you know, it's like, oh, maybe we'll add a tiny little portion to a strategic reserve and see what happens. But we are still very much the younglings here. And, you know, whatever happens to the macro market at large, if they're like,

a major war or whatever that might be it will have a knock-on effect to more risky assets right um and crypto will always be the first one to hit um one thing i want to point out as well david is you know we we said just now that nvidia is macro and what we meant by that is it is the biggest company in the world it's the most expensive you know market cap for a single company um and it's ai related

So it has a direct correlation or one-to-one kind of branding with, guess what, AI coins in crypto. So if you look at like alt caps within crypto, you know, you've got like Bitcoin coming down a bit. It's still above 100K, which by the way, no one's talking about, which might be the most bearish sentiment on Twitter ever.

But, you know, other alts are kind of shaken out, but not as much as like crypto AI coins, right? And I think that there we have kind of like a one-to-one kind of relation between Nvidia leading the market on the traditional stock index and then like crypto coins having a direct branding thing here. And the point we're making here is,

the crypto markets are still very narrative driven. Fundamentals often like lag a bunch of these things, but in a bull cycle, narrative often drives like a ton of these different things. Now that is good or bad, right? Like if you're focused on the fundamentals and you have a thesis that crypto AI is going to be pretty huge, open source AI unlocked by this deep seek innovation is going to be

absolutely killer from now going forwards, then you should just be chilling, right? And maybe you would buy the dip, but again, not financial advice, but maybe you would have more conviction in your positions and see these prices as kind of like a bargain, right? But it's important to pay attention and understand how this game works, right? And understand that the price will not always reflect fundamentals. And you got to figure out what that means for your investing strategy.

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. What if the future of Web3 gaming wasn't just a fantasy, but something you could explore today? Ronin, the blockchain already trusted by millions of players and creators, is opening its doors to a new era of innovation starting February 12th. For players and investors, Ronin is a home to a thriving ecosystem of games, NFTs, and live projects like Axie and Pixels. With its permissionless expansion, the platform is about to unleash new opportunities in gaming, DeFi, AI agents, and more.

Sign up for the Ronin Wallet now to join 17 million others exploring the ecosystem. And for developers, Ronin is your platform to build, grow and scale. With fast transactions, low fees and proven infrastructure, it's optimized for creativity at scale. Start building on the testnet today and prepare to launch your ideas, whether it's games, meme coins or an entirely new Web3 experience. Ronin's millions of active users and wallets means tapping into a thriving ecosystem of 3 million monthly active

addresses ready to explore your creations. Sign up for Ronin Wallet at wallet.roninchain.com and explore the possibilities. Whether you're a player, investor or builder, the future of Web3 starts on Ronin. Are you ready to swap smarter? Uniswap apps are simple, secure and seamless tools that crypto users trust.

The Uniswap protocol has processed more than $2.5 trillion in all time swap volume, proving it's the go-to liquidity hub for swaps. With support for growing numbers of chains, including Ethereum, Mainnet, Base, Arbitrum, Polygon, ZK-SYNC, Uniswap apps are built for a multi-chain world. Uniswap syncs your transactions across its web interface, mobile apps, and Chrome browser extensions, so you're never tied to one device. And with self-custody for your funds and MEV protection, Uniswap keeps your crypto secure while you swap,

anywhere, anytime. Connect your wallet and swap smarter today with the Uniswap web app or download the Uniswap wallet available now in iOS, Android, and Chrome. Uniswap, the simple, secure way to swap in a multi-chain world. Let's get into some of the actual fundamentals that I think picked up this last week because, you know, usually it's me, Ejaz, and maybe a hundred other people on crypto Twitter, AI agent crypto Twitter, talking about AI agents. But this last week,

different people were talking about okay ai agents can you want to run us through this yeah okay so we're putting our metaphorical bull hats on now david we're done with the news it's time to like get into like why this is so exciting why you know deep seek and everyone is just pushing this entire ai space forwards okay david not just in a single day within a two hour window two of the largest

AI setups, not crypto AI setups, AI setups, open AI and perplexity announced their agent products. What do you have to say about that? So we had open AI demoing or teasing their agent product, David. This is something that we've been speaking about for months now. And we've been kind of teasing with the idea like open AI is going to launch something, blah, blah, blah. Open AI teased their agent. And guess what it did, David?

Aside from, you know, showcasing Sam Altman and our future overlords. It showed this agent doing a bunch of shopping. It showed this agent ordering pizza. By the way, you saw it here first on Bankless, an agent ordering pizza. So I'm just saying open source kind of like leads the way. We've already done that, Sam. Listen, Sam, just chill. We're ahead of you. It's fine. You can take a leaf out of our book. It's cool. But anyway, and a third point, which we'll get into later, which...

their agents are using computer use, David. Now, remember, we've always spoken about APIs, APIs, APIs, and it's important, but just an interesting little thing there. But let's move on to Perplexity, right? Which has a very slick kind of like Apple-like demo here, which basically demonstrates the same thing. So we have two of the biggest companies announcing agents at a time where DeepSeek is coming out with open source and all these new products are being pushed forwards, techniques, products, whatever it might be. Just an insanely like...

opening week for us. Let's watch the introduction of the OpenAI agent, just the first 20 seconds of Sam Altman as he talks about this. Good morning. We've got something exciting for you today. We're going to launch our first agent. AI agents are AI systems that can do work for you independently. You give them a task and they go off and do it. We think this is going to be a big trend in AI and really impact the work people can do, how productive they can be, how creative they can be, what they can accomplish.

