Web3 Game Development Livestream - Metaverse pioneer Tony Parisi discusses the future of Web3!

Unknown: Make my thumb. Look at this, we got it. We're going to be here. Look at this. We got.

Jon Radoff: All right, welcome back everybody to the Web 3 game development live stream.

Unknown: We're trying something brand new today.

Jon Radoff: So I am in Austin. We are here for South by Southwest. We have never done an on site program before, but this is going to be the first of many because I'm going to be doing this at GDC next week as well. So today, I'm really happy to be joined by Tony.

Unknown: Tony, thanks for joining us.

Jon Radoff: Tony is a legend in virtual reality in Web 3. We'll dive right into the conversation about Web 3 and how the whole landscape is changing.

Guest: Yeah, sure thing. John, it's just great to be here. John and I have become friends over the last couple of years kind of doing this, working on Metaverse stuff together. Real kindred spirits, I think. So it was a delight to get invited to be here at the show. I've been working in the field of Metaverse technology for, believe it or not, over 30 years. It's really been a thing for that long. I'm an expert in 3D graphics and I created 3D graphics standards, VRML, GLTF. And the people at GLTF people are using every day for any use of 3D on the web. So it's very proud, I think, for me after working in this area for so long. I was also a technology executive at startups and large companies. I did a stint at Unity, the game engine company for six years as they're head of VR and AR fairly recently. I left that and was tapped on the shoulder by Neil Stephens in the author who coined the term Metaverse. And he said he wanted to build a Metaverse company for everyone, built on blockchain technology. So imagine the convergence of blockchain and VR3D types of technologies. And I worked as a chief product officer there for a year and a half. But then the siren call of getting back to making art called to me. I'm a musician as well. Now these days I'm working as an independent artist primarily, including having written a full musical, rock opera called Judgment Day about the end of everything. Maybe timely, let's see how that goes. And I'm doing all that in Metaverse technology on the blockchain as well as in real life. It's intended to be a real stage show. So it's a fascinating time for me and it's great being here in Austin with music, live music capital of the world as well. This is so much fun and I see you rock in the Austin look today, Johnson. Yeah, I'm a, I'm a, I'm, I'm when in Rome. I'm wearing the hat.

Jon Radoff: I didn't wear the cowboy boots today just because no one can see my feet. But I've got a, are you going to wear that to the panel? We're on a panel later today. Yeah, I'm going to wear it to the panel. And I've actually worn this while on a horse. So I feel like I'm not a complete poser. So this is so I feel okay with it. Yeah, all right. Well, let's, let's begin the journey from the beginning. So we've got a lot of Web 3 game developers that tune into this. We also have just people who are Web 3 enthusiasts generally trying to learn about what's going on. But you got inspired by the idea of the Metaverse really early on. When, when you invented VRML, was that inspired by the work of Neil Stevenson? Like what gave you the ideas for that?

Guest: Yeah, it actually was. So personal journey side of this was in the early 90s. I was already in tech. I was in the tech industry in Boston actually. And my wife and I got in a van and drove across to California. She didn't want to stay living there. You know, she'd look through up her whole life there. And I'm from New England as well. We're like new adventure. And I've been working in the internet tech a little bit already landed in California. Started playing around with some 3D real-time 3D graphics on PCs. MetaFellon and Mark Pesci and other fellow Bostonian actually, who'd also moved to California. And we hit all red. No romance or by William Gibson. And it's no crash by Neil Stevenson. Of course. Who was steeped in cyberpunk novels. And the internet was coming online in real-time 3D graphics were coming online literally at the same time. Like on a PC, I mean, real-time 3D. Where they were happening at the same time. Now, backdrop to this is the worldwide Web is actually just getting stung. People are learning how to make a Web page. And Mark was like, we're going to put 3D on the internet. And I was like, you mean like Nouromancer? And he said, yeah, Tony, like, duh. And I mean, he's a very visionary guy. But I had coding chops that he didn't. He was like a networking programmer. Whereas I was a 3D graphics and UI person. And we put together a prototype for the first ever 3D Web browser. And I'm telling you, people were barely making Web pages. And we were taking 3D models and rendering them. And at the time, a standalone PC app that was a companion to a Web browser. So mean 3 in the 90s. No, you click on a link in a Web browser and launch this companion app. And it sent it to URL as a parameter. And it would load a 3D model off the internet. And then the big round trip was you could click on the 3D model. And that would launch another Web page. So even though they weren't an integrated 2D 3D experience on the one window yet, it was showing that this stuff could work. And before long, Web plugins came along. And I built the first 3D Web plugin for browsers. And at the same time, we started going to these meetings of World Wide Web developers. And Dave Ragget, who was on the original World Wide Web team creating the standards out of MIT. The original Web Consortium put out a call for a 3D standard for the internet. He dubbed it a VR standard. And we went with that and called it the Virtual Reality Markup language. And then we renamed it to modeling language to be a little more 3D appropriate. And VRML became the thing. It became an international standard over the course of five years and large companies. And independent developers and artists. And everyone was super excited to build the future. A lot of them really fueled by this Neil Stevenson metaverse vision. Which took a lot longer to reach fruition. I think we're here in a lot of ways. Now I'm sure you and I want to talk about this in this segment. But at the time, we were showing really promising technology that were really cool experiences being made. That you needed to put a graphics card into a PC to really experience. You were getting your content over a dial-up modem. 9600-bottled modem kind of thing. So the experiences were crude. Not everybody had the coverage on the hardware side to even be able to do this. And yet, we still did a lot of great stuff. But the hype and boom around that fizzled eventually by the late 90s. And VRML just sort of died commercially.

Jon Radoff: Why, do you think?

