theCUBE + NYSE Wired: Crypto Trailblazers | Daniel Marin, Nexus

Clips
More from theCUBE + NYSE Wired: Crypto Trailblazers

Daniel ‌Marin

Founder & CEO

Nexus

play_circle_outline Exploring Nexus: A Layer One Blockchain Innovator Focusing on Scalability and Proofing in Technology

play_circle_outline Comparison to Nvidia's shift from gaming to AI and supercomputing

play_circle_outline Nexus' goal of building a distributed supercomputer system

play_circle_outline Revolutionizing Compute Performance: Nexus' AI Vision and Verifiable Computing Intersection

play_circle_outline Expanding Nexus Global Testnets, Mainnet Release Plans, and Supercomputer Joining Process Innovations

Info

Daniel Marin, Nexus

Daniel ‌Marin

Founder & CEO Nexus

search

>> Welcome back, everyone, to theCUBE here at Palo Alto, California. I'm John Furrier, your host of theCUBE. This is our Crypto Trailblazers two-day event. Everyone's in town from the Theorem Foundation in Singapore, or activating Stanford and Berkeley Cal. All the entrepreneurs are out there. All those other thought leaders coming through, sharing their vision and what they're building to usher in this next generation. Daniel Marin's here, Founder of Nexus. Welcome to theCUBE. Thanks for coming on. We're having a good chat and we should have just turned the camera on before we got on, but thanks for coming on.

Daniel ‌Marin

>> Thank you, John. Yeah, it's great to be here, and it's great to be here at theCUBE. As I was saying, we love cubes at Nexus.

>> Matt Pats was just on. You guys were roommates, post Stanford. Big community here in the Bay Area. Obviously, I mean-

Daniel ‌Marin

>> High density.

>> Yeah, very high density of intellectual power. But it's just not just tech, there's this pragmatic thoughts around how to make money in a new way, the impact around technology. So again, the movements are happening. Ethereum's been a movement for a long time-

Daniel ‌Marin

>> It has....

>> and stories power movements. And we're so glad you're here to tell your story and also do the commentary and contribute to the community. Tell us about Nexus first, what you guys do. We'll jump in, set the concept. Take a minute to explain what you're working on.

Daniel ‌Marin

>> A hundred percent. It's very simple. We have layer one blockchain and we advance serial knowledge proofs. So, we work in the field of verifiable computing. So, we have a team of a lot of mathematicians and scientists that work day and night in accelerating this new field of verifiable computing. So, Nexus actually started this research project that I was working on at Stanford. And two years later, we're a team of like 35 people here in the Bay Area. And it's pretty simple. We actually have a lot of machines that we've put out in the industry. You can think of them as chips, but these chips compute serial knowledge proofs. And serial knowledge proofs are utilized to scale blockchain architectures and bring what we call verifiable computation to life. It's an abstract concept. It's not easy to grab, but verifiable computing is something that allows you to prove whether something is true or false. And that has a lot of applications on many markets. What we say is that we're like an open AI of verifiable computing, as opposed to intelligent computing.

>> So, talk about the relevance of that. Because in the math, which is blockchains, a lot of math involved, the importance of the proving, you hear proof of stake, everything is about proof.

Daniel ‌Marin

>> That's right.

>> this is the scale piece, because value store, okay, check. We understand that concept. But this idea of scale and value proofing becomes a huge thing. What's the driver of this? What was the problem to solve? Was it scale? Was it a mechanism? Was it software? What was the main driver?

Daniel ‌Marin

>> Yes, yes. So Nexus, we as a company, work at a bigger space than blockchain itself. Verifiable computing is a new form of computation, and that's something very weird to say. Something like AI or AGI is a new form of computing. Verifiable computing is a new form of computing in and of itself. And just as how AI models like 03, 01, et cetera can be utilized in many markets, verifiable computing can be applied to many markets as well. So to give you a few examples, human identity. You can prove whether something is true or false about who you are. Private computing, being able to do computations without revealing your private data. High security computing. So if you're part of the defense industry, to be able to show that a computation was done correctly without mistakes. And blockchains are a subset of verifiable computing. They're distributed verifiable machines. And it's well known that all of the scaling propositions for long term for blockchains are related to serial knowledge proofs and verifiable computing. And so for us, blockchains is a huge industry financially speaking, that we target.

>> So it's a segment for you.

Daniel ‌Marin

>> It is a segment .

>> So you brought up AI, so I want to kind of maybe make a comparison and get your reaction to. If you look at Nvidia, Nvidia was doing gaming, they were a graphics card. Now they're bigger than Intel. Look at the stock price.

