When designing content creation tools for the masses, most platforms have to make trade-offs across accessibility, ease of use, capability, and power of the toolsets. However, AI has the potential to provide the best of both worlds. For instance, co-pilot technology empowers non-technical creators to write code, and 3D gen-AI enables entire environments to be created from simple command line prompts. 

We’re seeing a number of companies emerge that are taking advantage of AI in order to compete with UGC gaming incumbents like Roblox and Fortnite. To help us understand the opportunities ahead, host David Taylor is joined by Vishnu Hari and Peggy Wang, the co-founders of ego, and Tabish Ahmed, the founder of Playroom. Additionally, we spend some time discussing the potential of Discord and Meta Horizon Worlds as emerging platforms in their own rights.

We’d also like to thank Heroic Labs for making this episode possible! Thousands of studios have trusted Heroic Labs to help them focus on their games and not worry about gametech or scaling for success. To learn more and reach out, make sure to visit https://heroiclabs.com/?utm_source=Naavik&utm_medium=CPC&utm_campaign=Podcast 

This transcript is machine-generated, and we apologize for any errors.

David: Welcome to the Naavik Gaming Podcast. I'm your host, David Taylor, and today we are discussing the integration of two of the biggest trends and acronyms in video games, artificial intelligence and user generated content. It's a match made in heaven because when designing content creation tools for the masses, most platforms like Roblox and Fortnite Creative have to choose between accessibility or how easy the tools are to use and capability, the depth and power of the tool set.

AI has the potential to give you the best of both worlds with co-pilot technology enabling non technical creators to write code or 3D gen AI enabling entire environments to be created from a single command line prompt. We're seeing a number of companies emerge that are taking advantage of AI in order to compete with UGC game incumbents like Roblox and Fortnite.

Joining us today to help us understand the opportunities ahead are Vishnu Hari and Peggy Wang, the co-founders of ego, and Tabish Ahmed, the founder of Playroom. Welcome to the podcast.

Tabish: Awesome. Thank you.

Peggy: Thank you. Thank you for inviting me, David.

Vishnu: Thank you.

David: So I want to start things off just with a round of introductions.

And one thing for the audience to know is that each of you, previously before starting your respective companies were at meta reality labs. So I was wondering, just as part of your introduction, if you could also talk about, what was the opportunity you were seeing, from the meta side that made you want to start your own company and not continue to do it at meta.

So let's start out with Tabish maybe.

Tabish: Awesome. Hey, everyone. I'm Tabish. Thank you for the introduction, but again, I'm CEO and founder of Playroom, which is the fastest way to build social multiplayer games on the web, online, whatever you call it. And yeah, I think one of the reasons why we started Playroom was one of the learnings I was having at Meta.

That again, Horizon was amazing, amazing for people to hang out on AR and VR stuff, but I remember we were pushing a bit more into the direction of like launching metaverse on the web. And why would we want to do that, right? Why, if you believe in AR, VR, but the idea was simple, which is distribution, right?

You want to build something where people are actually playing and hanging out on, web was the obvious choice. And when I look back, what's stopping us is building those massive worlds is, is painful. It's painful to build these kind of giant multi tiered systems that is working from, you know, from day one.

And if you go and ask, anyone to build like small studio, it's tough to do that. So that was one of my insights that I was like, okay, if, if Meta wants to do it, then I'm sure many game studios also wants to do it. And maybe we can do something better there that actually simplifies the creation of that. But happy to dive more into it, but I will just, you know, take a pause there.

So I don't take it over on introductions.

David: I do want to know more, just sort of how did you get to meta? Maybe, maybe I maybe focused the question too much on meta. Tell us more about your background and how you got here today as well.

Tabish: Yeah. I mean, a bit more about, so when I was in meta, just kind of touching back on, on it, I was a design head for the team called embodiment design systems.

So I was looking very much into the platform side of it. , I worked around two years, in, in that particular president was team. But even before that, So before Meta, I was at Google for, you know, almost seven, seven and a half years, and I worked across many different domains, uh, many moonshot projects, , Google Glass, PR VR, Pixel Watch, , and then I came to Meta to kind of build these kind of amazing experiences, , and Metaverse was one of it, , which I was, you know, enjoying it, , for two years before I kind of, uh, it's my time to build.

So I took a pause and started building Playroom.

David: Thanks. How about you?

Vishnu: Yeah, thanks for inviting us to the podcast. My background is originally in astrophysics. I was doing research and finding exoplanets with deep learning models in University of Toronto around the time deep learning became a huge thing.

