Block by Block: A Show on Web3 Growth Marketing

[PODCAST] Nirav Murthy on the Future of AI and IP Licensing on Camp Network

Peter Abilla

Summary

Nirav Murthy, co-founder of Camp Network, talks through how Camp is building a user-friendly IP licensing platform that blends AI and blockchain. The project shifted from a Layer 2 to a Layer 1 so they could control the tech stack, tailor the network to IP use cases, and support developers building on top of it. As AI becomes the default way new IP is created, provenance and licensing move from “nice to have” to “mandatory.” Camp is positioning itself as the infrastructure layer for this shift by making licensing programmable, cheap, and instant. Nirav also walks through what creator adoption looks like in practice, how Camp keeps users engaged, and why community plus product feedback loops matter more than hype in this category.

Takeaways

  • Verticalized blockchains are emerging because general-purpose chains fall short for specialized use cases like IP
  • Nearly all future IP will originate from AI tools, which makes on-chain provenance essential
  • The current IP licensing market is slow, expensive, and controlled by intermediaries
  • Smart-contract licensing unlocks new monetization models for creators
  • Retention is a better signal than raw user acquisition in crypto
  • Gamified product mechanics help drive repeat participation
  • Feedback loops with early users guide product direction
  • The entertainment sector is most likely to be disrupted first
  • Community creates durable network effects around IP ownership

Chapters

(00:00) The Genesis of Camp Network

(03:03) Moving from Layer 2 to Layer 1

(05:58) What Camp Actually Solves

(08:46) IP Licensing in an AI-Native World

(11:39) Power for Creators

(14:44) How Buyers Fit Into the Model

(17:22) Ecosystem Growth and Metrics

(20:24) Creator and Buyer Journeys

(23:14) Go-to-Market and Marketing

(26:10) Lessons from Building Camp

(29:09) Competitive Dynamics

(32:07) Industries Most Likely to Be Disrupted

Thank you for the support from Infinex

Follow me @shmula on X for upcoming episodes and to get in touch with me.

See other Episodes Here. And thank you to all our crypto and blockchain guests.