Daniel ‌Marin

>> Yeah. It's crazy.

>> They started as a graphics card, all the gamers used it, and then became graphics systems for gaming. Crypto, great part of mining. Nvidia, thank you, Nvidia. Thank you for the blockchain enablement. Then they go to AI and now it's a super computing revolution.

Daniel ‌Marin

>> Yes.

>> Okay? So, Jensen Wong talks about accelerated computing all the time. With that kind of paradigm, compare what you guys are doing. Because I see a little bit of comparison here-

Daniel ‌Marin

>> A hundred percent, a hundred percent....

>> with what you're doing. You're accelerating, you're verifying, which is actually, I'm just riffing on this, but I can see how what... Because he talks about accelerated computing as a paradigm. And it's a new thing, it's a category.

Daniel ‌Marin

>> It's a new category that they're building.

>> It's not better compute. And now-

Daniel ‌Marin

>> It's a different type....

>> you've got AI factories. So compare, how would you tell about what Nexus does if it was a categorical shift as equivalent verifiable computing... Obviously, see there's an acceleration in there, but not in the sense of GPU. Talk about the comparison. How would you frame that?

Daniel ‌Marin

>> Yes. So, there's two verticals there. You can think about it as like the Nvidia of verifiable computing. So, not computing, but verifiable computing. And we actually shipped machines. We've shipped four machines, we call them zkVM. So we've shipped zkVM0, zkVM1, zkVM2. Last week we shipped zkVM3. These are verifiable machines like verifiable chips, they get better and faster on each iteration. Our latest machine was about a thousand times faster than our version two, which was about 100X faster than our version one, which is about 10X faster than our version zero, which we represented at Stanford about a year and a half ago. So these machines actually get faster and faster and faster at generating these proofs. Now, the other set of comparison with what you just described is, we happen to be building a supercomputer. If you go to our website, we talk about the supercomputer here and there, this is very serious. We're attempting to build the world's largest distributed supercomputer to ever exist. And how is this related to these chips that we put in the market? These are virtual chips, by the way, not physical chips. So, we have machines who work together as one to compute as a single machine. In fact, we were inspired by Ethereum back in the day. World computers. We can talk about it.

>> Let's go through the use case. Layer one supercomputer, I love that concept.

Daniel ‌Marin

>> Thank you. Appreciate it.

>> I love your website. Planetary scale. Mars is coming. We'll get to that in a second. I mean, talk about the use cases. Who's the customer? Where's the needle moving moment? What's the vision? Is it faster blockchain? What other markets do you see? Because you have the supercomputing, you're going to have the ability to power stuff.

Daniel ‌Marin

>> Exactly.

>> Is it efficient with power and cooling? What's the power envelope look like? These are all questions that are jumping in my head. I'm kind of going off the rails here a little bit. .

Daniel ‌Marin

>> Totally, totally. So, it's actually pretty simple. Our team invented this new thing, these chips, this serial knowledge virtual machine, CKBMs, that allow you to prove whether something's true or false and use this insane mathematics. And so with this, you can do many things including asking from the ground up, how to build a blockchain. And it happens that with these things, we can build a blockchain by utilizing idle compute around the world. So for example, the computer that you have right there, or your phone, they have compute that is not being utilized. And we can tap into that compute and pay you for that idle compute so that your computer can power our blockchain system. So, we have built a blockchain and are building a blockchain that gets more and more powerful with the more and more compute that is connected.

>> So, exponential scale.

Daniel ‌Marin

>> .

>> You got kind of a grid... I mean back, and not to date myself, but grid computing was a concept.

Daniel ‌Marin

>> Yeah. It's related to that,

>> back in the day, the more you put on the network, the more powerful it becomes, with all the benefits of like the internet.

Daniel ‌Marin

>> Exactly. The huge network effects. So in our case, this is what we call like the first blockchain supercomputer out there. And it's possible because there exists like this brand new technology, which is the CKBM that requires a lot of math and cryptographers. But it can scale exponentially through two verticals, vertical scaling and horizontal scaling. So, vertical scaling is the chips and the technology getting faster and faster and faster, which are series of CKBM .

>> By the way, AI loves vertical scaling because-

Daniel ‌Marin

>> Yeah, GPUs-...

>> domain expertise....

Daniel ‌Marin

>> for example.

>> Yeah, GPUs.

Daniel ‌Marin

>> Oh, that too.

>> Well, it's just everything's in vertical, but then the horizontal.

Daniel ‌Marin

>> It's just more compute.

>> Yeah, yeah.