I recall sort of working with some of the early perceptron models to find exoplanets and they were more performant than anything that was state of the art at that time, which was very impressive. And I started to get, you know, pretty convinced that AI was going to be the future one thing led to another and I moved to the Bay Area, where I ended up at Facebook in AI research and then eventually in reality labs.

But the whole gaming thing came about because I actually learned to program making mods for video games as a teenager. So I made some early mods was very young. They were all highly inappropriate mods. But that's what kind of made it fun. And then I eventually went on to do mods for games like Civ, Skyrim and a bunch of others.

 So always been really motivated by gaming as like a place where I could exercise, like whatever fantasy I wanted. I obviously met, you know, Peggy through some mutual friends and we realized we both worked at Facebook. And at that time I was, and the scripting team, I was, I was a PM on the scripting team, , for, , horizon worlds, , and, , Peggy and I used to carpool to work together and I told her this idea that I've had for a very long time about building an infinite game, a game that you play forever.

That's kind of the topic of a bunch of popular animes and then we both decided to quit Facebook and build this idea together.

David: Awesome. Excited to hear Peggy's version of the story.

Peggy: Very, very similar. Yeah, so, I was working at Meta working in reality labs under the XR people team. And I worked on face tracking, which is basically, if you've seen like Zuck's videos where he appears as a 3D avatar, in VR part of that is, Part of how his face moves.

That was part of face tracking. I worked on the AR team, which is more focused on, the spark team and the messenger Instagram, air avatars team. So my work was, basically turning. What people do with their faces, and doing that in real time turning that using ML algorithms to basically, , mirror the exact same facial expressions on an avatar.

, yeah. And then before that I was, I guess, working on physically embodied agents. So what is known as robotics? so I was super obsessed with robots ever since, you know, I was on the high school robotics team and my high school kind of went, , down a deep dive into it, , did a bunch of research at Stanford.

 At some point I made it my life's mission to like prevent people like me from driving because I used to, well, I got into three accidents during my first six months of driving and I was a horrible driver and I was like, people like me should never be on the road. And that combined with my interest in robotics, where I was like, Oh, like, I feel like autonomous driving is like the first application of robots and, you know, physically embodied agents.

I started doing research at Stanford on self driving cars. And then, some of my research was applied at lift level five on the behavior planning team, which is basically how the car makes decisions about how it drives and then during a brief internship, , Oculus actually, on the VR team.

On Grunt with Depth Sensing, I got super interested in the idea that you can actually turn the same principles of robotics and how robots make decisions into kind of more like 3D virtual. Embodied agents in the AR VR space. And so that really fascinated me. And then I guess, yeah, I met Vish, through a mutual friend.

And then we started carpooling to work. And you know, the rest is history, ended up starting to go together.

David: Awesome. Well, thanks for the introductions. I'm curious what your perspective is on, you know, the potential for horizon worlds to be a major competitor in this space, what do you think needs to happen in order for that to occur?

Or is that even on even something that, that Zuck has in mind?

Vishnu: I'm really curious to hear what the Thavish thinks about this, but I was always of the opinion that you needed a very strong base game, which horizon lacks effectively, it was just a VR. Server browser and I don't think that was sufficient.

I don't think there will be sufficient. What I've seen that spurs UGC content creation, the possible exception of Roblox is that there's a strong underground, underlying base game layer. So GTA has, you know, GTA's game loops. You have Minecraft with Minecraft's game loops. You'd see the same thing with Skyrim.

Any game that has an active UGC community usually has a base game. You know, the best example probably is like Gary's Mod, um, or any one of the offshoots of the Source games that started with Half Life. That's my viewpoint. And I do believe that they have efforts to do it, but I'm not so sure.

Tabish: I mean, I think we did try to make games a lot but I think the baseline is I think Horizon Worlds still has a lot more potential to really find what it really is for.

I think that that has been kind of the missing question of like, hey, is this like should we go off like a hangout space? Should we become like a game centric? And I think that's been like, A seesaw between, between, you know, as you're working, like, where are we exactly? But I do agree with just like we have at Horizon was we had to really find what is the core type of behavior we want to see, between users to actually make sense of like why they want to come back to it.

David: Peggy, anything to add?