Welcome to Block by Block. I'm thrilled to have Nirav Murthy, co-founder of Camp Network with us today. We're diving into how Camp Network is redefining the landscape where intellectual property meets AI. Welcome, Nirav. Thanks Peter, great to be here and can't wait to have this conversation. Let's dive into the origin story. Could you kick us off with the founding story of Camp Network and what inspired you and your team to help to build a platform that integrates IP with AI? Yeah, it's actually a relatively long story to be honest with you because we went through many different iterations to arrive at what we ultimately are today at Campon. you know, we actually started this business in Q1 2023 when we were actually building a form of call it Provable Fandom on chain in the first place. You can think about it as Patreon built for the creator economy on chain. And what we found is we were really working with a lot of these different partners, know, folks like Sonon Ethereum, folks like Audius on Solana. And what we found is there are so many different teams we were working with and it sort of gave us the initial aha moment where it's like... Why don't we just have our own block space? And so with our good friends over uh at Optimism as well as Gelato, we spun up a layer two. And what we found is people really liked the concept of the block space, but unfortunately there was a lot of stuff missing. We couldn't really prioritize all things in the consumer crypto side of things. We wanted to handle royalties directly at the execution there. We wanted to customize the chain. And unfortunately none of that was really possible in sort of the RAS L2 world. And that's actually what let us decentralize the blockchain and become a layer one overall. And that's the current iteration that can't be today. Currently we're in test net. launched our test on May in May 2025, but it was a two year learning journey of, you know, first app to block space to needing that customizable level of block space that brought us here to what campus today. You know, there's been a lot of criticism against the Ethereum Foundation's strategy, the L2 strategy. Do you think uh we'll see more and more projects move from L2 to L1 based on the experience that you had? It's such a good question. So I'd almost make a distinction here. I think we'll see a lot more projects, whether it's applications or anything like that, move to having their own block space. I wouldn't necessarily make the distinction between L2 and L1. And I say that for a couple of reasons. Number one is running an R1 is extremely expensive. Nobody really considers that possibility. within all your tooling and whatnot, you can easily be spending north of $2 million a year just on tech contracts that you're really working on. I think you'll start to see all applications, as they get big enough scale, start to actually want their own block space. And this is one of the theories that we have as camp. One of the things that we do is we consider ourselves like the L1, the base camp level. But if any app gets big enough, obviously they want their own block space. And the reason for that is because you want your users experience as little latency as possible and the only way to get that is with dedicated block space. Otherwise it's just crowded. And so one of the things we've done is with our good friends over at Gelato, anytime one of our teams gets big enough, you know, we can basically have them spin up a side camp for them that ultimately settles back to us. Now the reason why teams like this in addition to the block space is also they get bigger value accrual on the fees, which is something that I think more apps will start to focus on as it gets bigger and bigger. ah Best example of this, UniSwap. They launched their own L2s simply for this exact reason alone. Gotcha. You on the cost side, you're right. mean, a lot of people don't recognize how expensive running an L1 can be. I led growth and business development at Harmony. And we had one application, DeFi Kingdoms, that absolutely destroyed the chain. So much money was being moved over the bridge, uh tons and tons of users. we were spending, it was kind of like a, it was a bittersweet. contrasting moment because we were so excited that so many users were bridging over and participating on DeFi Kingdoms. At the same time, we recognize that, man, with each user, the cost just skyrockets. um But a lot of people don't recognize that, and it's actually a big deal. It's a huge deal. mean, like, you look at a lot of these teams, like the L2 wars, we can call it, were probably like Q1, Q2 last year, I would call it. ah And then sort of some of them portrayed on and did their token launches and mainnet launches towards the end of the year and even earlier this year. But you look at some of those teams, you know, they don't have to raise that much money. You look at sort of why, because you go work with a Gelato, a Conduit, or Caldera, you're paying them effectively probably $50 to $100K a year to sort of maintain your chain. And then obviously when you work with one of those, whether it's like Arbitrums, you get SYNC or Optimism, in addition to getting those grants, you're basically just doing rev-sure off the top. So the fixed costs aren't actually that much, and really what you're really spending money on is growth fundamentally, because they take care of a lot of things like... your subgraph, your block explorer, all the things that nobody really cares, nobody really thinks about. When you do the L1, those all come from scratch at the end of the day. And so in order to actually do that, you're going to have to go raise north of 20 million bucks overall, just to sort of build that initial hype, build that treasury, build that foundation, technical foundation, not like literal foundation. But it's substantially harder to do and do. And so I think one of the things that we'll start to see over the next three to five years or so, Peter, is a big emphasis on spinning up block space in the easiest way possible. I think you're starting to see the L1 premium sort of die down in real time. And that's even sort of happening with some of the token listing agencies. And so the fact that now you can basically have college not the most differentiated block space as an L2 or as an L1.5 as a sovereign roll-up. You know the tooling has gotten so good but you can quite literally raise 20 to 40 percent of that amount required for an L1. It could be the sort of sign of the situation that more folks will go to what's convenient as opposed to what they think might trade higher in the market overall. Yeah, let's let's jump into camp network. How would you describe its core functionality and pretend like you're talking to someone who knows nothing about camp or IP or AI. ah its core basis, and I'm sure you know this from your growth days, everybody always tries to big brain what a blockchain really is. And at its core basis, blockchain is just a technical infrastructure layer. That's really all that it is. We believe that blockchains should start being more verticalized. What you saw in the last five years or so are these general purpose L1s that have done phenomenally well. Solana, Ethereum. probably two us examples. Bitcoin is a store of value fundamentally. But now what we expect to see is a lot more of these vertical blockchains. And what I mean by that is block space for a specific industry or for a specific purpose. What comes to mind over here, you think about like, Athena, is like a yield-based stablecoin and that's why it has its own thing. Uniswap as a DEX, right? On the RWA side, you have Ondo and you have Plume On the consumer side, you have a whole number of these, but abstract is probably the biggest one