Daniel ‌Marin

>> Right? And so, the product of the two is how powerful the system is. And so, our blockchain is designed to capture the exponential effects of Moore's Law. But it's a new form of Moore's Law, a new form that only applies to verifiable computing.

>> Okay. I got to ask because, were you guys sitting around saying, "Hmm."? Or where was the moment that this kind of clicked for you? Obviously you got the research, you went to Stanford. You got the brain trust, you got the network of folks working on a lot of cool stuff, the density here in the Bay Area, super powerful. What was the motivation? Was it like a eureka moment? Where were you? When did it come together? When you said, "Hey, we're doing this."? Or was it an evolution to a flash point?

Daniel ‌Marin

>> Good point, good point. I would actually say it was a little bit of a eureka moment, but it's sort of built to it. I was part of the Stanford Blockchain Club, I was a student of cryptography at Stanford. Before that, I studied theoretical physics and mathematics. And so at one point, I was working on my senior project at Stanford, and a new paper came out from Microsoft Research actually, that showed how to combine serial knowledge proofs together. So you could grab a lot of proofs, compute it for some programs, you could all combine them into one. If you extrapolate that idea, you can realize that you can have, for the first time in history, a lot of computers work together, and combine their computer output into a single output. And so the goal of Nexus from day one, has been to build this supercomputer system by combining all of this computing power into one. And that's why we sometimes call it the Nexus Project. So even though we're a Delaware C Corp, we're a startup, et cetera, et cetera, our goal is to build this gigantic computer system that our hope is outlasts even our lives.

>> Yeah, I mean, you're at the Nexus, pun intended. And you mentioned AI. I want to bring that in because you mentioned horizontal and vertical scaling, which from a hardware standpoint, layer one and physical layer and gear, software's got to run on it, right? So you need software to run on hardware.

Daniel ‌Marin

>> .

>> Not to go old school, you got hardware, middleware, apps. Okay, that's the old school kind of breakdown. If you're going to nail down this layer one supercomputing, okay, love the vision, love the execution there. Now, what happens next? Because now if you intersect AI, some of the things we're talking about AI, you mentioned a little bit earlier, you now have a runtime and generative process. So, the runtime assembly is not a new concept in distributed computing.

Daniel ‌Marin

>> Totally.

>> I mean, it's been around for-

Daniel ‌Marin

>> It's been around for a long time....

>> operating system theory, link load, everything happened, compile. But now with generative AI, the category, it's generating new things. It's not pre-programmed. So this idea of the proofs coming together makes me think there's got to be an intersection, the nexus between some of the AI advancements.

Daniel ‌Marin

>> A hundred percent.

>> Can you share your vision on that?

Daniel ‌Marin

>> Yes. The vision, there's so many intersections that we're going to spend a lot of time talking about. But something that I like to say, I'm a technical founder. We as a company, are a very practical company. The technology is just an element for what we think is a huge business opportunity that's coming in the next years and decades. The layer one blockchain that we've built is only, as you said, the runtime for what we hope will be many world scale applications and businesses getting a lot of distribution power. And so answering your question about AI, we are in San Francisco, the epicenter of AI. We're literally like a layer one blockchain company surrounded by a bunch of AI companies.

>> You have a lot of friends.

Daniel ‌Marin

>> We do have a lot of friends.

>> Super computing, hey.

Daniel ‌Marin

>> Exactly. Exactly. On GPUs and high performance computing.

>> First drink is free.

Daniel ‌Marin

>> Something like that, right? Something like that. And so, something that I was just chatting with someone yesterday, is like, we expect that 99% of users on the Nexus blockchain will be Ais, actually. Just because 99% of software and backend system is being written by systems such as cursor and Copilot. And then you can have agents trade with each other, so on and so forth. And so, answering your question once again, more concretely, AI forces advancements in compute performance that directly benefit the Nexus layer one blockchain because it was designed for that. AI allows us to do way much more trading, let's say, on the blockchain than before. AI allows us to deploy applications way faster for blockchains than ever before. And for a team like us, which is a lot of high density in terms of talent, when we give access to AI tools to our teammates, we're able to develop applications at 10X the speed that we were before. And so taking a little bit of a macro perspective, AI nowadays is pushing the world to move so fast at an exponential rate, that if you are a startup that is capturing a market, you need to be thinking about, how do I keep up with the velocity at which things are moving? And how do I utilize this exponential momentum to build a company that scales, well, to the scale .

>> You know, one of the things I love, I've been in California, Palo Alto for 25 years and studying entrepreneurship by living it and hanging out here is, that it's serendipity, the propensity for collaboration. And when you have collisions of people connecting, whether they're roommates or whatever, just sharing is interesting.