Okay. So my understanding of this, Ejaz, is that whatever open AI builds with this agent thing, we just get to have that technology in the crypto AI agent sector, right? Whatever capacity that they're imbuing inside of open AI, do we get that or is it still closed off to us? So open AI is still closed source, just to make things abundantly clear, right? But what it

DES do for us in the short term, David, is it convinces us that we're either already on the right path or it corrects us to make sure that we're currently on the right path. And they've already given us a bunch of information just in this tease

alone, right? I mentioned computer use. I mentioned some of the things these agents can start doing. They are already crypto platforms that are doing this and we'll talk about it later. But it's just cool to know that we are at least directionally correct. But these guys are still closed source. I have a feeling they're going to do something like they've done already, which is like, you know, here's an API. You can query it. It's still censored. You can't open source it, blah, blah, blah. But I bet you that Meta is probably going to follow up with a model that can do something similar and

release their agent product? David, what if it becomes open source? What are you about to say? You look excited. Well, isn't China just going to do it? I feel like we're going to build all the proprietary high-touch white glove LLMs that cost $200 a month. And then China is going to be like, well, thanks for that. We're going to spend...

six million more dollars by riding on the back of all your guys's hard work and then we're just going to open source all of your labor yeah well i'm curious like how far this strategy is going to get taken right like meta did it within the us against open ai so that they can kind of attack them and like you know bring them ahead of of everyone else and you know there's been this like um thought that eventually meta will eventually

close source it you know because they're a private company they have shareholders etc and you know we'll go back to normal but like whilst this we're in this golden age like there's a lot to learn from here right now to your point on deep seek right um i think they will end up releasing an agent model and typically and it's not something we mentioned they open source this model which typically means that they've got a better model under closed like wraps that they have been already been working on for a while and i don't know if you saw some uh sam

basically mentions that R1 is very impressive, DeepSeek is super impressive, and he acknowledged that they have made advancements that they didn't quite figure out. And kind of this has been the take from like a number of OpenAI employees, as well as like, I'm pretty sure Jensen Huang of NVIDIA also said something similar being like, wow, this is groundbreaking, blah, blah, blah. So it's put people on their kind of like alert,

kind of scenario right now where they're like, okay, can we be doing something better here? And what does it mean for like the kind of like wider sense of things? Part of this is just heavily ripping off the, ripping the realization off that, no, we are in a deep, deep,

arms race about AI and the strategy of announcing that it only costs $6 million to train the R1 model, which is not true. That's like not, that's probably the one part of this whole thing. That's probably not real. Uh,

was an intentional strategy to knock down the valuations of US-based AI companies. There's a conspiracy that the hedge fund took a short position as well. Yes. Stuff like that. And then also, you know China will... China has, for decades, heavily, heavily subsidized their tech industry in order to commoditize...

some of the value created inside of the United States, copy it, and then produce it at scale, and then just have that become highly competitive with United States models. So it makes a ton of sense that they are heavily intended to take the value of United States AI companies

and open source it as much as possible because that slows down investment. They are trying to slow down investment in United States AI companies. And so anything that I think comes out of China, in my mind, I'm kind of viewing that, perceiving that as,

both trying to bolster their own AI tech industry, but also be an attack on American market values of AI tech companies. Yeah, yeah, for sure. I mean, I think it's a, I would personally brand it as a splash of very, very cold water on America's AI darlings, right? They need to figure out, you know, are they too bloated?

Are they directionally correct with their research? And are they building the right types of products? Because this is a wake-up call. And I kind of think you summarized it perfectly with the phrase, it's an arms race. That's literally what it is. If AI is going to penetrate every single sector, including...

you know, literally defense and, you know, open source tech and consumer apps and all that kind of stuff. Well, then you're going to need to like lead the frontier here. And so far up until like literally last week, it was pretty clear that the U.S. was the leader. And now it's kind of been brought into question. But I think this is going to be net net good for all of us, David, particularly. The fact that it's converging on open source at the end of the day. This is my point. Open source just wins. OK, OK, OK. So let's do the breakdown here, right? OK, we had

Trump come into presidency and announce a bunch of like really pro AI and crypto things, right? And he said, hey, we're going to have a specific person appointed to figure out the AI and crypto strategy, but we're going to be pro. Okay. Literally a few days later, he announced a $500 billion set of investments that's going to happen purely within the US for AI, right? Which is just

crazy, whatever, super bullish, right? Okay, we're leading the charge here. Sam Altman is leading that fund as well. Okay, wonderful. Then we have this deep seek thing come, right? And it's a cheaper, more accessible model. It's open source and it can compete with the best of the best. Okay, and