Guest: Well, for a couple of reasons I just said, I don't think the infrastructure was quite ready yet. You're trying to deliver this content over a very low bandwidth connection. Though we actually engineered it to do pretty well for that. Right? Most people didn't have the hardware to have the experience. But I think there's actually a third and most important reason perhaps is that folks weren't ready. I mean, folks on the consumer, you know, consumption side of the content. They weren't enough for them that even knew how to deal with the web yet. And so this idea that they could be like spinning 3D models are walking through the universal universes. That was so far down the line for most end users. And the flip side of that is the content creators as well. Most folks did not have 3D modeling skills. That picture has changed a lot. And I think with AI that's going to change even more. I'd love to talk to you about that. But at the moment, even just having the craft the crafting ability to make 3D models. We didn't have scanning technology. Any of that was really, really special. And it still is, but now there's millions of 3D modeling professionals on the planet. It's a very different situation than it was 25 plus years ago. So that, you know, that fizzled, but the RML continued as a standard. And then we worked on some upgrades to that standard for a while. But pretty much that languished for about a decade. And nothing happened until we get to GLTF. But we can stop it there. So that was, that's the origin story. That's the big old origin story. A handful of crazy, you know, Bay Area Northern California folks wanting to build cyber space. The other name we use for back then. Back then, no more with the information super highway.

Jon Radoff: Remember all these old terms. So let's do a quick station identification. You're watching the Web 3 game developer livestream. We've already got a couple hundred people watching. So that's really awesome because you are here with an opportunity to ask questions live too. So Tony and I will have no problem keeping this conversation going over the next hour. But there's absolutely no reason you can't be a part of it. So you're present with one of the legends who created virtual reality as we know it today. The inventor of the RML who's been involved in really all of the big things that have happened on the internet since then. And that brings us maybe forward to Web 3, which we'll talk about a little bit. But please, we're broadcasting on Facebook on LinkedIn on X on Twitch on YouTube. All of these places you could ask a question. Oscar, our producer, he's there. You see him online. Oscar will happily take your question, put it on the screen, and we can talk about the questions you have. In fact, you could even join this livestream. If you have a camera, Oscar will even give you a link to StreamYar. You could jump right in and be part of the conversation. So this is a very open format. We're trying to almost build the metaverse version of this by making it a place you can come in and out of. So please, ask your questions, be part of the conversation. All right, so I just said the word metaverse. We began with it. It almost feels weird in 2025 talking about the metaverse because it had all this hype a couple years ago. And now it almost feels retro to use that term. And it's sort of like that's the panel we're going to be on it. So by later today is like, is it coming back?

Guest: I kind of shock they accepted our panel submission with the title metaverse in it. But I think we were intentionally provocative and how we framed it to say this is an update. Hey, it ain't dead yet. Kind of thing because it seems retro, but the fact is this isn't evolving work in progress. And I think that's going to be one of the throughlines of what we talk about in the panel today. It's no one end state. I mean, okay, maybe we can get to the holodeck and we can all just dream up a world that just appears around us in full 3D for real. Like around us as hologames. Maybe then we've actually achieved it, but there's so many steps in between. And I think it's more of an aspirational idea of a set of technologies, a bunch of different experiences. Right? I went through in gaming, for example, being one of the major things I think it's going to be thriving and driving action in there. So yeah, it is wild that we're going to be talking about this today. But as someone who's been working in the field, I've never shied away from it. It got pretty funny when Zuck started saying metaverse and he rebranded the company. And then a lot of people were like, oh, we need a new word for it. You know, so I'm in this like insider crew of industry people that argue over terms all the time. I don't know if it's the same in the world of beamable and the appear web 3 gaming, but I got to tell you like when it comes to XR, like VR, AR, you know, all the puns, it's all attack experts like me. They start getting into these stupid little debates over like what are we going to call this thing? Because it's a marketing problem because otherwise everyone says it's a marketing problem. We're all going to be using it if we just come up with the right name, you know, or the right three word message and ain't true. So like when Zuck said that a lot of people are like, oh, no, we can't call it the metaverse anymore because Facebook is going to own it or they're going to make it shitty or something. We are rated here. Can I say that? Anyway. So yeah, it's funny, but I think it's so exciting. And ironically for me to say that we are still kind of early stages on so many things. Though I think the tech underpinnings are really getting super solid.

Jon Radoff: So let's talk about that because I think there's a whole bunch of buckets of tech change. There's also some social and culture change that's accompanying this, which I think is worthy of identifying too. So you referenced artificial intelligence and how that may change things. Clearly the graphics technology has gotten way more advanced. The amount of polygons you can put out on the screen in very high speed is completely different. Like we imagined that happening in the 90s. I don't think I could even conceive of the hardware that would be capable of that. We're talking about literally, I don't know, millions or billions of times faster hardware today than what we had available then. So there's that. There's also blockchain and web three. And this is the web three game developers live stream. So let's maybe just begin there because you've also had your own journey through web three and blockchain technologies. For you, what do you feel the relevance is, I guess, to metaverse online world creation game development, all of these that you touched all of those. But where does blockchain fit into that? Why is that something that adds to this?

Guest: I think there's several aspects of blockchain that excited me enough to go join a blockchain company. It wasn't just because my eye don't even see them and called me up and said you want a job. This is the direction the world needs to go in for a few reasons. Decentralization. It's a term we throw around a lot, but it has so many implications. Decentralized infrastructure technology, game servers that nobody owns and controls.

Jon Radoff: Like what beamable does? Like what beamable does.