Daniel ‌Marin

>> A hundred percent.

>> And so now we're at a point where, okay, you have two worlds colliding in a positive way where there's a lot of propensity. So, open source software where the AI friends, your AI friends are hanging out, new models are leapfrogging each other every day. But also how to use off the shelf stuff to make it go better. So you're seeing a whole top of the stack, if you will, middle of the stack, get innovative. You've got the supercomputing vision here, scaling from the bottom. As that comes together, I'm expecting a lot of the problems to be solved.

Daniel ‌Marin

>> Like a clash.

>> So, here's what we know. We love AI. We know there's problems. Power, cooling, energy, bounded by energy. I mean, AI's bounded by energy at some level.

Daniel ‌Marin

>> It is, ultimately.

>> Data centers are sold out. I mean, CapEx, they're looking at nuclear reactors. So in every wave I've seen, the so-called standard like ethernet, that's dead. I mean, it dies every year, but it never dies. And now it's the center of the fabric in chips. The role of standards are key, things evolve and problems get solved. What do you see as problems that solve, if you take your vision forward? Because you have efficiencies.

Daniel ‌Marin

>> Certainly.

>> That's a potential power opportunity to solve that problem, one. AI wants more horsepower-

Daniel ‌Marin

>> .

>> They want more context. You see what DeepSeek does. I mean, it's very clever. I mean, it's just an illustration that smart people can figure stuff out. What do you see as stuff that will be figured out? If you kind of connect the dots a little bit, go out to the future and say, "What will be solved by evolution?" Interactions with people, creativity, entrepreneurship, collaboration? What are some of the things you see will easily be checked off the box?

Daniel ‌Marin

>> Well, ultimately, all of this technology, et cetera, it's all meant to enhance human superpowers. AI enhances human intelligence and productivity. Blockchains enhance financial superpowers of businesses. And so to give you an example, and I think a lot of blockchain people don't talk about this too often, is if you are a business and you're utilizing a blockchain, you get access potentially to millions of users. You get a lot of distribution power, you gain the ability to raise capital from many parts of the world. You gain the access to millions of users to pay you instantly from everywhere. And so if you're a developer and you know how to utilize AI and blockchain, you gain both technical superpowers and financial superpowers that allow you to build a business with very little resources. And it's pretty insane.

>> I mean, you have a backend innovation theater going on, front-end innovation theater. I mean, SaaS was just front-end in my mind. That was just web apps on cloud. Cloud was a backend innovation, but there's no front-end innovation. It was just SaaS on the cloud. Both created massive change. Now, you have two theaters of innovation.

Daniel ‌Marin

>> That are happening simultaneously.

>> At the same time.

Daniel ‌Marin

>> Exactly.

>> I've never seen that.

Daniel ‌Marin

>> It's crazy times, it's crazy times. What I tell you when people ask me about, "Hey, you know what's coming?", is like the human mind cannot comprehend exponential effects. If you're a company leader, you need to understand that exponential effects are happening, whether you want it or not. So, how do you best position your team to be able to capitalize on those rising technologies?

>> Daniel, I really appreciate the conversation. Great innovation.

Daniel ‌Marin

>> Thank you.

>> Congratulations. Looking forward to seeing how it goes. Put a plug in for Nexus. What are you guys looking to do? What your goals are? Share some stats. Just take it away.

Daniel ‌Marin

>> Totally. A hundred percent. Well, since November, over the last something like eight months, we've launched three global testnets. Our first testnet had about a hundred thousand users, then we had about 500,000 users. We just launched a testnet, our third testnet about two weeks ago. It had about 2 million users. And out of that, there were like 3.5 million nodes running our software in the network. And so, what's next for us is probably one more testnet release, and then likely, mainnet. Still to be determined. But we are a team that is very focused on shipping constantly and very fast.

>> Well, congratulations. I love your work, I love what you're doing.

Daniel ‌Marin

>> Appreciate it.

>> How do people get involved? How do they engage with you guys? Give the coordinates. It's all yours.

Daniel ‌Marin

>> So the magic of Nexus is that you can join quite literally in just one click, even from your browser. If you go to nexus.xyz and you click, connect, you already start supplying compute to the network. And you join the .

>> I got like three Macs sitting around and a Dell, and I'm going to just click it up and join the supercomputer-

Daniel ‌Marin

>> You certainly should....

>> hive mind of infrastructure layer one. I guess it's kind of hive mind of layer one.

Daniel ‌Marin

>> It is. It is like a hive mind.

>> Daniel, thanks for coming on. Great conversation.

Daniel ‌Marin

>> Appreciate it.