to date, and then dating back before that flow. And what we think is, you're going to start to see more of these things because there's candidly not that much of a demand for just general purpose block space that's for everything. And so at Camp, we're built for the intersection of IP and AI. Now, what does that really mean, right? So built into our chain, we have the proof of provenance protocol, which consists of origin and matrix. Origin is simply the easiest way for a user to onboard their IP. and plugs in as a developer SDK for the application layer itself. That can be anything from songs, uh images, videos, text, data. All of that is your IP and origin makes that the easiest way to actually register any of that on chain and set terms and conditions around that. Then the second part of that, what we built in is matrix. And this is our AI framework that makes it extremely easy to train and deploy agents on the blockchain in the first place. Couple things to unpack here. Our belief is that the future of the world is agent to agent interactions. What does that really mean? know, in about five years, I think instead of us doing this podcast, we would have an agent version of myself and an agent version of yourself basically talking to each other. But in order for that to exist, you need to have these agents trained on user-owned IP. And that's what we basically integrated Matrix into Camp at the execution layer, effectively, at the protocol level, because that way it plugs in directly from all things related to origin itself. The second thing that we did is we spent a lot of time talking with our teams in blockchain and one of the things that they thought is you know usually if you're a validator you're going to prioritize what is the highest MEV and 99 % of the time that's going to be a DeFi related transaction and so one of the things that we did is at the execution layer itself we put it such that a minimum number of IP and AI related transactions must be included per block. And the benefit to doing that is you have priority order tagging. So if you're building in the consumer space or anything related to IP and AI, which I broadly include under the umbrella of consumer crypto, we are the best DevEx for you because we basically make it such that you have the lowest latency for any of your end users overall. The core proposition behind Camp Network is simple. Come prove that you're the first creator of something. 99 % of IP in the future will be generated via AI. so provenance as a concept or establishing that you're the owner or the originator of something has never been more important than ever before. We're seeing an ongoing battle right now happen. You saw chat, GPT and Reddit, which was the first kind of these. Then the second more profile one was Stability AI and Getty Images. And now you're seeing Mid Journey and Hollywood. We expect this IP copyright versus AI model debate to continue for the next decade. And we believe that camp is the home of where all new IP will be created overall. Got it. A couple of things to unpack there. First of all, I totally agree with your thesis that blockchains are going to become more verticalized or the way I call it, it's like very domain specific. And in general purpose, blockchains are, you know, they become commoditized and really don't bring any differentiation. One thing that I think could be helpful is if you helped us kind of help us picture like a side by side of like what the current IP licensing looks like right now. Like what is that process? And then what does the process look like with Camp and in the benefit to the various stakeholders. Yeah, and one thing to unpack there before we sort of dive into that uh second part of the question, just what you said you agree, that's not just a blockchain specific thing, by the way. I to make that clear. Generally, you see things start off as general purpose and then start to get more verticalized or more specialized. Like AI is going through the exact same thing right now. You see the gen purpose models like GPT and DROPIC, even Grok to some degree. ah You're starting to see much more verticalized, whether it's industry specific, like the hardweeds of the world for legal. There's a ton of help here specifically for patient data, for drug discovery and all that type of stuff. And then you're also seeing modal specific. which is like voice data, video and whatnot. So fully agree with you that domain specific is a really interesting way to phrase it overall. In terms of what IP licensing looks like today, it is probably the most antiquated industry in the world. You could argue outside of some of the entertainment industries, but it's ridiculously expensive to do. ah It's extremely bespoke. You need to have an army of lawyers to do so. and you're dealing with so many different stakeholders that your minimum to do like a basic IP licensing contract is a few months overall. Then also you have to think about it. Let's say I'm using the example of Disney, right? If you're looking at Moana, something like that has different licensing terms for the music, for the video, for the image, and then also for different sub buyers of that, like soft toys or anything like that. So it's a super antiquated process in all that link where there's no clear automation. And it was built for these like behemoths like Disney and, you know, Fox where they can have an army of lawyers basically negotiate all these things. Unfortunately, that's the IP of the 20th century. The IP of the 21st century is built around users and UGC overall. We think of Web2 as actually being the original sort of perpetrator of the UGC movement, and now it's those very people whose IP is actually being trained on for all AI, and they're getting the short end of the stick because there's no compensation for it. Where we envision this coming in on campus all through the origin version of the origin component of the uh provenance protocol. What that means is a user uploads their text and images on it. Simply there, you can basically say, hey, if somebody wants to use this, it's free, but they have to give me 5 % of any earnings that they make that's created on chain in the first place. Or it can be, you know what, I don't want any royalties associated with it, but if they're going to use this song, me a dollar. It's all about the user and how they think about what the terms and conditions should really be. And because we do this all at the smart contract level, it's ridiculously easy to do. This is actually the best benefit of using smart contracts overall, because it's almost the if P then Q type situation. And you're restricting people who are not actually agreeing to those terms. Obviously in your head, I'm sure you're thinking, well, that doesn't really solve the problem, it? Because theoretically, similar to like a LineWire or Video10p3.net situation, know, somebody could basically right-click and if there's the bad actor problem, we unfortunately can't solve that. And again, on blockchain, we can't really solve that. NFTs, everybody's like, it's a one-of-one. I could right-click and steal somebody's Twitter profile picture and nobody would really know unless they were to dig on chain. Instead, what we're trying to solve is the UX problem, or what I call the Spotify conundrum. where for $9.99 you can basically go on Spotify and listen to every song ever streamed a month. And so the principle is on Camp, as long as you comply with the terms, you can basically find the world's greatest user-owned IP repository and verified IP repository. And that consists of anything like text, images, music, data, and pretty much everything over there, as long as you comply with the existing terms and conditions that a user sets on their