And it's from China, David, which is like the arch nemesis politically of the US. So now there's this branding of like, wait, this thing can be cheaper and accessible to everyone. So everyone's like perking up and like, hey, maybe I have a shot at building something cool in this industry and not just allowing meta and open AI do it. But also like, whoa, it's China. We're scared. Like what's going to happen? Blah, blah, blah, blah, blah, right? But they open source the entire model, which means that all the...

and we're going to bring it back to crypto here, people that are working on open source AI things such as agents, the combination of this and open AI's agent product is very telling, is just like the perfect mixture to allow this entire industry to kind of boom. Because it's like, okay, where's some of the most important agentic open source activity happening, David?

It's in crypto. It's in crypto, right? It's like some of the top teams are working on pioneering all these frameworks and platforms. So I think we're going to see a lot of investment push out into this. I don't know if you saw Jack Dorsey's tweet, by the way, just adding this in, David. He released an open source, he's calling it an agent update

on machine system, which basically, I don't know what exactly it does, but it's going to be his agent system that he's open sourcing for a bunch of different folk to kind of use. And I think he's calling it Goose. So it's super interesting regardless. And I think we're going to see a push for more open source AI development.

Oh, yeah. Let's get into some more crypto-native subjects. You have this tweet in the agenda from DGN News. Breaking Solana becomes the first chain in history to break $200 billion in monthly DEX volume. Why is this in the...

ai roll-up agenda how is this ai agent related okay so last week we spoke about the kind of like comparison of solana versus base and i actually think this theme is very important to introduce or reintroduce rather into this episode because there's been a number of announcements this week david which is suggested or is suggesting that solana might be the home

for a lot of AI agent activity going forwards. And I'm not saying that as a fact, I'm just saying it from like certain observations. So- You're saying that the tug of war between Solana and other chains, mainly base, is meaningfully moving into the Solana camp. Like Solana's winning the tug of war. Correct. So let me-

to tee us up for this, right? So number one, Solana's had like its best month of decentralized exchange token volumes. It's actually broken the all-time record, $200 billion. It doesn't take a genius to figure out why Trump launched his meme coin on Solana, which like broke everything. So it's not exactly AI specific, but I'm teeing us up for some of the analysis that we're about to break into, right? Secondly,

I think something worth noting, you mentioned earlier, David, like, you know, AI coins are down. We don't know whether it's going to be the same meta. People are like looking at social consumer apps separate to AI. They're looking for the next meta. They talk about hyperliquid. Every tweet has hyperliquid at the end. But it is still the number one most talked about thing, AI, specifically within crypto. If you pull up like this Mindshare analysis, basically,

you know, baby, we're back. Baby, we've never left, right? So it's still like the number one. 42%. It is still consuming everything. And you'll see on the right there, you know, we see meme, we see DeFi. That represents basically people being like in panic mode and trying to figure out, damn, is this AI thing even worthy? Should I chase other things? But AI is still up online

on the rise. It's green. Is DeepSeek, I think this is a DeepSeek-induced spike though. So whatever it was, we are higher than what we were. Yeah. Is talking about DeepSeek- I would actually disagree. Yeah. Oh, really? Okay. So disagree with me. Why is this not just a DeepSeek-induced attention, mind you? So it is DeepSeek-induced, but it's ingrained within the crypto stuff that's being developed within AI agents, David. So a lot of this chatter, remember, CryptoTutor doesn't just, sorry, Kito, which we're referencing here, doesn't just track

general AI stuff within crypto, it tracks like whether it has any relevancy to projects, to teams, to applications being built. And what I think this is capturing accurately, David, is all the teams and protocols which announced, hey, by the way, it took us five minutes to integrate DeepSeek into our framework. It's there if you want to use it for any of the agents that you've spun up.

And David, that's in fact the case. You know, I can name like a bunch of protocols off the top of my head. Eliza framework, Arc framework, virtuals,

Venice AI, which we'll talk about later, which had their launch with their token this week. All of them have DeepSeek already ingrained and running locally, which means that it's accessible to all, can't shut it down, et cetera, et cetera, go and enjoy it. So it's capturing basically crypto's number one use case or maybe number two use case, which is its resilience.

You know, it doesn't matter what's happening in the world. If there is something good that comes out of it, it'll adapt it or adopt it rather and use it for its own good. And it benefits every single agent. So now, you know, let's assume there's like, I don't know, 100,000 agents there built on crypto rails. Now all of those agents pretty much or 90% of those agents get access to this amazing pioneering new advancement. It is worth, I think, really reflecting on that idea that

it doesn't matter what open ai builds or meta builds or even like this deep seek a new ai lab builds the natural convergence if it does naturally converge on open source gets it and like why do you pay open ai for 200 a month well so you can get the shiny new model like three months six months sooner than open source but the idea here is that eventually the value of all the ai llm and all that stuff eventually shows up

in the open source world, which is our world. And that value comes to be reflected on open source, open blockchains in the form of tokens. That's very bullish. That's very bullish to me. There's a bunch of other subjects that we're going to get into. Virtuals. We have to talk about virtuals. AIXBT becoming a product is something you want to talk about. Venice, Eric Voorhees project launched a token that is part of the whole sovereign AI thesis that I want to map together. So we're going to get into those subjects. But first...