Guest: You know, it would want to be overt about it. But absolutely what you're working on, John is super excited. Yeah, thank you. Decentralized infrastructure, you know, to storage and delivery and all that. I mean, we talk about a lot of things around ownership or money or crypto. I'll get to that. But the infrastructure needs to be decentralized to enable freedom of choice for anybody creating experiences, building technology that can interoperate with an infrastructure like this. And so that's critically important. That second factor is ownership. That is digital ownership of your own goods where you're not beholden to, you know, some platform who actually owns it and you're renting there. And a related element is monetization. Right? That we have digital currency backing digital goods through smart contracts, you know, the convalidate those. So all of those together are fundamental to create a free and open anything, you know, in the future of the internet. Right? So the term web 3 is actually going out of the idea of a third generation of the internet. And I forgot that you might remember who coined the term. I forgot in the name, but it really is the next generation web. And the fundamental innovation in that. I hate to say it being a 3D graphics expert. It's not 3D graphics. It is ownership. Now there's another important piece of technology, which is sort of live. Live stream live experience to me that the defining characteristic of the metaverse and this may be a spoiler for us today because we're cutting to the chase, but to me it's about being connected live. Whether it's live streaming like this or being in discord or being in a Twitter space doing audio. It's this real time communication that is then represented in 3D graphics or a video that brings us all together there. But back to web 3, which was your point. Sorry I could talk all day. We can take it a lot of actions, but then we can go back to it too on the streaming side and the live side. Having decentralized control is fundamental to the future of anything that anyone wants to build unless they want the future control by a large entity. And that could be a tech service company that could be a platform that's been gating the distribution and what people can access, which then leads to real problems and free speech and some other things that we could touch on as well. Yeah, web 3 is fundamental from decentralization ownership and monetization might be.

Jon Radoff: We can talk about decentralization a little bit. I think sometimes people in blockchain talk about decentralization as if they invented the concept, but actually it the internet was decentralized. Emails a great example domain names. So could you imagine a world in which to have an email account you have to ask one company whether you're allowed to have the email account and they decide what you're going to be allowed to do with email or domain name. Would you imagine that world in which to have a domain name you have to subscribe to, I don't know, Microsoft's domain service or something and they let you know whether you'll get to do it. What your price will be every year as it goes up.

Guest: I love Microsoft and what they've been doing the last decade, but you're also are probably old enough to remember that if it was 1997 and they had their way, that's the way it would have gone. Yeah, I'm not taking anything though. They tried to own the entire internet for a while there and then they actually got broken up for it and the company's done so much better since then. Actually better, better for everyone. So yeah, this is this is the point like DNS well DNS itself is obviously mediated and curate them. There's a process you have to go through to get the domain name. There's governance around it, but it's not controlled. That's one of the key aspects. It's like, you know, any service and web three. There's also governance for that. There can be at least right like an E and S or something like that, but the protocols behind it are open and anyone can serve it up if they participate in the rules of how to serve it up and a decentralized infrastructure, which if it's okay. I want to kick it back to you to talk about beamable and educate me a little bit. I mean, obviously your listeners know you already, but I would love to hear like where beamable is in that piece of decentralized infrastructure. What's that?

Unknown: What's that?

Guest: That term gone.

Jon Radoff: No, deep in deep in is the category of technology that we're building at beamable. So deep in stance for decentralized physical infrastructure network. So the idea is that you can have community ownership of infrastructure. So historically, if you go back to the original internet, when there wasn't that many users on the internet, people actually own their own computers. They'd have a computer. It was everything from mom's basement. You might have a little wrap back. And that was the web service.

Guest: Right. They were serving it.

Jon Radoff: Yeah. My first web server was running on a salarist workstation in my office. So we actually ran it, hosted it on site on the T1, which would be a very low speed connection by by today's standards. But that's how we get started. But then more and more users started coming. So you had a choice. You either have to build out a data center. Very expensive, very capital intensive. Also requires a lot of labor to run things for you. So what came out of that was companies like Amazon and Amazon absolutely gets a ton of credit for pioneering in the space. As much as I'm going to critique the state of affairs today with big companies. Like all these companies have done great things like we wouldn't have the internet as it is today with a lot of these guys. But so Amazon created Amazon Web Services. And today it's millions of servers basically there what you call hyper scale or where you can essentially rent capacity on their machines. But what comes with that? Well, they tell you the tech stack that you have to use. Although there's a lot of flexibility in it, but they essentially have their own tech stack. It's in their interest to get you more and more dependent on their own tech stack over time. Because then you're stickier to them as a business. Yeah. It's higher priced because you're renting servers versus owning stuff yourself. And there's a lot of value added layers and things like that on it. So in the specific case of Amazon, they've also released technologies that they've decided to discontinue that people have been dependent on. So good example.

Guest: Yeah, you get locked in with someone and then yes, they don't serve us a piece that you need because it's not an interest in the game industry.

Jon Radoff: They bought a platform called Game Sparks and like thousands of games were built on this. So they decided ultimately to discontinue a post-acquisition. My understanding is the technology didn't actually scale that great or there were some issues with it. So they decided to end that, but of course they stranded many, many games as a result of that. A few could do the hustle and figure out how to completely change out their back end, but it's not easy to change out of back end once you've made choices. So this is when you run an internet business of any kind and you choose a technology provider, which is centralized that gets to dictate the business terms to you, whether you get to be there at all, whether they like what you're doing. So it's interesting to me, you mentioned renting, and it's kind of an interesting metaphor to explore for a second because you are renting from them.

Guest: But imagine this in the context of a home. So if you're a homeowner, you have a lot of hassle on your own. If you rent, the landlord takes a lot of that hassle away. So that's a choice you can make for a lot of people. But this is like your landlord also telling you what kind of furniture to have in your house. And maybe what style of clothing you wear around the house or something. So the making choices for you that maybe go beyond.

Jon Radoff: It's like a worst H-O-A.

Guest: Yeah, it's a good word to H-O-A. Or if you're a condo on it, there you go. That's the worst possible. Yeah. So I think that's a kind of fair analogy though it breaks down to a certain point. But yeah, so then there you are locked in and your world choices are being dictated in a lot of ways by someone as a platform provider like this. So back to then, D-Pen, and B-Emole. So like bring us into that world then.