piece of data. You know, that sounds really fascinating. um Previous to me jumping into crypto full time, I did a lot of work in kind of process improvement in healthcare, et cetera. And standardizing a process that's very, very bespoke, like IP licensing, is actually really, really hard. But is like a requirement for it to grow and especially for like... to automate it through smart contracts. It has to be standardized to some degree. And that helps, I think, everybody. Because all these bespoke kind of approaches is just, on the one hand, it's good because it uh meets kind of the edge case needs of the individual. But it is so hard to generalize over thousands of IPs. Yeah, mean, what? What's impact? You hit it on the head. I mean, look like One of the reasons why we're starting off with what I call the long tail of users or like users and creators, can even call it, is because those are the easiest folks to onboard in the first place. You think Disney is going to come to us tomorrow and put all of their contracts on chain for each of their Avengers terms, whether it's sync, as I mentioned, toys or anything like that. Absolutely not. We actually have a team that we work with called Remaster who works with some of these big enterprises. They recently did Minto, which is one of the larger IP holders in Japan actually, but they almost take this like white glove McKinsey style approach where it's like for every single client that we work with, we're going to build our own smart contract level for it, where we can sort of negotiate existing terms and whatnot. Naturally, they all build on camp because obviously the mission aligns and there's a ton of things that we can do there overall, especially around the co-creation of IP. We actually just announced a couple of weeks ago a pretty big campaign around that and sell like 125,000 pieces of unique IP created. where we really take this approach is let's get the long tail of users on board because they can actually prove what's possible. And then this slowly goes upstream. Everybody always wants to try and get the T1 clients to start off, but the T1 Codifying is never going to be the first movers in any space because they know that they have some more negotiating power and as they see some of these better examples coming downstream, become substantially, substantially easier to do it. And that's sort of one of the things that we're really seeing. Codifying or standardizing what IP licensing looks at its most basic forms is great for individuals and creators. It gets a lot harder when you have like 500 different revenue streams that you need to account for. And that's sort why we haven't seen some these big guys come on to you get in the first place. You know, a lot of folks still think that when it comes to IP licensing, it means you have to hire a lawyer. Is that still true with camp or no? The goal for it is that it, so let me rephrase this. A lot of people have put IP on chain on camp and they have not used a lawyer. So I think that's like the most direct way to answer your question. Having said that, there's still a lot of work left to do here. We expect to see. a lot less of legal work involved in IP in the next three to five years, specifically for what I call like this prosumer segment, where it's like professional consumers or, you know, even like long tail creators. But right now you're still seeing a lot of folks use lawyers for pretty much everything. I think as you see agents get better and better though, that meta will slowly die down over time as well. You know, back in the day, like in a prior life, I was in a traditional finance person. was an investment banker and an investor. And we used to have our lawyers review NDAs and that was like a thousand bucks an hour or something like that. Now it's like Harvey cost 20 bucks a month or something like that and they've seen over a million MDAs so why would I need to go pay a thousand bucks to do so? So as we see some of the tooling get better and better we think that all of that metal will slowly die down as well Peter. Let's talk about the market. eh know, as I think about Camp, I'm imagining a two-sided market. You've got creators on the one side, and then on the other side, you've got the demand side of organizations or people that want to use the IP that's on Camp. I don't know what the right phrasing is. On Camp or maybe under buyers. Yeah, tell us about the market. Yeah, it's interesting. you mentioned creators and something that's really important and nuanced to mention here is creators are not just people who are like the creative creators. It's not just artists or photographers or video creators or game developers. Anybody can be a creator in the future. Let's say I create an agent to book restaurants in New York City where I live. Or let's say I create an agent to track the weather patterns in Salt Lake City when I'm skiing where you live. know, those are still my forms of IP that I've created and developed. And I should be paid accordingly when anybody actually uses those or pays for those services. So the big thing to realize here is anybody can be a creator. Creating a creator is no longer limited to just a musician or an artist overall. And that's something that we think is super important, especially as it relates to if I'm the creator of an agent, I should be comped for that, for the usage of that agent overall. On the second side of things, ah on the buyer side, the buyer demand is pretty unlimited. We spend a lot of time with the big Silicon Valley AI companies on pretty much every side of the model, whether it's general purpose, whether it's industry-specific or whether it's motor-specific. And people are kind of spooked about the big lawsuit and settlement that GPT had with Reddit. And so there's a massive demand of people basically looking to buy what I call a training data or a training IP in the most verified and legal way possible where there's no long-term legal ramifications of it. So what do mean by that? You see Silicon Valley is going through a very interesting shift. There's sort of three different armies in the AI world. There's the model layer, which is clearly commoditized by DeepSeek. It's very expensive and whatnot. And so, you know, there's a pretty big barrier to entry. Then you have compute. Nvidia obviously is like bigger than the entire market, cap of crypto, until this recent bull run. But, you know, decentralized compute is very difficult. And then the third, and where you can really distinguish the quality of your agents is the IP on which it's trained. And that's where we think we can really come into play. You what distinguishes one model from another? The training IP. And when you can do that on verified IP as opposed to having to do that on synthetic data, that makes it that much better of a model overall. And so that's why we think the buyer demand is absolutely massive. Can you share with us some metrics of uh how things are going so far? Yeah, so we're currently in test time right now. We have 6 million wallets created to date. We currently have 75 native apps, 300 apps overall. The thing that I really care about is the number of native apps on blockchain. I'm sure you know this, Peter, but like... Nobody cares for the 111 iteration of a DEX building on a chain that's built on 110 other chains to say the very least. so what I really care about is that sense. And then specifically in that native daft lens, how many of these people are going to be net new users to crypto? One of the things that I think is pretty cool about Camp, and I think this is a testament to the quality of our BD team, is a lot of our teams are what I call Web2 teams that are now using blockchain to basically streamline their business. So these are