>> Of course. The Crypto Trailblazers blazing the trail, plowing the fields, creating innovation. Again, not for the faint of heart, but as the breakthroughs come, mainstream continues to adopt. You're going to see a lot more distributed computing architectures, blockchain, new kinds of applications, all coming to the table, all from the trailblazers here on theCUBE and beyond. I'm John Furrier, host of theCUBE. Thanks for watching.

Daniel Marin, Nexus

search

>> Welcome back, everyone, to theCUBE here at Palo Alto, California. I'm John Furrier, your host of theCUBE. This is our Crypto Trailblazers two-day event. Everyone's in town from the Theorem Foundation in Singapore, or activating Stanford and Berkeley Cal. All the entrepreneurs are out there. All those other thought leaders coming through, sharing their vision and what they're building to usher in this next generation. Daniel Marin's here, Founder of Nexus. Welcome to theCUBE. Thanks for coming on. We're having a good chat and we should have just turned the camera on before we got on, but thanks for coming on.

Daniel ‌Marin

>> Thank you, John. Yeah, it's great to be here, and it's great to be here at theCUBE. As I was saying, we love cubes at Nexus.

>> Matt Pats was just on. You guys were roommates, post Stanford. Big community here in the Bay Area. Obviously, I mean-

Daniel ‌Marin

>> High density.

>> Yeah, very high density of intellectual power. But it's just not just tech, there's this pragmatic thoughts around how to make money in a new way, the impact around technology. So again, the movements are happening. Ethereum's been a movement for a long time-

Daniel ‌Marin

>> It has....

>> and stories power movements. And we're so glad you're here to tell your story and also do the commentary and contribute to the community. Tell us about Nexus first, what you guys do. We'll jump in, set the concept. Take a minute to explain what you're working on.

Daniel ‌Marin

>> A hundred percent. It's very simple. We have layer one blockchain and we advance serial knowledge proofs. So, we work in the field of verifiable computing. So, we have a team of a lot of mathematicians and scientists that work day and night in accelerating this new field of verifiable computing. So, Nexus actually started this research project that I was working on at Stanford. And two years later, we're a team of like 35 people here in the Bay Area. And it's pretty simple. We actually have a lot of machines that we've put out in the industry. You can think of them as chips, but these chips compute serial knowledge proofs. And serial knowledge proofs are utilized to scale blockchain architectures and bring what we call verifiable computation to life. It's an abstract concept. It's not easy to grab, but verifiable computing is something that allows you to prove whether something is true or false. And that has a lot of applications on many markets. What we say is that we're like an open AI of verifiable computing, as opposed to intelligent computing.

>> So, talk about the relevance of that. Because in the math, which is blockchains, a lot of math involved, the importance of the proving, you hear proof of stake, everything is about proof.

Daniel ‌Marin

>> That's right.

>> this is the scale piece, because value store, okay, check. We understand that concept. But this idea of scale and value proofing becomes a huge thing. What's the driver of this? What was the problem to solve? Was it scale? Was it a mechanism? Was it software? What was the main driver?

Daniel ‌Marin

>> Yes, yes. So Nexus, we as a company, work at a bigger space than blockchain itself. Verifiable computing is a new form of computation, and that's something very weird to say. Something like AI or AGI is a new form of computing. Verifiable computing is a new form of computing in and of itself. And just as how AI models like 03, 01, et cetera can be utilized in many markets, verifiable computing can be applied to many markets as well. So to give you a few examples, human identity. You can prove whether something is true or false about who you are. Private computing, being able to do computations without revealing your private data. High security computing. So if you're part of the defense industry, to be able to show that a computation was done correctly without mistakes. And blockchains are a subset of verifiable computing. They're distributed verifiable machines. And it's well known that all of the scaling propositions for long term for blockchains are related to serial knowledge proofs and verifiable computing. And so for us, blockchains is a huge industry financially speaking, that we target.

>> So it's a segment for you.

Daniel ‌Marin

>> It is a segment .

>> So you brought up AI, so I want to kind of maybe make a comparison and get your reaction to. If you look at Nvidia, Nvidia was doing gaming, they were a graphics card. Now they're bigger than Intel. Look at the stock price.

Daniel ‌Marin

>> Yeah. It's crazy.

>> They started as a graphics card, all the gamers used it, and then became graphics systems for gaming. Crypto, great part of mining. Nvidia, thank you, Nvidia. Thank you for the blockchain enablement. Then they go to AI and now it's a super computing revolution.

Daniel ‌Marin

>> Yes.