Jon Radoff: So the idea of D-Pen is you have community ownership of the infrastructure resources. So instead of one siloed company that owns all of that capital equipment operates it and detakes all the terms, you have a community that owns the infrastructure. And it can be spread across multiple companies. So in the B-Emole D-Pen, for example, there are bare metal service providers. So bare metal service providers are data centers basically where they just give you computers. Give me the box. And then you figure out what to do on them. It's much less in terms of software infrastructure. It tends to be a lot lower cost, but there's a lot left to you. But if you aggregate together a lot of these bare metal service providers, you're able to create the equivalent of a virtual hyper-scaler. Like an Amazon or like an Azure or like a GCP. Because when you add up all these data centers around the world, it adds up to a lot. And they're secure and they're high bandwidth. They do all the things that you'd need as a game developer to operate your back end, but at lower cost. And because it's spread across many providers instead of stuck with one. If one of them decides they don't want to be in that business anymore, you're not screwed. You just shift your traffic to somewhere else. It functions as an order book, which is where some of the blockchain technology gets involved. Blockchain happens to be an ideal technology for running things like order books and identifying what is the price that I want to ask for when I'm offering my infrastructure back to the community. And conversely, if you're a game developer, what am I willing to pay the order book then matches. Make sure that you get an infrastructure at the market optimal price, but open. So what we're doing is taking our tech stack and deploying it in an open source manner across what's an open network. I think you could think of deep in as the. Take the logic and the ethos of open source, but apply it to the world of hardware and infrastructure and data centers. That's what we're building.

Guest: So I wonder about this and this reminds me of node operators on a, you know, Ethereum network or a Bitcoin network. And I'm sure that's actually pretty close as an analogy, right? So if I'm providing services into the beamable deep end, what is my economic incentive? Why are people doing this now? Because this is, this is one of these keys to this whole thing is most game to, and I want to get into it. I'm packed how it gets in, how we get into this for game development, specifically what game devs have to go through to serve things up, right? They're not going to come in and rely on a decentralized infrastructure. What are the people providing them that infrastructure getting in return? Because this only works in sustains if there's an economic incentive for both sides. Yeah, great question. Look at what we talk about that because I'm pretty curious. Yeah, totally. How have we not had this discussion in depth, you and me before, right? But this is, yeah, fascinating.

Jon Radoff: Well, let's go back all the way to some of the earliest crypto mining that existed. Like you go back to Bitcoin. So what is Bitcoin? You run some software that happens to be solving a lot of cryptograph puzzles and eventually you solve it and you get rewarded with a Bitcoin. So that's, that's, or multiple Bitcoin. So that's, that's how Bitcoin mining worked. That's then how Ethereum effectively works through proof of world algorithms. Now over time, people discovered that for a lot of use cases doing proof of work was just not really the optimal. The thing to do because he ended up with very high cost fees and expenses to do small transactions. And it wasn't good for a lot of applications such as games, for example. So games have a lot of little, little micro transactions and you can't be paying 50 bucks of Ethereum every time someone buys a sword, for example. So they started coming up with different algorithms called proof of stake. But what Deepen is doing is it's sort of going back to this original idea of essentially mining for cryptocurrency, except instead of just running cryptocurrency puzzles, which is only of use to unlock the next coin in the proof of work algorithm. You're doing actual workloads that people need done.

Guest: And presumably way more efficiently.

Jon Radoff: Yeah, but exactly. And there's incentives all around because the game developer has an incentive to make sure that they operate their game servers in a efficient. Cause they're going to get for it. Yeah, exactly. Likewise, the data center service providers want to be paid for offering the service. So in essence, instead of mining for cryptocurrencies, where you just run in crypto puzzles, you're actually running game servers. So that's really what the beamable deep in network is is game servers that you can provision. And you pay the data centers for spinning up game servers for you, just like they could for any other service that they offer. And this model is really taking off in crypto, not only in our use case, which is game servers, but all kinds of things storage. There's really interesting ones like helium was probably the first one helium made it possible to have essentially a Wi-Fi hotspot. And you get paid in their cryptocurrency for offering Wi-Fi. So if you run like a small business where a lot of people are coming in and out all the time, you want to offer Wi-Fi service there. You actually be paid to offer them Wi-Fi service. And they've actually done now major deals with traditional telecom companies like T-Mobile. So it's starting to have a way beyond.

Guest: And you need those multiple layers of the stack in this economy. And if those folks are starting to participate, that's going to make better economies of scale for the software providers and so on, the infrastructure providers. So that's great to hear. That's really encouraging. And all right. So then Web 3 gaming is going to take off now.

Jon Radoff: Well, that's a funny thing to say. So I wanted to sew out of that loop.

Guest: Right. I can't wait to hear what the leg is.

Jon Radoff: We can talk about Web 3 gaming and we can talk about Web 3 enabled gaming, which I think is maybe two different things. So at Beamable, we have both kinds of game developers. We have Web 3 game developers. So for example, mythical games is probably our largest Web 3 game developer. They've got things like the NFL license, Pudgy Pagglands, a FIFA license. So worldwide, absolutely massive sports franchises that they're building games on based on the mobile. They're a Web 3 game developer. But actually, most of our traffic is what you'd call in Web 3 Web 2, meaning just traditional MMOs, online games, multiplayer games of all kinds.

Guest: But using the deep end of the Web 3 infrastructure, you're saying.

Jon Radoff: That's the cool thing. So for years, we've talked for at least the last five years, people keep bringing up the subject of onboarding game developers to Web 3. The challenge with that has always been that you're asking game developers to change their business model because having an open market game with a trading economy in it. First of all, that's a very complex kind of game to build well. I'm in favor of them. I think Magic the Gathering was the non-digital version that came before all the others. There's lots of good examples where this can work. Roll blocks shows kind of a way it can work at scale. Minecraft in some ways even shows a way it can work at scale, not in blockchain, but in similar kind of trading economies. The issue though is it's good for a category of games. There are plenty of games that don't need a trading economy or an ownership economy. I don't believe it's actually true to say that every game that exists is going to be an ownership based trading economy game. I played Baldur's Gate 3, probably my favorite game of all time. I don't think it would have been made better with a trading economy in it. It was good as implemented, and I wouldn't change much at all about that game. But those are games that represent a massive amount of the server footprint in the 200 billion owners.

Guest: Different infrastructure improved infrastructure. Yeah.