existing applications that are now using wallets to better track as opposed to relying on the octas, or the Okta is the aggregator, or like the Google sign-ons of the world where, you know, they control pretty much everything. Or they're also using sort of blockchain for microtransactions because it's significantly easier and cheaper. And that's some of the things we get really excited by. You know, we have a team called Runa that has two and a half million monthly active users. We have a team called Rewarded TV which has 400k monthly active users. These are real businesses that have revenue and are now using blockchain as a way of streamlining and making their businesses more efficient. And that's what I get really excited about to see at scale overall. Because as long as I've been in the space, we've been talking about how the application there needs to exist. And I think we're finally seeing that the DevEx is getting good enough. And some of the account obstruction partners and on-ramps and off-ramp aggregators are getting good enough where this can actually be equipped for what I call the broader masses. I love to hear that. You one question I ask often with founders of layer ones is what is the point at which you will stop talking about the layer one and instead be spending most of your time talking about the applications on your layer one? And I just witnessed that you just did that. You just shared about several applications, a couple that are doing super well on camp. And I think that's great. I mean, I'll put it really simply. ah I have one of the benefits of not having friends outside of crypto as well. And for the first time I've seen people finally get into a crypto app without realizing that it's crypto. And that's like Blackboard, particularly here in New York, because they've got a pretty tremendous amount of coverage on it from the restaurant side of things. But my belief is if in about 6 to 12 months, let's call it to 18 months just to be safe, if we're still talking about users onboarding through blockchains just for the sole purpose of yield farming or airdrop farming, we are screwed as an industry. In fact, realistically, we have to have people start onboarding through the application layer. Otherwise, the user comes on blockchain, what do they do? They probably put money in a vault, they probably deposit, maybe do a couple of deck swaps overall, just to hope that they get some sort of airdrop and then they leave. That's exactly why there's no stickiness on blockchain overall, right? People are really just yield farming. And so the application layer is essential and that's why we spend so much time there. know, nobody really cares that much anymore about, you know, whether my TPS is this much better than some of the else or the giga gas per block overall. It's more so, apps are in this ecosystem which makes me want to use it overall and think Solana has done a really good job of that overall like everyone knows Solana is a really fast blockchain that's all that matters and then on there they have a really good experience on things like pump on or on Jupiter or on any of the other apps overall and I think that's why Solana has such a sticky developer community and user community overall. Let's talk customer journeys. So if I'm a creator, like what does my customer journey look like interacting with Camp? And if I'm a buyer, what does that customer journey look like in searching for IP that I want to use? Yeah, it's such a good question. And we're sort of in the early days of this, but so our belief is that if you're a creator, it still should be through the application layer. So I'll give you a really quick example. We did this campaign for Scioka, which is like an IP, it's like a comic book. It's pretty cool if you haven't checked it out, and I'm happy to share it after in the comments, but we have a team called Merv. which is AI co-creation actually. And so one of the things people did is they basically said, okay, here's the existing IP, let's run a co-creation campaign around this. And so we did a ton of really cool stuff around it. Some people put like evil versions of the kind of characters. Some people did like action stories around it. Some people did love stories around it. And there were a whole new pieces of content created, 125K overall from like 60K unique users. We were thrilled with the campaign. But all of those creators fundamentally came in through either Merv or Storychain, which is like another AI remixing, AI native IP for remixing effectively. That's the way that all of these people came, and then they all put it on chain using origin and put their terms and conditions around it. That's how easy it is. If you have a good app that actually works, people are willing to use it. Shocking. know nobody else ever seems to consider those things on blockchain. From the buyer realm, it's something that we're thinking about pretty extensively right now. And where we think this trend is sort of similar to like an Airbnb style UX or interface where it's, you you can basically decide what type of data you're looking for, whether it's voice data, whether it's music, whether it's images, whether it's text, any of all these filters. And then you can basically have sort of things around it where it's verified. How many other people have used this? How many other people have viewed this? That's what we envision this coming into this uh user-owned IP repository, almost similar to an Airbnb. And then where we think that goes beyond that is called agent abstraction. so, again, similar to UX, almost like a marketplace for all these things, if I say I need to make a reservation at a bar in New York tonight or something like that, instead of having to know what agent directly to go find, I can basically query this in and then redirect it to the agent already. And given the way that AI is advancing, we think that piece will be shipped by Q2, Q3 next year as well. I understand you've been building since 2023, but I'm curious how you were able to grow brand awareness to a degree where applications have decided to build their business on top of camp. What was the strategy there and what have you seen work in attracting application developers to build on camp? Yeah, it's such a good question. honestly, there's a different way. Let me rephrase that. There's a different sort of mechanism at each step of the way. So let me explain what I mean by that. When we were an app, the whole premise was we were helping users sort of identify, or sorry, creators who their biggest like. fans were. And that was via a combination of permissioned APIs, like ZKTLS, then just questing platforms and whatnot. But what we would do is we would go to people and say, hey, this, this, and this are interesting. Does it work? And then we would basically just do almost a traditional B2B SaaS model where we're servicing the hell out of our customers and making sure that everything was actually working well and whatnot. When we built the L2, The cool thing was we're like, hey guys, we're going to give you all this data as long as you build on our L2 and you don't have to pay for any of it. And then everybody's like, cool, we'll integrate with you directly. And that's actually how we found our first like five to 10 teams. And then by working with those teams really closely, whether that was introducing them to like potential investors, whether that was introducing them to development partners, whether that was introducing them to potential customers. The word of mouth went pretty crazy. And so we started to have some of these people say like, hey, there's another team that we know. It would be really cool if you could chat with them. And so we built this initial cohort of what I call 10 to 25 