Bankless Nation, you guys got to know the tax season is around the corner and Crypto Tax Calculator is probably the best place to get your crypto taxes done. If you have been trading AI agent tokens, buying and selling them, you have a taxable event that you need to account for. And you can use Crypto Tax Calculator 30% off with a link in the show notes if you are a bankless citizen. Otherwise, you can get 15% off with the code bankless.

We've been big fans of Crypto Tax Calculator. We've been using that at Bankless for years. I have an accountant that I just pay money so I don't have to do this at all, but they just use Crypto Tax Calculator. And so this is the product that my accountant uses. And so if you guys want to check them out, there is a link in the show notes to get started. EJ?

EJS, tell me about virtuals. Some very big news happened on Sunday or Saturday, I think. It was a weekend announcement. Walk us through the news that happened in virtuals this last week. Okay, so context here. Virtuals, one of the biggest agent...

protocols within the crypto AI sector. It's pretty much led the market and it was birthed on the base chain, David. Well, on Saturday, they announced they were making a pretty strict move or a pretty bold move rather, which is they are going to be also launching on, you guessed it, Solana.

So this obviously churned up a bunch of news, right? Mixed reactions, like some were incredibly overwhelmed. They were like, this is the best news ever. If you remember, Virtual's mission is to build up an agentic

society or a genetic economy, and it should know no bounds regardless. So it should use any chain. And this is the first major move, territorial move rather, that we're seeing some of these protocols move from one chain to the next in a very dominant form. So what does this mean, David? It basically means virtuals

as it exists today, you know, it has a platform you can go onto. You can design an agent, launch an agent, integrate with its frameworks or any other framework, as we mentioned last week, you know, they have this new API. You can now do all of that on Solana. So obviously it bodes a bunch of questions. It's like, well, hang on a second. What about the virtuals token that I own? Right.

On base. Because the virtual token built, deployed contract address is on base. On base. A token can only have one contract address. Yes, correct. And it's where, whatever is deployed first is its home now and for forever. Exactly. So everyone's like, okay, well, are they going to be minting more tokens? The short answer is no. Don't worry. Your coins are good. It's going to be the same thing. There's a one-to-one comparison, but they're using, I believe, layer zero to transport

some form of liquidity or portion of their tokens, be it treasury, to seed the pool or seed a bunch of tokens that they have on base onto Solana. So think of it as having the Solana version of their token, but it's not using any more or any less. It's just literally the same coins being transported over to another chain. But they are literally expanding their economy here, David.

They're going to have the same kind of rails, the same kind of ability. And I think they're adding a few other things, David. They've mentioned that they're going to be introducing something known as a venture partner, which I believe is going to be something along the lines of they'll recruit someone like the best minds or what they believe to be active contributors in the Solana AI ecosystem to help curate and launch projects online.

on their Solana version of virtuals, which will allow them to have some kind of like a seeding here in terms of like building something awesome. But I mentioned, David, that there was some really overwhelmingly good reactions. There were some overwhelmingly bad reactions as well. Okay, understandably, from the base ecosystem, they were like, hang on a second, are you guys just abandoning us? What about the agents we launched on base? Like, what are we going to do there? The

TL;DR of the response from the team is, "We still love you. We are still putting the majority of our treasury and support and resources to you guys. We're just moving a chunk of it now to Solana for now." But also, they're figuring out ways where these agents on base, with their tokens on base,

could also potentially move to Solana as well, David, which means that fees that are aggregated in the same way in Solana or on base will get fed back into these same agent tokens, whether they're on base or whether they're on Solana. So everyone technically or theoretically should win.

Every agent on virtuals technically should win. I think if you are a chain maxi, you're just seeing this as a big win for Solana and a big L for base. I mean, nonetheless, like AIXBT, that token is on base. That contract address is on base. I think, correct me if I'm wrong, but

The idea here is you can now issue a virtuals agent and his token on Solana and that be a Solana native token address, token contract address. And then also tokens, agent tokens that are on base can use layer zero to become, have liquidity on Solana, access liquidity on Solana. So that last part, I'm not entirely...

confirmed that it's live yet. Okay. But I believe it's in the works. I saw Jensen say that he hired a token economist to map out how this is going to happen in the kind of phase structure. But it's their full intention that all these agent tokens on base can operate and work on Solana. Yeah.