Jon Radoff: Exactly. So we're onboarding them to Web 3. They don't have to know a thing about Web 3. They certainly don't need to change their game design. They just build the game the way they always would. We give you an SDK for Unity and for Unreal. All the blockchain stuff is abstracted out. We provision off of the deep end network automatically. You don't have to know anything other than we make your life a lot easier and we get all your servers set up for you for your online game.

Guest: That's amazing. But now let's go back to the other one. The Web 3 gaming one. I get the Web 3 enabled gaming part. And just as an engineer, as an infrastructure person in my DNA when I'm not making musicals, I think that makes just total sense. If you can again line up all the pieces, the economic incentives to the service providers, economies of skill that actually work technically, get it working efficiently and get developers to embrace that because it's just going to benefit them in the medium term. They'd have some switching costs, but then they're going to have some great benefits out the other end. But this other thing that's Web 3 and gaming that had some real hype behind it for a while and it's gone through a circuitous and torturous path the last couple of years. Where are we at on that? Because I have no idea. I've been busy making music and not paying attention to this stuff for a while.

Unknown: Real fun things at least. So what's happening there?

Jon Radoff: Well, I mean, it's safe to say Web 3 games have super struggled for a lot of reasons. Some have been successful. Like I just mentioned, mythical games like NFL rivals is a game that figured out a path to millions of users by frontending it with a game that just looks and feels and plays like a traditional mobile game. Based on a big IP and based on a big IP and then after your deep in, you discover the open trading network with that that also surrounds the game, which means you're going off mobile and into other places to engage in that. But the interesting thing with their business model as I understand it because I actually had John Lyndon the CEO of Mythical on just a couple of weeks back. He has explained that their ARP Dow, so average revenue per daily active user has started out with very little Web 3 revenue and more on the mobile revenue side over time what the contribution to ARP Dow from the trading economy is actually become most of the games. Wow, it's become a sustainable source of revenue for the company. So I think that's really promising. Yeah, so I think that's an interesting model there. Obviously collectible sports personalities is a well-worn idea from gaming, even before gaming going back to trading cards and stuff. So it's an idea that I think translates really well to an ownership based trading economy game. So I think there are signs where it's working in there. It's a game that has had several million installs. FIFA, you could reasonably forecast is even bigger. I think globally it's three or four times the size of NFL just in terms of audience. So maybe it'll factor up to that.

Guest: There's nothing massive massive and also I assume you're not implying and I don't think you are that you have to have a big IP attached to this to make it successful, but it really is helping right now in the sense that now we're seeing proof points. They are based on big IPs today, but it's just a matter of time that as people are seeing the money here, they're going to come back in. Some indie game developers going to come up with some hit that we aren't even expecting. That is trading based. That is pure Web 3 based.

Jon Radoff: And we'll get back into the cycle in a big way, I think. This is what I'm hearing from you. It's possible. Yeah, I think I mean I built IP based games for years on mobile. So I built Game of Thrones and Star Trek and walking dead games. And anytime you have a brand new platform where there aren't a lot of entrenched competitors yet having well known worlds in IP such as an NFL franchise or a game of Thrones in my mobile example. That is a way to gather a lot of attention for a product. Now I also found that over time the benefit of that IP can go down a lot because you end up in the state of like competition between every IP that you can imagine and every game IP. But that's not true in the earlier stages. So we probably are at that earlier stage of IP within the Web 3 ecosystems. Because they'll tend to have their own unique distribution channels. We don't really know what the distribution channels for Web 3 are yet other than that telegram over the last year has gotten enormous traction with telegram. Yeah, that's really something, right? Which points to that. But you know, but Web 3 gaming generally other than there are some real interesting really success stories like mythical games. There's they've struggled for a lot of reasons. One is on the business model side. We're still in the discovery mode there like people have lots of hypotheses about what is the game design that's really going to work in Web 3. Well, we know that something like magic the gathering works sometimes. But there's also been lots of other ideas and not a lot of them have shown to scale out for big mass audiences. So there is still the experimental process that we're going through to discover the business model. Another is technological. We pointed it a little bit of it with a deepened stuff. Just the reality of Web 3 games is they're all inherently online games because they've got market systems and social systems and multiplayer systems typically. So your complexity level for the game has gone up a lot. And the technology to support that often just gets in the way of building a great game. So what you want a game developer to be able to do and what I was when I was a when I was building games like game front is you just want your team to sit down in front of unity for example every day and be working on a game. Not be figuring out back end infrastructure and scaling up to millions of users or frankly that's just kind of boring. You have to do it super well where you fail. But it's not fun to work on the game. And it's entirely specialized. And yes, it's not fun. And your customer doesn't really care. They only care if it breaks. Right. Exactly. If you log into the game and it's like servers down your pissed off is the player. But when it works. He's the player just expects it to work. Yeah, absolutely.

Guest: And I care about what the pixels look like or what the gameplay is like, but they don't care about that. Like you said in the super ex. Absolutely.

Jon Radoff: And the infrastructure side though, like we're hoping to have a future where you choose the most important parts of your game. But you choose the front end technology unity unreal good. Whatever you think fits the way you want to build your game and that delivers the game experience. And then infrastructure is just something you can build on top of without really having to think of it. As opposed to either the other options are build it all yourself. Very, very capital intensive, very, very risky. Or you could try to build in one of these metaverse environments like Roblox, for example.

Guest: Well, then you're back to being having the whole stack controls, right? Whole stack control. The Fortnite, you know, creator stuff. It's all basically controlled by one entity. Right.

Jon Radoff: Yeah. And you and Roblox is what they've achieved is really fantastic. I don't want to take anything away from the enormous success that Roblox has had. I think it actually proves a lot of the things that we've talked about for years. But they in addition to the 70 or 80% take rate that they have. They also just make a lot of tech choices for you. So there are certain kinds of games you're not really going to be able to build in a Roblox. So giving people choice of technologies that they bring together gives them the ability to realize their vision.