developers. Now obviously, asking people to redeploy from an L2 to an L1 is never fun. It's definitely not something, especially when you're pushing back development timelines a year, that they're going to be shipping main nets substantially faster. But the nice part about all that really is we had this really core group of developers who became our biggest champions. And I think we, I'm obviously biased, but I think more than anybody else, do a really good job of working every day with our developers to make sure that product is good. That's not just on the product UX side of things, it's on feedback. It's on working, as I mentioned, investors, development partners, customers and all that. And so our belief is that BD kind of wins in blockchain, especially as I mentioned earlier, you know, the more that tech is getting commoditized, that what really matters is Some builders saying, wow, do I feel like I'm actually being prioritized? Do I actually feel like any of my concerns are being voiced and whatnot? And a lot of working with our teams on the L2 side of things is what led to us building the L1 because we heard what their biggest qualms are and what their biggest constraints were and we're like, why don't we just solve this overall? And so that's really how we built this and cultivated this really robust group of builders overall. Now on the native side of things, look, we don't have as many native builders as Solana. We don't have as many native builders as Basin. We don't have as many native builders as Ethereum. But tell us that it doesn't really matter. because what you usually see is in each of these ecosystems, there's really like only four to six apps that people actually use overall, and everything else is like trendy for one to three months at a time, but not really as that staying power. And so that's why we really care about the killer apps that we can see people actually using, and that's what we tend to try and make as happy as we can on the camp ecosystem. Here's something that I really like. I love that you guys went from an application to a layer two and then a layer one. And like the lessons that you learned and I think the key insights that led you to, know, we're an application, but we need to be a layer two. Layer two is not quite giving us kind of what we need or what our customers need. We need to be a layer one. I think that kind of journey is really under appreciated. And it's also really, really hard. And so kudos to you and the team for having gone through that and having followed your gut and intuition. now that you guys are your own layer one, you're meeting the customer needs that a layer one can only provide. Yeah, it's a really good point you brought up. like one of the things that... calls it annoys me about our industry is in web three, I'm sure you've noticed too, but like the narratives change faster than anything else. It's like by the time you actually define a narrative, it's too late already. And so like, let's look at this year, even last year, was like AI stable coins, DeFi, institutional flows back to stable coins. And it's like, if you're a builder, how are you not like, Oh my God, what is happening? I feel like my project will never make it. And so too often to not encrypt in particular, we look at some of these teams and it's like, they try to raise and they end up don't raise it. they do raise and they're like, wow, we don't have enough hype. And they'll see like one of the big funds puts a huge chunk of money into something. They're like, that's a new narrative. Let's everybody go do that. Like this was really popular last year with like the whole ABS and like restaking ecosystems and all this type of stuff where like there was so many different ones basically like eat LSTs and whatnot overall. And so the problem is people just jump narratives around faster and faster. And the result is you never go through cycles of development where you can actually figure out what works and what doesn't work. You contrast that to the web to development of a life cycle, right? It's like, Hey, I'm going to ship something. Okay. Then we go raise the pre-season round. Then it's like, let's go test this with some users. We do alpha feedback. Then we're like, okay, this is interesting. This is not bad. We take some of that feedback. We rejigger the thing. And then we go back and we do beta testing. And over that life cycle of five to seven years, you have an app that comes out like Snapchat V1. or something like that and because it's been tested so many times it works and people get excited about it. In crypto those cycles are so fucked up, they're so like messed up because effectively what you're seeing is like, hey let's raise, let's go do something, shit people got really excited about it, let's go hint at a token, let's launch the token, now everybody's made money from the airdrop and then they all leave and now it's like shit I don't have that much hype for it, can go out and iterate on those things again. And so I think what the result of this really is is it's important to listen to your customers and make sure you can do the feedback in terms, uh it is much better I believe the long-term profitability and long-term scalability than just trying to make a quick buck. Because anyone can go out and just launch a token instantly and just dump everything, but that doesn't have any value long-term as opposed to trying to build something that matters and is sustainable overall. I really like the journey that you guys have taken. Application to layer two to layer one. It shows that you guys are in it for the long term and you've developed some battle scars along the way. Like what are some key learnings that you've developed having gone through that journey? Oh man, I think I could do a TED Talk on this. Um... No. this is not something that's very common in crypto. you guys have actually like put in the work. no, we've noticed it's all like, it's funny. Our pre-seed investors invested in about two and a half years ago and they're like, wow, this is the longest path to liquidity we've ever seen, but it might actually be profitable because of the fact that like we've seen like at each stage of the junction, we've seen huge growth on every one of those things overall. But I think there's a few things that really matter. Number one is Build a team who's aligned for the same goals as you are as a business. If you have people who are in it for like the ponzi-nomics and quick pumps and dumps, trying to build on a long-term team, it's really bad for culture. And that's something that we've had to learn the hard way in terms of like folks joining the team and being dismissed or leaving the team and whatnot. That's the first thing I would say is make sure you have your company mission and that everybody's on the same page. Otherwise it leads to a lot of kerfuffles is probably the most polite way to phrase it. Number two is you can't rush building a good community, especially of core contributors. know, a lot of times I'm sure you see this, but like Kaito is the new craze right now. And you basically have people who are spending north of a million bucks a month on just like KOL promotions and whatnot. But none of those guys are going to care about you the minute that your token launches. In fact, the minute that it launches, they're going to collect their rewards or percentage of the network that you profit to top the appers, and then they're never going to talk to you again. What really matters is when we were building an app, being those first 30 people in Discord, who now kind of work for us because they're all effectively Discord moderators for different languages. So your core community becomes your biggest supporters and also they become your core