>> Okay? So, Jensen Wong talks about accelerated computing all the time. With that kind of paradigm, compare what you guys are doing. Because I see a little bit of comparison here-

Daniel ‌Marin

>> A hundred percent, a hundred percent....

>> with what you're doing. You're accelerating, you're verifying, which is actually, I'm just riffing on this, but I can see how what... Because he talks about accelerated computing as a paradigm. And it's a new thing, it's a category.

Daniel ‌Marin

>> It's a new category that they're building.

>> It's not better compute. And now-

Daniel ‌Marin

>> It's a different type....

>> you've got AI factories. So compare, how would you tell about what Nexus does if it was a categorical shift as equivalent verifiable computing... Obviously, see there's an acceleration in there, but not in the sense of GPU. Talk about the comparison. How would you frame that?

Daniel ‌Marin

>> Yes. So, there's two verticals there. You can think about it as like the Nvidia of verifiable computing. So, not computing, but verifiable computing. And we actually shipped machines. We've shipped four machines, we call them zkVM. So we've shipped zkVM0, zkVM1, zkVM2. Last week we shipped zkVM3. These are verifiable machines like verifiable chips, they get better and faster on each iteration. Our latest machine was about a thousand times faster than our version two, which was about 100X faster than our version one, which is about 10X faster than our version zero, which we represented at Stanford about a year and a half ago. So these machines actually get faster and faster and faster at generating these proofs. Now, the other set of comparison with what you just described is, we happen to be building a supercomputer. If you go to our website, we talk about the supercomputer here and there, this is very serious. We're attempting to build the world's largest distributed supercomputer to ever exist. And how is this related to these chips that we put in the market? These are virtual chips, by the way, not physical chips. So, we have machines who work together as one to compute as a single machine. In fact, we were inspired by Ethereum back in the day. World computers. We can talk about it.

>> Let's go through the use case. Layer one supercomputer, I love that concept.

Daniel ‌Marin

>> Thank you. Appreciate it.

>> I love your website. Planetary scale. Mars is coming. We'll get to that in a second. I mean, talk about the use cases. Who's the customer? Where's the needle moving moment? What's the vision? Is it faster blockchain? What other markets do you see? Because you have the supercomputing, you're going to have the ability to power stuff.

Daniel ‌Marin

>> Exactly.

>> Is it efficient with power and cooling? What's the power envelope look like? These are all questions that are jumping in my head. I'm kind of going off the rails here a little bit. .

Daniel ‌Marin

>> Totally, totally. So, it's actually pretty simple. Our team invented this new thing, these chips, this serial knowledge virtual machine, CKBMs, that allow you to prove whether something's true or false and use this insane mathematics. And so with this, you can do many things including asking from the ground up, how to build a blockchain. And it happens that with these things, we can build a blockchain by utilizing idle compute around the world. So for example, the computer that you have right there, or your phone, they have compute that is not being utilized. And we can tap into that compute and pay you for that idle compute so that your computer can power our blockchain system. So, we have built a blockchain and are building a blockchain that gets more and more powerful with the more and more compute that is connected.

>> So, exponential scale.

Daniel ‌Marin

>> .

>> You got kind of a grid... I mean back, and not to date myself, but grid computing was a concept.

Daniel ‌Marin

>> Yeah. It's related to that,

>> back in the day, the more you put on the network, the more powerful it becomes, with all the benefits of like the internet.

Daniel ‌Marin

>> Exactly. The huge network effects. So in our case, this is what we call like the first blockchain supercomputer out there. And it's possible because there exists like this brand new technology, which is the CKBM that requires a lot of math and cryptographers. But it can scale exponentially through two verticals, vertical scaling and horizontal scaling. So, vertical scaling is the chips and the technology getting faster and faster and faster, which are series of CKBM .

>> By the way, AI loves vertical scaling because-

Daniel ‌Marin

>> Yeah, GPUs-...

>> domain expertise....

Daniel ‌Marin

>> for example.

>> Yeah, GPUs.

Daniel ‌Marin

>> Oh, that too.

>> Well, it's just everything's in vertical, but then the horizontal.

Daniel ‌Marin

>> It's just more compute.

>> Yeah, yeah.

Daniel ‌Marin

>> Right? And so, the product of the two is how powerful the system is. And so, our blockchain is designed to capture the exponential effects of Moore's Law. But it's a new form of Moore's Law, a new form that only applies to verifiable computing.

>> Okay. I got to ask because, were you guys sitting around saying, "Hmm."? Or where was the moment that this kind of clicked for you? Obviously you got the research, you went to Stanford. You got the brain trust, you got the network of folks working on a lot of cool stuff, the density here in the Bay Area, super powerful. What was the motivation? Was it like a eureka moment? Where were you? When did it come together? When you said, "Hey, we're doing this."? Or was it an evolution to a flash point?