Guest: Fun fact. 2010, Dave Bazooki hired me to port Roblox from the PC to the Mac. He had a failing project in house. You know, it was always like sort of second fiddle anyway. Like, maybe we need to get it running on the Mac. Most people are on a PC. So I went and fixed that project and my son at the time who was 10 or so or 9. Was the first alpha tester of Macintosh Roblox back in 2011. So I'm a huge fan, but I also know what the code base looks like inside. And I would just say, you know, yeah, you want to maybe you want to have some of your own infrastructure for that. No knock on Roblox. Like you said, they've done what they do is amazing. But also they have not only made tech choices for you.

Unknown: They've made art direction choices for you.

Guest: So unless you want to build the game like Roblox or in Roblox. And then they made distribution choices for you. You have to have the game in Roblox. It's not like you're using that as an engine to put on your own site for your own download. So it's back to that same thing we were talking about for fan companies and infrastructure. They control way too much of the stack. And it has that has palpable results for the end user. So you're never going to build a AAA game in Roblox no matter how I try. So but the open worldness you get and all the ease of creation you get. Maybe the right choice for you if you wanted to build something like that. So yeah, I mean, it's just back to having a big platform. It's controlling way too much of the stack. It's not good for the next gen of developers or for the open metaverse.

Jon Radoff: I mean, the interesting thing with Roblox is it's almost this midpoint between the world of traditional game studios on one side versus this market of UGC content. So UGC user generated content. I mean, there's huge games now that started out as UGC projects basically counter strike or even League of Legends was inspired by Dota. Which was basically a mod. So UGC has spawned huge huge industries. Roblox kind of feels like it sits in the middle of that. It's like harder than a lot of UGC more way more constrained than a game studio would have the options for. But more flexibility than say just building a mod pack based on top of the game.

Guest: It's funny, it is harder, but I mean, if you're eight to 12 euros old, you're in it. Like in order like Minecraft, the same thing that the hardness of it is the challenge. It's part of the gameplay actually to create in there. You know, the constraints and all the things you have to do. That's the kids who do that have this sense of mastery. There's an engine behind it's a social engine.

Unknown: If you get good as a creator in those systems, you are a god.

Jon Radoff: I love that though, too, because that's the other change that we talked about technology is changing. But culture is changing. So today we have a lot of game developers who started out where it was Minecraft or Roblox. It wasn't download unity and start hacking together a project. It was Roblox is where he built a game.

Unknown: And in some cases they ended up with fairly sizable.

Jon Radoff: And the most extreme cases, kids became millionaires, they had millions of people playing it. But even kind of sub scale, like your building stuff, where probably you got hundreds or even thousands of people to try it out. So you're interacting with audiences and you're crafting and creating. So that's a really interesting change as well. It's sort of like that YouTube moment, right? Like it used to be that creating media for that matter. Like we're live streaming video right now. What I'm doing with a iPhone on my camera and a MacBook and some online software and a couple of snap on mics. Like to do this 20 years ago, we would have had a production crew in here. Exactly. Tammerman and all kinds of stuff.

Guest: So you should have seen 15 minutes before we went live. John had the manual out for these digital laptops. He's a piece of paper. I'm like, okay, we're going live in 15 minutes. So yeah, it really has changed.

Jon Radoff: So, but yeah, and YouTube and live streaming technology has made video a lot more accessible. Well, now that's happening with games and immersive world creation.

Guest: Except, sorry, I got to get you up in one second.

Unknown: Yeah.

Guest: Except with YouTube, you make a video in a standard format. True. And you don't do that in Roblox. You don't do that in a Unity game engine. So this is full circle back to what I have been spending 30 years of my life working on. So, the formats that you build that 3D content and need to be completely open are really aren't going to have a true YouTube analogy here. I mean, you can take the video you put on YouTube and you can put that anywhere. It's not like you can't run it on another video service because it's MP4s. Yeah. Right? I mean, the other codec issues sometimes and all that. But in the large part, you need to have a world like that where the content was created in a way that was portable and an unoperable across networks with the 3D content and we're not there yet. Yeah. So it's a glimmer of that, but it's not quite there when you come to Roblox or building in the Unity. It's like they're missing that last step. You need to be able to create that content in a way that could be put somewhere else.

Jon Radoff: Is there always going to be a race between the standardization effort for that kind of immersive 3D graphics experience and just the people who are

Unknown: going to push the envelope on what's possible and intentionally get out of the box like an unreal engine, for example, with like

Jon Radoff: Nanite and Newman and these new technologies that they've created to do more real time geometry and lighting. Like that's not based on any standards they did it because game developers wanted it needed it.

Unknown: Maybe it's not even all the way there yet, but it shows you sort of this tension that exists between a standardization effort and just wanting to push it.

Jon Radoff: I mean, it is a race. Yeah.

Guest: It is the tortoise versus the hare. Right? It is game engine tech is always going to move like the hare. But one day the game is just going to wake up and be like, oh, we should have done more of this with these kind of standards. Right? Either the devs are going to say, like, I'm sick of porting or whatever it is. Now, look, in gaming, it is this difference is really pronounced because game engine tech is way past what you can do in any web browser. But, and this is where it gets a broader vision of what the metaversus. If I am creating a character, if I am creating a virtual good, from creating assets that I want to exist outside of my game, I should not need the engine to render those. That's content versus the gameplay. And so that's the world we need to get to. And it doesn't always have to be rendered photo realistically like a metaume and on a, you know, the latest PC. Right? I might want something that can run an asset that can just be in a mobile game or just put on a website and spun around to see if I can, if I like it before I buy it.

Unknown: Right?

Guest: And then when you get to use cases outside of gaming to the broader metaverse, create a 3D model of this space. I mean, I want people to see it all over the internet. It does it need every shadow and every little reflection. It might not just to convey the sense of what is in here in this Hilton world right now. I think, you know, depending on the use case, right? It could be just a simple 360 to a simple spin. So a lot of it depends on the use case. But yes, the rendering quality of standards base 3D is always going to lag behind what game engine companies can do. And if you're making a high quality high production triple A game and you need the highest production value, I would never say don't use the best game engine you can find. So it depends on use case in your economics really. So it's an it's an end. Right? For many years, and this has been somewhat the bane of my existence. I think people have treated it like an ore like it's like, oh, I need it to look like that. So I'm not going to go bother with these using these other technologies. And then they get it to look like that. And they realized that gave me this one limited use case and didn't actually give me what I needed because you know what? I didn't need all that render fidelity to make this game. I need a better gameplay.