Okay. So it's really the idea that new virtual agents can issue their tokens on Solana. I think that's kind of the big punchline for everything to me. Everything else is kind of more marginal than that. It's that, but also just expanding the TAM.

of your potential users. It's been no secret that Solana has kind of been the degen chain of the year. Right, it's got the hop all money. Yeah, it's got like the most activity, the most, you know, you could argue obviously this, but like the most kind of innovation or new primitives that we're seeing. And I think it's where a lot of the attention in general from the public on crypto is,

at mass has been, right? You know, Trump literally launched his meme coin on Solana. So it'll be interesting to see how this plays out and whether it, you know, attracts the same type of quality of builder, maybe a different kind of quality. Remember, you know, Solana is written in Rust. So there's going to be like, kind of like, is it going to be a one-to-one comparison? Are we going to get

better quality engineers or worse. It'll be interesting to see how this kind of works out. But David, some of the top protocols or rather teams that built on virtuals are kind of moving over or started to migrate over to Solana as well, David. If you pull up this tweet from Veda, so Veda is one of the top investment DAOs on virtuals. They're going to have some liquidity prevalent on

the salon chain. And there was also another team that migrated, David, I think it was Mika, which, you know, is kind of outlined the entire kind of like roadmap or plan of how they were going to do it. So the point I'm making here is there's obvious demand from salon ecosystem, but there's demand from the base guys to also kind of like, get into this kind of like salon attention economy. So just

a massive and major announcement, which I think a lot of people weren't kind of like reacting to. People are hypothesizing, David, that also it's because Coinbase didn't list virtuals. And so they were kind of getting- Huge L, huge L. You know, and I'm kind of like thinking about like, has this happened before in a similar analogy? And it kind of got me to think of Pengu.

You know, the meme coin that the Pudgy Penguins came out with and the fact that they just launched it on Solana kind of reminds me of that, except that they didn't have the listing exchange issue. But yeah, virtuals, you know, didn't get the Coinbase listing. And, you know, some people are still old school. They're still trying to be like, hey, you know, what's up? Like, why can't you pay us the attention that we need? Well, like one of your biggest tokens, we attribute for like, I think 50% of your DEX volume at one point. Like, how are you guys not figuring this out? Right? So anyway, like, I think like,

Stepping away from this Solana launch, one final thing to mention is the founders and the team, David, are still so locked in, dude. It's pretty nuts. I read this interview. I don't know if you have it pulled up, but with one of the co-founders of Virtuals on their recent trip to Asia, they were doing a bunch of ground business development work. And I just loved...

that he took the approach of first principles to everything that he was building. He wasn't focused on the token or any of that. He was like, how can I build a good product? And I found this quote pretty funny where he was asked, why didn't you launch on Solana? And he was like, well, it was because we didn't know how to code in Rust.

And obviously, like since then, there were things that were in the works, they were either collaborating or working with Solana engineers to figure out how they can actually, you know, deploy those smart contracts. But there are three big takeaways from this interview that I kind of want to highlight before we move on, David, which is they're working on three things on their roadmap, which I think shouldn't go underscored. Number one.

Better agent interoperability. So this means how do agents talk to each other and how can they end up like working with each other? You know, we've seen Luna, their flagship agent, hire another agent to do something for it. What if that became commonplace amongst agents? Something super cool. Number two, sharing agent revenue with token holders and community that helped build the agent.

Now we're thinking of like these agents as like investable assets, potentially, right? And like, obviously, there's a discussion around securities and all that kind of stuff. But if your agent is a productive,

being or entity and generates a ton of revenue, now you potentially may be able to get exposure to that kind of thing. Like typically they've been doing buyback and burns, but what does a revenue distribution look like is super interesting. And the third and probably the wackiest one, David, is physical integration with the real world. So AI agents running pop-up shops is going to be that first iteration. Yeah, yeah. I think...

I believe, I forgot the name of it, but they're basically going to be setting up a coffee shop in Mexico as like their first kind of like test case. And it's going to be run by an agent. What that means, I have no idea, but hopefully the coffee tastes great.

Okay, so one question I have, if you go to the virtuals price, I'm looking at CoinGecko. Down bad. Let's go by even a month. It's down only for a month. And the integration with Solana, which I'm guessing just like using my left curve brain, it's just because that's where the hot ball of money is. That's where the liquidity is. It's good for new users. It's good for increasing the, you know, you want to be next. Everyone wants their token next to the hot ball of money and that hot ball of money is on Solana. Yeah.

Uh, this actually just didn't show up in the virtuals price. Uh, and also virtuals has been down only for the entire month of January. Uh, so, I mean, I'm going to ask you the impossible question that no one can actually ever answer, but just like, what do you think is going on with the price?