Daniel ‌Marin

>> Good point, good point. I would actually say it was a little bit of a eureka moment, but it's sort of built to it. I was part of the Stanford Blockchain Club, I was a student of cryptography at Stanford. Before that, I studied theoretical physics and mathematics. And so at one point, I was working on my senior project at Stanford, and a new paper came out from Microsoft Research actually, that showed how to combine serial knowledge proofs together. So you could grab a lot of proofs, compute it for some programs, you could all combine them into one. If you extrapolate that idea, you can realize that you can have, for the first time in history, a lot of computers work together, and combine their computer output into a single output. And so the goal of Nexus from day one, has been to build this supercomputer system by combining all of this computing power into one. And that's why we sometimes call it the Nexus Project. So even though we're a Delaware C Corp, we're a startup, et cetera, et cetera, our goal is to build this gigantic computer system that our hope is outlasts even our lives.

>> Yeah, I mean, you're at the Nexus, pun intended. And you mentioned AI. I want to bring that in because you mentioned horizontal and vertical scaling, which from a hardware standpoint, layer one and physical layer and gear, software's got to run on it, right? So you need software to run on hardware.

Daniel ‌Marin

>> .

>> Not to go old school, you got hardware, middleware, apps. Okay, that's the old school kind of breakdown. If you're going to nail down this layer one supercomputing, okay, love the vision, love the execution there. Now, what happens next? Because now if you intersect AI, some of the things we're talking about AI, you mentioned a little bit earlier, you now have a runtime and generative process. So, the runtime assembly is not a new concept in distributed computing.

Daniel ‌Marin

>> Totally.

>> I mean, it's been around for-

Daniel ‌Marin

>> It's been around for a long time....

>> operating system theory, link load, everything happened, compile. But now with generative AI, the category, it's generating new things. It's not pre-programmed. So this idea of the proofs coming together makes me think there's got to be an intersection, the nexus between some of the AI advancements.

Daniel ‌Marin

>> A hundred percent.

>> Can you share your vision on that?

Daniel ‌Marin

>> Yes. The vision, there's so many intersections that we're going to spend a lot of time talking about. But something that I like to say, I'm a technical founder. We as a company, are a very practical company. The technology is just an element for what we think is a huge business opportunity that's coming in the next years and decades. The layer one blockchain that we've built is only, as you said, the runtime for what we hope will be many world scale applications and businesses getting a lot of distribution power. And so answering your question about AI, we are in San Francisco, the epicenter of AI. We're literally like a layer one blockchain company surrounded by a bunch of AI companies.

>> You have a lot of friends.

Daniel ‌Marin

>> We do have a lot of friends.

>> Super computing, hey.

Daniel ‌Marin

>> Exactly. Exactly. On GPUs and high performance computing.

>> First drink is free.

Daniel ‌Marin

>> Something like that, right? Something like that. And so, something that I was just chatting with someone yesterday, is like, we expect that 99% of users on the Nexus blockchain will be Ais, actually. Just because 99% of software and backend system is being written by systems such as cursor and Copilot. And then you can have agents trade with each other, so on and so forth. And so, answering your question once again, more concretely, AI forces advancements in compute performance that directly benefit the Nexus layer one blockchain because it was designed for that. AI allows us to do way much more trading, let's say, on the blockchain than before. AI allows us to deploy applications way faster for blockchains than ever before. And for a team like us, which is a lot of high density in terms of talent, when we give access to AI tools to our teammates, we're able to develop applications at 10X the speed that we were before. And so taking a little bit of a macro perspective, AI nowadays is pushing the world to move so fast at an exponential rate, that if you are a startup that is capturing a market, you need to be thinking about, how do I keep up with the velocity at which things are moving? And how do I utilize this exponential momentum to build a company that scales, well, to the scale .

>> You know, one of the things I love, I've been in California, Palo Alto for 25 years and studying entrepreneurship by living it and hanging out here is, that it's serendipity, the propensity for collaboration. And when you have collisions of people connecting, whether they're roommates or whatever, just sharing is interesting.

Daniel ‌Marin

>> A hundred percent.

>> And so now we're at a point where, okay, you have two worlds colliding in a positive way where there's a lot of propensity. So, open source software where the AI friends, your AI friends are hanging out, new models are leapfrogging each other every day. But also how to use off the shelf stuff to make it go better. So you're seeing a whole top of the stack, if you will, middle of the stack, get innovative. You've got the supercomputing vision here, scaling from the bottom. As that comes together, I'm expecting a lot of the problems to be solved.