Jon Radoff: Yeah.

Guest: Right. But pixels, I mean, we were just so visual, like the pixels really captivate people. And so I mean, come on, it's like you go to ease three and you see all those trailers for the games. Those aren't the game, right? People make a slice of this thing that they do it with like the pre rendered CG. And then you go play the game and it's like this doesn't look like the trailer. But it plays like it was promised. So I think it's fascinating. And I think the world has changed. And I think even the biggest game engine companies are appreciating the openness of this clearly, you know, it's a big fan of Open because he's been fighting that fight with the app stores with Apple and Google and everything from the economic side. But even Tim and that gang are so obsessed with rendering them out of human that they often lose sight of the big picture. So a lot of people at Epic are heavily invested in metaverse standards too. So I think the tide has shifted. But as someone who's kind of in that sense brushed up against gaming gaming companies, you know, game engine companies in particular for a long time for a couple of decades. It was frustrating for me. So that of course it was very gratifying way ended up getting a job at Unity. And I know David, how could some for years, by the way. But there was never the time the time was like, oh, now we have XR. We need more Open 3D. We need, you know, what was becoming the consumer known metaverse. And that signal to shift, but the shift is taking a lot longer than I personally would have liked. Yeah, so it's a race, but again, I think it's kind of toward us versus the hair in the sense that eventually even the rendering will get commoditized. And then everyone really will understand that it's just like I need to make my assets and my content in a way that they can be used in the developers in the most places possible. And the consumers will care if they've invested time and energy and it purchasing or creating virtual good that they want to move from one system to another.

Jon Radoff: And we also, and the lowest fidelity experiences also the most popular in the world, we shouldn't forget that too, which is Minecraft.

Unknown: Yeah.

Jon Radoff: I don't have a billion users or whatever it is every month. I mean, it's absolutely massive, but it's a great platform for just creativity, which reduces things down to the purity of the experience. And the game play and the interactions in a lot of these worlds, just the social scheme of that's in place, the progression mechanics.

Guest: Yeah, and immersion in our heads, really, right? It's all in our minds, right? It could be physical playing cards, right? All the way up to the highest rendered game. And Minecraft is a particular example of what we're talking about, absolutely.

Jon Radoff: So we mentioned artificial intelligence a lot earlier. We've been through Web 3. We can always come back to the blockchain aspects again. We talked about deep in, we talked about some of the technological changes. You mentioned artificial intelligence. How do you feel that that's going to change things in terms of metaverse or just general 3D graphics creativity?

Guest: Well, like a lot of people in this world, I would say I'm a very big hypocrite when it comes to AI. I'm old, so I have a little bit of a get off my lawn feeling about it in some ways. Here comes the next monster to potentially destroy us all. In my darkest moments, I'm fearful of a big extinction event, long term, not because there's anything in AI technology inherently bad. It's just that what will happen is people with a lot of money will give people with not much emotional intelligence, free reign to go try and create a bunch of crap that might destroy us all. That's just the nature of the technology business. But that aside, I have been warming up to a lot of AI over the last couple of years.

Unknown: Mid-Jerny, great, know the founder, David, this is brilliant.

Guest: He actually came to my backyard one day and was telling me I'm working on this new thing, generative AI art. It was like good luck with that kid. I didn't get it at all. And then a year later, mid-Jerny's blowing up. I'm like my dad when he said the electric guitar was never going to go anywhere. He's a musician too. I just didn't see it. I just blinds him. So, you know, big mid-Jerny's or big chat GPT user, but also as an independent artist, I'm worried about copyright infringement and other things that are happening with the machine learning. I'm worried about folks mistaking AI as this panacea that now, you know, anybody can be an artist. I think the big difference between anybody being able to create with a prompt and being an artist, and that is the thing of philosophy. We don't have time in this session to get into it. So, all these things taken into account though, I'm using AI tools a lot now in my creation process. I use mid-Jerny a lot to generate art. Now, I can sav my conscience because my wife is a visual artist. And she's over my shoulder art directing the things I'm doing for any concept art or even production art. So, I mean, in a sense, there's an artist in the loop in there anyway. So now, with 3D, and I'm really excited for the possibilities of world generation in 3D. I think that is not a cure all either. I think we still need artists involved in making that happen. I'm also, I've got a slight bit of butt hurt on one part of this, which is that a lot of people I talked to were like, well, AI is finally going to solve the 3D, you know, creation. It's always been too hard. And I'm like, you know what? There's this middle ground of people just needed to make some good no-code drag and drop tools for a lot of 3D use cases. Instead of folks beavering away in blender with this really crappy interface to make their art. And nobody's done a good job of doing that kind of easy 3D creation in a drag and drop tool in my opinion still. And everyone thinks, well, you know, just magically add AI. It's not. You need all of those things. You still need good intuitive, you know, easy creation tools, level above the Minecraft and Roblox stuff, where it's really easy. And that now, imagine, I use AI to generate the hard stuff, the 3D model part of it, the pixels, the polygons that go into that. That's still artistry, and if AI can help with that, that's really important for 3D because there's just many more assets to me. The 3D world has a lot of assets in it. If you want to make them interactive, independent objects, people can drag around, then wrap in smart contract, and turn it into a virtual good, right? Put code behind it. There's a lot of those assets that need to be created. And so, you know, I'm just imagining this time when we walk up to the computer and say, hey, holodeck. There you go. Make me a world. And boom, and it's all around you.

Unknown: And I don't, maybe I didn't too far off. I mean, we've seen these things go, and they go, and they go and fit and starts.