Yeah, I mean, I would say that this price chart and pattern that you've just pointed out, David, is going to be reflective across a ton of other AI coins. And we should double check that, right? But basically, we had a massive run up with these crypto AI coins by the end of last year. And then we've pretty much been in a sell off period since. I mean, look at AI16Z. It's kind of been like a very similar story or pattern there. So it kind of brings up the point that you mentioned earlier, David, which is,

crypto is very narrative driven. And what drives the narrative the most? It is macro. So if macro is not doing too well, or if there's a bunch of things happening in the US government that apparently has a massive reflection on the rest of the market, we will see similar things here. So crypto still isn't grown up enough yet or treated as a mature enough asset just yet, which makes sense. Regulations are

haven't caught up, but they will. But it's not mature enough to get treated as its own separate kind of trend that's going on. So that could be your best knowledge, right? So if you're like, oh, well, I think, you know, the fundamentals aren't being reflected in the market, that could be a really good buy opportunity. Or you could just be impatient and let the price, you know, reflect your own narrative or investment thesis. It really depends.

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.

As the home of the stablecoins, Celo hosts 13 native stablecoins across 7 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,

I think of all the charts out there, AIXBT stands out to me the most as

I mean, it's off of its all-time highs from just like 15 days ago, but otherwise the chart looks pretty intact. Talk to me about AIXBT. What's going on over there? Dude, I'm super pumped about the AIXBT team. So...

Major news this week is they announced their kind of like product. So if you remember what we mentioned on the pod a few weeks ago was AIXBT is cool. It delivers financial alpha, but my guess or our guess was they would eventually dog food a product to people through this, right?

And their product at the time was, hey, you need 600,000 of these tokens to get access to private, sorry, private access to the terminal, right? So it could give you your own tailored alpha. That was a lot of money. That was, I think, $150,000 at the time. And so, you know, very inaccessible to most.

So what they did was they launched this new AIXBT terminal tiered system, which now allows a range of token holders of AIXBT to get differing access to the agent terminal itself. Now, I'm not going to go through all the tiers, but basically, if you scroll down, David, it gives you varying kinds of access to different

the interface, what kinds of alpha or questions you might be able to ask it, and gives you a general idea of where this thing is going from a product perspective. But obviously, that wasn't enough information for us. So we jumped on a call with the AXBT team this week, and they gave us their insight into what they were thinking. And

I'm not going to get into it because I want to save it for a pod episode later where we formally speak to AIXPT and the team. But the two main takeaways was, number one, I really like that they're trying to get AIXPT into different types of mediums. So right now it exists on Twitter.

in written form, but what does it look like from a voice perspective or a digital representation on video? What does it look like when it is existent on different social media platforms or just like entirely non-social platforms? Like, what does that look like? Is it your assistant? What does that look like?

The second thing I thought was really cool was they're kind of like plans for like a B2B approach, which is like, "Hey, if you're a protocol that's trying to figure out your token economics, wouldn't you kind of want to speak to the smartest agent in the room that's already digested a bunch of these things from a million other teams and knows where the kind of market is trending and could give you advice on like, 'Hey, I think you should structure it this way or that way, or you should do A, B, and C.' What if you, David, had your own private AIXBT agent just

pop up, you know, just sitting on the edge of your screen there, like kind of like Clippy back in the day. What would that be? Like, how would that look? What would that price pricing system look like? Some really cool, you know, products coming out of this. One big bit of news that rocked the AI and just crypto world generally, I think this was maybe the news of the week, at least so far, is the Venice airdrop. Venice is kind of like a private duck

DuckDuckGo version of ChatGPT where you don't need him sign up for an account to make a query with ChatGPT. So it's branded as like private, private ChatGPT. This is built by Eric Voorhees and this is very downstream of his old project Shapeshift.

which was once upon a time forced to implement KYC. All it would do is you would input one crypto asset on one chain and then you would give it another address on a different chain and then ShapeShift would swap the assets and send it to a different chain. You never had to do any sign up because Eric Voorhees hates KYC. Venice is something similar. Just no sign up for using this LLM.

Connect to wallet. And Venice launched the VVV token, which instantly zoomed up to $2 billion, fell back down to about $1 billion until Coinbase announced that it was listing Venice, and then it went back up. Much to the virtual fan's disgrace. Yes, right? Yes, exactly. Like a very quick token listing, which allowed it to break all-time highs, again, in the same day it was launched.

At $23, $2.3 billion, I think is about the total valuation. And that has been the story of Venice. And I think everyone kind of scrambled. If you are a listener of this podcast and you hold something like AIXBT, which is a token on base, Clanker, a little AI bot to drop tokens, also a token on base. Many AI agent-based tokens, if you held any of these things, you got an airdrop.

And so many people have claimed their airdrop, pretty fat airdrop. And then it made its way into a listing on Coinbase and it already trades on Coinbase right now. What did you think of this event when you saw this event? Okay. So there's a few things here. Number one, love that Eric Voorhees is building out this product. You kind of summarized it well, but rather than think of it as like

the ChatGPT, think of it as like ChatGPT, but it has access to any model that Venice connects it to, right? So think of Venice as like the front end and it will query whatever model you want it to, right? And that could be potentially centralized, but preferably open source locally run models, right?