Daniel ‌Marin

>> Like a clash.

>> So, here's what we know. We love AI. We know there's problems. Power, cooling, energy, bounded by energy. I mean, AI's bounded by energy at some level.

Daniel ‌Marin

>> It is, ultimately.

>> Data centers are sold out. I mean, CapEx, they're looking at nuclear reactors. So in every wave I've seen, the so-called standard like ethernet, that's dead. I mean, it dies every year, but it never dies. And now it's the center of the fabric in chips. The role of standards are key, things evolve and problems get solved. What do you see as problems that solve, if you take your vision forward? Because you have efficiencies.

Daniel ‌Marin

>> Certainly.

>> That's a potential power opportunity to solve that problem, one. AI wants more horsepower-

Daniel ‌Marin

>> .

>> They want more context. You see what DeepSeek does. I mean, it's very clever. I mean, it's just an illustration that smart people can figure stuff out. What do you see as stuff that will be figured out? If you kind of connect the dots a little bit, go out to the future and say, "What will be solved by evolution?" Interactions with people, creativity, entrepreneurship, collaboration? What are some of the things you see will easily be checked off the box?

Daniel ‌Marin

>> Well, ultimately, all of this technology, et cetera, it's all meant to enhance human superpowers. AI enhances human intelligence and productivity. Blockchains enhance financial superpowers of businesses. And so to give you an example, and I think a lot of blockchain people don't talk about this too often, is if you are a business and you're utilizing a blockchain, you get access potentially to millions of users. You get a lot of distribution power, you gain the ability to raise capital from many parts of the world. You gain the access to millions of users to pay you instantly from everywhere. And so if you're a developer and you know how to utilize AI and blockchain, you gain both technical superpowers and financial superpowers that allow you to build a business with very little resources. And it's pretty insane.

>> I mean, you have a backend innovation theater going on, front-end innovation theater. I mean, SaaS was just front-end in my mind. That was just web apps on cloud. Cloud was a backend innovation, but there's no front-end innovation. It was just SaaS on the cloud. Both created massive change. Now, you have two theaters of innovation.

Daniel ‌Marin

>> That are happening simultaneously.

>> At the same time.

Daniel ‌Marin

>> Exactly.

>> I've never seen that.

Daniel ‌Marin

>> It's crazy times, it's crazy times. What I tell you when people ask me about, "Hey, you know what's coming?", is like the human mind cannot comprehend exponential effects. If you're a company leader, you need to understand that exponential effects are happening, whether you want it or not. So, how do you best position your team to be able to capitalize on those rising technologies?

>> Daniel, I really appreciate the conversation. Great innovation.

Daniel ‌Marin

>> Thank you.

>> Congratulations. Looking forward to seeing how it goes. Put a plug in for Nexus. What are you guys looking to do? What your goals are? Share some stats. Just take it away.

Daniel ‌Marin

>> Totally. A hundred percent. Well, since November, over the last something like eight months, we've launched three global testnets. Our first testnet had about a hundred thousand users, then we had about 500,000 users. We just launched a testnet, our third testnet about two weeks ago. It had about 2 million users. And out of that, there were like 3.5 million nodes running our software in the network. And so, what's next for us is probably one more testnet release, and then likely, mainnet. Still to be determined. But we are a team that is very focused on shipping constantly and very fast.

>> Well, congratulations. I love your work, I love what you're doing.

Daniel ‌Marin

>> Appreciate it.

>> How do people get involved? How do they engage with you guys? Give the coordinates. It's all yours.

Daniel ‌Marin

>> So the magic of Nexus is that you can join quite literally in just one click, even from your browser. If you go to nexus.xyz and you click, connect, you already start supplying compute to the network. And you join the .

>> I got like three Macs sitting around and a Dell, and I'm going to just click it up and join the supercomputer-

Daniel ‌Marin

>> You certainly should....

>> hive mind of infrastructure layer one. I guess it's kind of hive mind of layer one.

Daniel ‌Marin

>> It is. It is like a hive mind.

>> Daniel, thanks for coming on. Great conversation.

Daniel ‌Marin

>> Appreciate it.

>> Of course. The Crypto Trailblazers blazing the trail, plowing the fields, creating innovation. Again, not for the faint of heart, but as the breakthroughs come, mainstream continues to adopt. You're going to see a lot more distributed computing architectures, blockchain, new kinds of applications, all coming to the table, all from the trailblazers here on theCUBE and beyond. I'm John Furrier, host of theCUBE. Thanks for watching.