Guest: And then boom, one day there's an explosion because someone makes the next piece of computing hardware that makes that better, or there's a breakthrough culturally, and all of a sudden everyone's doing it and pushing these things farther. So, I think AI is going to get us the last mile, excuse me. AI is going to get us the last mile on a lot of this stuff, and I don't think the generative AI tech is quite there yet. I think there's more machine learning needing to be had. But I'm hearing murmurings that the 3D mid-journey on the way. Yeah. You know, so we'll see. I'm assuming the initial attempts will generate. It's like what's going on in music right now. The initial attempts will generate a world you can walk around in, but it won't be made up of separate objects. And then someone's going to have to take it into that thing into a tool and then bust it up. Yeah. Like you do in music now, where the AI generates the music, and then you bring it into some other tool to separate the stems. So then you can bring it into logic or pro tools to actually fix it. Yeah. So it's kind of goofy and clunky. So the AI sees all this 3D content, and then the robot makes the 3D content like it's seen, but varies it for you, and then you like it. And then you have to go back to these crew production tools to break it up. So that's probably the next step we're going to see, but that's still, you know, a step along the way.

Jon Radoff: The analogy there to some of the music tools that are currently available is interesting, though, because on last Thursday's program, we had a really interesting couple of guests. Carismatic.ai, not charisma, sorry, charismatic.ai and then Floyd. So Floyd is sort of operating in the world of procedural model generation, where they have a lot, a library of components that are parameterized. Yeah, parametric. And then they combine them together, but they also use generative AI for prompting and essentially like a user interface into working with a procedural system.

Guest: So there's really interesting. So there's a language model, but then they extract this parameterized data and put it into the forms that are prompted. That's so cool.

Jon Radoff: I think that's right. And then the language models generally becoming a user interface. Right. To a lot of capabilities, which is that's kind of what the holodeck did. Like you could imagine the computer running the holodeck on Star Trek has a language model interface. And then under that are a zillion other like subsystems that it has to figure out.

Guest: I mean, can we cut out that middleman and just go to a conversational interface though? I mean, yeah, you can do text speech to text and get your prompts that way, but it's sort of then you're still just speaking the prompts. We just get to the like, hey, I need this kind of world. Please put it there and then you make a Victorian style. Okay, cool. It's right. Nice. Yeah, like I thought about this. I'm not going to pat myself on the back just odd how this works more than 10 years ago as I was steeped in all the XR stuff, the R stuff. And I was thinking, why am I still like using sublime text to like write the code for a website? Like, can I just walk up to my computer and say, hey, give me any commerce website. Make sure it works on these platforms. Use these packages and boom. That exists now, I mean, typically you're typing that or their website builders now that do that. They haven't they have an LLM interface. You go type a prompt and it gives you all this code and there's a little code window and you see the text and then it gives you all the builders have the time to fix it, but it's getting there.

Jon Radoff: Yeah, no, there's and there's Claude has something called cursor where it's where you being either non technical or just semi technical can almost dictate the kind of application you want it to build and work with it interactively. So yeah, much, much more of our world is going to become this language interface.

Guest: I'm, you know, I'm a generation that's annoyed by it, not only because when I making my music, I put in my thousand plus hours on all the instruments that I can play when it comes to coding. I learned the hard way. Now, like, are these kids not going to have to have any of that pain? Probably not, right? So I like using the IITuls because I know it's behind them and it's it's saving me time to do what I wanted to do. But what this new generation is it's something different has become more of a magic wand instead and I feel like it's, it's the magic wand now. It's wild.

Jon Radoff: Well, there was a day too when everyone had to code machine language and you practically had to be an electrical engineer to figure out how to code and then along with the pillars and people.

Guest: I never would have made it through that case. Yeah, I came, I was there already compilers. I write compilers. I know what goes into those, but I would not have written those in machine language, I wrote them in higher level languages. Yeah, never would have been able to do that.

Jon Radoff: The thing I'm most intrigued by with generative AI though isn't just the make more stuff or make stuff faster aspect. What happens when we bring generative AI into the game loop or in the process of making things or even music, I've imagined like is there a future form of music where you sort of create a certain amount of structure, but you have generative AI creating a versions of it that is a little bit different for almost every listener. Yeah, like there's a lot of interesting things and just in at the experience make that a feature. Yeah, that's the feature. So we'll make it hallucinate and if it's weird, it's weird and it's supposed to be weird.

Guest: The dreams are weird.

Jon Radoff: Yeah, they bring these 3D dream experiences to life. So that's an example of the generative AI being used in the play experience as opposed to like a production process enhancement. Wow, we could talk about that for hours, Tony, but and we've covered a lot of ground today seems like we're. Did we hit the hour where we're over by a minute, so just shows that we could keep going and we will keep going because we're going to be on this.

Guest: In one hour, we're going to do this on the panel with two other people.

Jon Radoff: It makes it up even more. It's going to be great. Yeah, but Tony, thanks so much for doing this with me really early in the morning itself by Southwest. But we're getting it out. We're getting it out of the way before all the events for later in the day. We stole the stand. We didn't we don't even officially have this by the way. We just decided to walk in and use it.

Guest: Tony, thanks a lot. Absolutely.

Jon Radoff: It's been a blast. Next week, we're going to be doing it live from GDC. So the Web 3 game developer live stream will be back. We're going to be doing it at the same exact time from GDC. We're going to talk to Dennis Diedack, who's worked on some really legendary games in Web 2, things like legacy of Kane. But now he's working on a Web 3 game studio called Apocalypse where they're using AI and blockchain and all of the cool stuff that we just referred to. So come back for that. And tomorrow we'll be talking to a couple of indie game developers in our traditional game development live stream, a sure lamb and Renee kittens where indie game developers for over a decade. They can tell you about their journey and how hard it is, but also how rewarding it is to be creative like this. So Tony, thanks a lot. This is super fun. Enjoy the rest of South by and we'll see all of you online. Thank you for joining us once again.

Unknown: I love that we got your.