The second selling point of Venice is that it's supposedly private, right? And there's a bit of debate as to how private it is, but it is private. And the way that they do this is it's privatized cryptographically at the node level. So when you make a request, it goes to a node, the node privatizes it, and then like, you

a query off chain, pulls the output, again, cryptographically, privatizes the response and then sends it back to you. So it doesn't technically know where the query came from your IP address or any of that kind of stuff. It only exists on your browser doesn't store any information. And to your point earlier, you don't need to KYC, you just connect a wallet address and you get on with it. And it has a free version and a pro version.

So, a really cool product. I think what was also cool about this is as soon as DeepSeq was released, Venice just plugged straight into it and ran locally. So now you had the best model and access to it. So you could just jump on Venice and have some of the smartest returns, maybe smarter than the leading OpenAI 01 model. So very cool. Now on the point of the Coinbase listing, David, there's a few things that I think about this. I think

like, you know, they were working together by some means because, you know, it seems like Venice is like being plugged in as an API as part of their AI agent SDK kit, which has been, you know, very popular in terms of setting up certain agents. But number two, it brought into question what their listing process is like. And, you know, this isn't the first time this is being done, but it's being really focused on for the last week. And I want to kind of like

point out, I think a tweet or a comment that Paul Graval made on this. Paul Graval is like the chief legal officer. And what he basically said was, listen, we hear you and we are now at a point where we really need to revise what our listing process looks like because the fact of the matter is there are a million tokens being released every

Is it every week? Yeah, it's every week. And we just can't feasibly keep up the same standard and structure that we're taking to diligence and list tokens. Otherwise, we're just going to miss out on a massive market. So whether it was Paul Graball or someone else that mentioned this next specific point that I'm about to make, which is,

they kind of want to maybe explore doing a list all approach and then just retroactively ban or prevent tokens that are in ill faith. So they're looking at ways to do this

via the centralized exchange, but also through decentralized exchange ways. I think Brian put out a point here where he was like, Brian Armstrong, the CEO, where he was like, we shouldn't be constrained by this. Whether it's a DEX or a SEX, we should be able to get access to listing tokens for all our Coinbase users. And I think net-net, that is a productive output from all of this, David. - I also wanna talk about just the price performance. So as soon as the VVV token was launched on Coinbase,

It instantly went up. It did like a two, two and a half X, but it's been down only ever since. So it peaked at $22, it's at $9. And I want to compare that to Toshi, which is a meme coin on base. And to me, the data here, the interpretation that I'm giving is like, well, let's look at what happened when an AI token was listed on Coinbase and was made available to retail versus what happened when a meme was listed on Coinbase.

And I'll say the Toshi token listing, well, it always kind of peaks at token listing, but the Toshi price has really maintained its value post-listing. And the Venice price is kind of down only. Now, this is a very, there's hugely, massively apples and oranges because Toshi has been a token for decades.

over a year, and then VVV is a token for like two days now. So it's not the same. But what do you think about like retail appetite for means versus AI tokens on Coinbase? Yeah, I think you raise a really good point, which is, you know, the general archetype of some of the degens and particularly degens that are trading on chain. So that's, you know, not necessarily using centralized exchanges.

are just completely dispersed. It's like a split personality, David, where it's like one time we're like, oh, we're pro-AI fundamentals. And then the next one, like, oh, it's all a lie and we should just go to memes and financial nihilism TM. Do you know what I mean? So there's definitely part of that. The other thing though, and...

you know, the Toshi or supporters may not like this. And again, I'm going to put my hands up and say, I don't know whether it's true or not, but I believe the supply of Toshi is pretty concentrated amongst, you know, a set of holders. So that might also be a reason why, you know,

price was able to pump so quickly and so vertically. Just because the flow isn't very high? Yeah, yeah. Is it illiquid? Yeah, and again, like fact check me on this, I could be wrong, but that could also potentially be a reason.

All right, EJ, as we're at an hour, let's kind of burn through the rest of the news here. So Griffin, ARK, AI16Z, we're going to burn through these. What's going on with Griffin? Okay, so Griffin context here, we've mentioned it on the show before, but think of it as like an agent execution platform. So you basically can go onto it. It's like a chat GPT, but hey, there's a bunch of agents that can do on-chain stuff for me. Well, David, that kind of changed this week when they added an off-

chain web two very boomer component david which was shopify integration so now this sounds what sam altman announced at open ai is just like your agent that can do things yeah it's it's almost as if the open source um crew are like front running um a lot of what the centralized corps are focused on which is you know kind of bullish

But yeah, so a bunch of the agents now are able to integrate with Shopify and do a bunch of Web 2 products and services with you. So what this looks like and what you're looking at in the demo, if you scroll down, actually, David, I think you should be able to see an agent ordering cold brew, a pack of cold brew coffee cans for the particular individual. This agent in particular, I believe, is Agent Kitsune. And if you watch this demo, it's basically like ordering a pack of cold brew.