contributors, which matters substantially along the way. And then I think number three is... Try and ignore the noise. Like a lot of times there are times when we're like, wow, we should just be doing this instead. Like when the AI boom was happening, we're like, should we just do decentralized compute? Should we just like sell data? We have all the data anyways. And it's like, well, that's going to be kind of gimmicky. People won't like it. It's cool for like a week or two, but beyond that, nobody really cares outside of that. And then I think the fourth is resilience is the most important thing in blockchain. No other industry goes through as many ups and downs as this. And if it's not for the thick skin, uh it's very difficult. And obviously, I believe that culture comes from the top. And so if the founders aren't resilient, it's going to set a really bad tempo for the team overall. I think that's good. And those are some seriously key learnings. Let's talk about the marketing funnel. I mentioned earlier about customer journey for both creators as well as the buyers. But there's going to be some set of creators that might be interested but not ready quite yet to participate in camp and the ecosystem. What does that funnel look like and how do you help them convert to become creators? Yeah, it's really good question. ah I think BD is the short answer to this, but there's a substantially longer way to phrase it. So a couple of things. Number one is if you're going to a bunch of creators and trying to pitch this pretty big tech unnerving of we're going to change the entire way that content is produced and released and managed and distributed and how you're going to get paid, they're going to be like, how does this work? So I think the big thing is being able to prove case studies is massive. Now, what do you need to do in order to prove a case study? You need an application that already works. ah Side note, when does this episode air? uh, probably in the next week or so. Okay, cool. So we have like a team called Core Protocol that we work with pretty extensively. They do a lot of really interesting things. They help Web2 IPs come into Web3 and find interesting use cases for them. That could be anything from like, if you own the token behind this Web2 IP, then you also get a portion of the royalties longer term. It also means you get early access to pieces of content. You get pre-releases. You're like a member of it. And then in addition, they have this AI native remixing app that works on Telegram directly. And you know, they work with like dead mouse, mouse trap and whatnot around the cap table, disclosures on the cap table. The amount of sales cycles that you can save by going to artists and say like, hey, we've worked with this person and this is the app and this is how they made more money in these three different ways. They're like, wow. those KPIs look really easy. That is substantially easier than going to somebody and saying, we're going to change your life by a thousand X and they're like, how? And you're like, well, here's our plan. And they're like, well, has anybody done this before? And you're like, no. So the more case studies you can show, the better. And so what I think that really means like at its core is work with apps to have something live that can actually be demoed. That saves a ton of time. Then the second is work with a good application, like have a good application in there where you can abstract some of the video way to them. You know, like I mentioned reward TV earlier, they have content Houses hitting them up every day where we don't have to go negotiate with the content houses. We're rewarded as doing that itself. And we're basically helping them on the back end saying here are the partners you need to work with when it comes to wallet obstruction, when comes to account obstruction, when it comes to FIA on ramp and whatnot. And then I think the third thing on the marketing funnel is make sure that you make your initial 10 to 20 customers happy. Because those testimonials go a longer way than even the case study itself. And I think that's what really changes. Repeat customers are the best form of marketing at the end of the day by a mile. Yeah. Yeah. And retention is one of the key things that is really not paid attention to in crypto. In crypto, we spent a lot of time and money and effort on the acquisition side, but very little on the retention side. And that's really where growth happens. Yeah, it's actually really funny you mention that. my prior life, I did a lot of gaming investments and lot of gaming on the investment banking side of things. it's like, UA and retention is everything in gaming. And in crypto, the thing is, UA is abstracted away via AirDrop. right? You're basically promising away future things. But the problem is the minute that everybody dumps their tokens, you don't spend anything on retention. Because no matter what, you can't convince somebody to stay in your ecosystem unless you're going to say, hey, I'm going to commit another 9 % of the network or whatever, in a low flow high FDB meta, 3 % of the network, in a high flow FDB era, high flow low FDB era, 10%. And so it's like, unless you're going to commit another percentage of that of the network, then nobody really wants to stay in your ecosystem overall. And so that's why the application layer matters so much more substantially. Because that's the only way to actually retain these users overall. Yeah. On user acquisition and retention, you mentioned Kaido earlier. uh What's your view on InfoFi in Kaido and has the camp leaderboard been helpful for camp? Yeah, I mean look, we're close with the Kaido team. We have a Kaido leaderboard, so I'm not going to act like we didn't drink the Kool-Aid at all. That would just be lying by me. I think Kaido proved something pretty interesting, and I would say this dates back as far as Galaxy ah with questing as a general concept. I think gamification in crypto wins. People want to compete, people want to do things where they can prove that they're the number one or number two or number three version of it. And I think that's really cool because, I believe that the two best forms of tokens or the two best use cases of tokens are capital formation as well as community incentives. And so anytime you can do those two things, you know, you're incentivizing people to be a bigger fan of your community. The problem I think is the Kaito and like, I think this is a crypto problem at large. It's like, even though something sounds great in theory, the incentives almost always get corrupted. And I think you're sort of seeing this with Kaito right now because projects are basically paying a shit ton of liquid cash in order to KOLs to get them on the apperboard. Or they're basically saying, hey, we'll give 10 % of our incentives if you're one of the top Kaito users. Now the result of that is if you're a KOL who has 100k followers, you have to do three tweets and that gets you in the top 10. But if you're one of the core contributors or the core community members who's been supporting the project for three years, but you only have 100 followers. then it gets substantially harder to get them on the Kaito work. And so I'm not acting like I have a solution over here. Like our belief is that even for our airdrop, you know, we're doing a combination of those on Kaito, those who are our biggest test net users, you know, those who actually paid for our NFC collection on Ethereum as well. Those are really who we're airdropping to. you know, for us, it's like, how do you manage expectations among those three and also make it such that this is not like a mercenary paid market in shill. And it is more so people who actually want to be here and whether