So why is this important? Why am I even mentioning this? I think there's a secret trend that no one's really talking about, David, which is what happens when you combine a Web2 product or service with a Web3 innovation or tool and strap a token to it, right? We've spoken about all the crazy things that have happened in crypto before, which is like we're launching NFTs, but it's Web3 native. We're doing, you know, DeFi, but it's on chain and blah, blah, blah.

Okay, but like the normies don't get it, David. They don't understand what this means for them. You need to make it somehow relate to them. Now we have a vector or a medium, which comes in the form of agents that will allow you to, you know, hey, could you cash out some of my crypto portfolio in Coinbase and then use that to buy me the groceries for the week? You know, what does that look like? Why is that important? Why should we be paying attention to that? Well, it's going back to our thesis that agents are going to make all of this really easy for people to use. They don't even need to know that blockchain is happening

on the backend. But I thought that was super cool about the Graphene thing. - And Arc, the announcement here is we're teaming up with the Solana Foundation to accelerate AI innovation on Solana. Arc, okay, I remember you telling me about Arc. Arc is all about Rust. These are the Rust AI agents. It's harder, but it's more performant.

Solana also built in Rust. It makes sense that the Rust AI agent platform is going on the Rust chain. It's weird that they are announcing they're teaming up with the Solana Foundation because I always already figured that they would already be fully integrated with Solana. So what's going on? Listen, there's nothing better than a formal headline on...

x to to pump your price david but actually the the complete opposite happened uh nothing happened if this happened uh a week ago if this headline was made a week ago uh or maybe even two weeks ago this would have absolutely sent arc but uh this is going very much under the the radar um and so what's interesting here or like the the wider point is

A formal partnership with Solana with the crypto AI space hasn't really happened yet. I don't think we've seen the same on base, so maybe I'm wrong. But Solana Foundation partnering with an actual protocol that embeds or will create a bunch of these agents has not been seen before. So very understated announcement, but I want to raise it on this. Super, super cool.

And then lastly, AI16Z, which I saw formally rebranded because A16Z, actual real A16Z, finally probably sent them a cease and desist. Is that what's going on? Okay, so these are all rumors, and so I don't know the exact specifics, but actually initially when AI16Z launched, the real A16Z reached out to them and said, hey, this is super cool. Actually, if you remember, Marc Andreessen called out AI16Z saying,

this is a really cool project. I hope you guys kind of make it. And actually there's rumors that the A16Z team helped them set up their legal structure within the US, which is like pretty awesome, right? Like very hand in hand, centralized, decentralized, like, hey, we're going to figure this out together, right? But of course, you know,

All things must settle down eventually. And there are some T's to be crossed and some I's to be dotted. And one of these things potentially is around, you know, namesake and copyright. And of course, A.I. 16Z originally started as kind of like a parody for A.I. 16Z. A very small, like a $10 million parody, which is no longer now that. Yes.

Yeah, so basically the likeness was too close and they wanted to move on to something that was more formally separate. And I also have a feeling, David, that this is also interfering with exchange listings potentially. So they want something that is like separate that they can not be sued after basically.

So that's pretty cool. The second important thing here is they announced a $10 million fund in conjunction with Jupiter Exchange, which is like the number one decentralized exchange or aggregator on Solana. And why this is super interesting is yet another conduit to build and support teams that are building within this space. And I think that that's super important. We've seen the rise of investment DAOs. We've seen the rise of...

traditional VC funds, raising funds specifically for the sector. So it's cool to see some of the core protocols dedicate so much funding towards this space. Pretty awesome. Cool.

All right. Bangalow Station, that was the AI roll-up. Ejaz, thank you once again for guiding us through all the weekly news. It moves faster than ever. What are you paying attention to now? What are you looking forward to? Let's get a little peek into the head of Ejaz. What are you doing over the next seven days or so? Okay, so I'm looking at unsexy agents, David.

Unsexy agents. That sounds very unsexy. So yeah, very unsexy. So sexy agents are the agents that you see on your Twitter every day that talk to you, that make you laugh, that order you pizza, you know, all those fun, real things. I'm looking at the guys behind the closed curtains, David. The ones that are- Middleware agents? Yeah, the agents that are delivering the food, you know, and the gig economy that's happening behind the scenes that are aggregating all the important data that you need to feed you the perfect responses that you need to.

I think that with all of this like, you know, deep seek news with NVIDIA crashing a little bit, there's going to be a flight to quality, to fundamentals. And I think people are going to be...

And putting on their framework hats and looking at a bunch of these unsexy agents that are driving kind of the fundamental usage. They're going to look at like infrastructure plays and all that kind of stuff. Second thing I'm looking at is computer use. I've mentioned that on previous pod episodes before, but with open AI and perplexity teasing computer use for their agents, I have a feeling there's a trend that's going to emerge here on the open source side of things. So digging into those things. Cool.

Bankless Nation, you guys know the deal. Crypto is risky. Crypto AI is probably even riskier. You can lose what you put in, but we are headed west. This is the frontier. It's not for everyone, but we are glad you are with us on the bankless journey. Thanks a lot.