it's want to invest or want to buy the token on the liquid market is the most important thing. Yeah, I've done some work with the Infinex team in Cane and they're taking kind of an upside down approach. They're actually, want to reward users of their application of Infinex, which is kind of like a super app. And they want to reward those that use it and then talk about it versus the ones that just talk about it. And so they're kind of taking a user acquisition first, but actual users of Infinex. I wonder if a similar approach could be taken with Camp where you use Kaido as a funnel and then the participants become creators. Yeah, you can even put it in the reverse as well. It's like people who actually created new pieces of IP tweet about it as well. Yeah, yeah. something that we can track pretty easily. I think that is the best way to do it for a word to it because then it's not just like, look, marketing matters in crypto. Nobody's gonna disagree with that. We could argue that marketing is probably the most important thing in crypto still today, but you need some combination of marketing plus your actual core users overall. Because if you're a core user who feels slighted by the fact that you didn't get an airdrop, I guarantee you, you sure as hell are not coming back and you're probably telling all your friends that it was a complete scandal for all. So that's why I think you need some sort of combination of the two. I think Kydo will sort of figure out what's the right balance where it's not all about paid and whatnot. Like obviously it's also a little bit self-selecting because to get a Kydo leaderboard is so damn expensive, then those products can really don't have the cash to do so. So we'll start to see what it looks like uh from a peer balancing perspective, but it is something that's interesting and I'm watching. And to your point, like as you mentioned earlier, a lot of people tend to do the same thing. We're already seeing like 15 different iterations of Kydo come out right now. whether it's like through the Kaido and like Kaido is launching their own launchpad for this exact reason to try and have tokens launched through Kaido directly. But also you're sort of seeing like, you know, new companies come and say like, oh, we track Twitter real users better than anybody else. We track the engagement better than anybody else. Like each was trying to distinguish one or two different features, but that's sort of proven that this idea of attribution is more important than ever before. um Let's jump into kind of what the competitive market looks like for Camp. um Who do you consider as competitors and what are some kind of differentiators between Camp and others? Yeah, would say that like stories, the most direct competitor. would say the IP token. ah think a few different things. Number one is I think we have a slightly more robust ecosystem. I'm obviously a little bit biased there, but I think we have more unique teams building on Story than they do. Obviously they did a phenomenal raise and they've raised way more money than us. So it's like a kudos to Aswan team there. We're huge fans of what they do. And in a lot of ways, they really helped us because they helped sort of instill the marketing that anything you do is IP. Your data is your IP. Some of your activities online is your IP. And I think the other thing that differentiates us from them is almost taking this like institutional corporate macro level view where it's like let's go get some of the biggest creators on the world. You know, they were doing some stuff with some of big K-pop folks. know they did something with Justin Bieber at some point. Some of the big brands. Ours is the IP of the internet. You know, we believe that most of this new, as opposed to trying to go buy or license this expensive catalog from the Sony Warner Universe of the world, you know, our belief is that net new IP will be created by new creators via AI native IP platforms on the blockchain itself. And that's sort of where it can't place. So we consider ourselves in that sort of long tail. And I think that's led to us having a lot more teams and a lot more success in early, applications ah than what they did. know, kudos to them, the token is obviously trading incredibly well and they've done a phenomenal job of sort of building the brand of why IP matters, especially in this era of AI. I mean, the only other competitor that we're sort of seeing right now is abstract. And that's because abstract has done a lot of really cool stuff around owning the IP in the first place. And I wouldn't be surprised if that's where the industry goes, where you kind of own the whole stack. know, they own Pudgy. know, Pudgy will live on abstract. sure at some point right now it's on sole. But then basically being able to lend against a Pudgy and do things with that Pudgian or Neel, that to me is a really cool flywheel for what can exist in the future as well. Yeah, I'm of the mind that, you know, especially in kind of emerging markets, there's going to be many winners. um And since campus focus on the long tail, stories focus kind of on kind of the big brands. um There's there's more than enough. You know, the pie is big enough for everybody, I guess is my point. 100 % like you look in blockchain is usually things come like twos to threes So like for example, if you look at paralyzed you have Monad, MegaEath and Cnetwork You know you look at the move based change yet after the sweetie and then obviously movement substantially later on the sort of uh DA side of things yet Celestia, Eigen and then obviously avail but you know it almost never is like a one-of-one category creator and then what usually we see is like you want a couple players in this space that uplift the entire category Linda, Ryzen, Tide, Lifestyle, Boats type situation overall um Tell us again where where camp is right now. I I believe you said test net. Is that correct? Yeah, we're in test net currently looking at main net at the end of this month, you know, currently slated for a date of August 26th along with, you know, everything associated with that as well. I think for us the name of the game is hammering our applications, make sure that from day one there's enough stuff that people can actually use as opposed to, as I mentioned earlier, like a boring decks or swaps platform overall, such that when people can actually engage with it, they'll see that there's something to do there. And that's really what's pretty exciting. I think we've built a pretty world-class team with 20 people today, very lean on the burn, but I think our team ships with the, both on the BD side and on the tech side, with the quantum of 150 people, even though despite us only being 20. One final question. Let's just suppose Camp is successful and achieves as Northstar in all of its goals. What industries or stakeholders currently will be disrupted? I would say the entire entertainment industry uh that includes everything from film and TV, images, video, gaming and music. And then I think that all of the sort of data marketplaces would have a similar situation where there's a pretty realistic world where in, call it two to three years. If you're looking to train an agent, instead of trying to find a way to do it with one of these existing resellers of Instagram data or Twitter data, you come to Camp directly and it in most verified and known way possible. But I would say those are the two biggest industries by far. That's exciting. Well, Nirav, thanks for taking us through Camp Network's vision, mechanics, and what makes it unique. For listeners, I'll put in the show notes all the relevant URLs for Camp and uh how you can get involved in the community. Thanks, Nirav. Thank you so much, Peter. This was great.