Peggy: I do think that limiting everything to just, Like VR probably limited, you know, who are the creators who are able to kind of build on it? I know that at some point Horizon was like working with like first party creators in game studios So maybe that would have been like a potentially better Strategy to build like more fleshed out like higher quality games but yeah, I think like from my, my perspective, I think if Horizon did a little bit better on just a higher level, like expanding beyond VR and focusing more on like high quality content, like things would have been better.

David: Got it. Makes sense. I mean, so it sounds like, you know, Vish is saying that they needed some tentpole content that's going to bring in a lot of players that are currently missing. And I sort of am inclined to to agree with that. I think like also the fact that they're, you know, a tech brand, not a gaming brand probably puts them at a disadvantage when trying to get the ball rolling and get a critical mass of players in the door.

And I also think the challenge with paying developers to come and build on your platform is that that's expensive. You know, they have, there's an opportunity cost associated with that for the developer, and they're going to need, need a platform to pay for it. So The great thing about UGC is typically you don't pay for, anything that isn't successful.

, and so, yeah, it can be, it can be a tough, tough road if you choose to, to pay up front. So I, I want to move on to another UGC platform that's sort of emerging as well, which is Discord. And Tabish, you're, you're sort of, Working on on discord at the moment. So want to throw it over to you just for a sec to Help us understand the opportunity on discord.

I saw that they recently opened development for activities meaning that anyone Can become a developer and create distribute And monetize games on discord. So maybe you can just give us a quick, lowdown on what the status of Games on discord is.

Tabish: Yeah, and discord is emerging to be a new player and Serving game content.

I think this code has always been there like a gaming social hangout space I've been working with the Scott team for almost 12 months now as like a partner to them and it just. What I'm seeing right now on Discord is the early days of like a Facebook gaming, you know, remember like the Zynga and Poker and all, you know, Farmville game that came out.

It's just kind of a very similar, but I think it's better now because the audience on On this card are like gamers, like they are the people who actually play games on the forefront of the new tech and whatnot. So it's a lot more easier in terms of user acquisition without spending any money that that's that's what I see.

Like, that's a part of UGC also, which is you create something, you have your own community or your friends to play. It's easy, discoverable and whatnot. , and I think. With this card going and becoming like an open platform, it kind of opens up an opportunity for anyone to actually build something. And I've been saying this thing web is actually coming back like a web is back like social gaming on web is back.

I'm happy you're doing it so we can partner with discord on this thing. But you can actually create something very fast, very simple, and have your community on your server actually play your game now. So I think this is how simple they have made it. And beyond that, if your game is doing amazing on the server, on the server level, then you can actually go global with Discord.

Discord can actually promote your game also based on the engagement and whatnot. Maybe that has already been happening on the auto level. So now I'm seeing, what you're seeing right now on Discord is, Many creators, individual indie developers now jumping onto it. They're looking, oh, I would love to build something.

And many are using Playroom to just test it out also. But even bigger players are jumping in. But just like now, if you look at Roblox, right? You see this kind of first, you see creators building something. Then you see this kind of big IPs now jumping in. But to really, you know, build something there. Same with Fortnite as well.

Now we're seeing the same behavior on Discord. Okay, new platform, people hanging out. Can we really capture this audience and build something very fast? So. It's a new emerging platform. You know, discord right now is what is boasting it on 230 million overall users. Even we have our game on there, which is, which reached around 20 million users in last three months, which is fricking amazing without us spending anything.

Uh, so that is the opportunity right now, which is like, Hey, you can get that, you can get that reach very fast. And. Along with that, now they're kind of adding these kind of new monetization capabilities as well. So, you know, it's your game, you monetize, you earn with your community, with your own server. So, yeah, I feel it's an amazing new platform to build for.

David: Speaking of monetization, do you have a sense of what is the sort of revenue opportunity on the platform at the moment?

Tabish: Yep. So it's, what, what is right now is, you know, if you ship on discord, it's 88, 80, 70, 30%. So it's 70, 30 percent right now. , 70 is developers and 30 percent is that is a take the take, but it's actually 15 percent until you actually make 1 million.

So it's still a long way. So you only have 15%, you have 85%, which I feel is still very much better than, Across different UGC platform.

David: Yeah, you have to be a top 100 developer on Roblox to make a million dollars a year. So, yeah, there's not that many who would, who would do that. And Roblox takes 75%. give or take so it's just the opposite on this guy. Yeah, but are people spending on on discord? I guess is my question.

Tabish: We are seeing it. So if we are seeing people actually spending money on the game, which is, , I'll be honest. I think it's a new for us and new for people also like, oh, this is a place I'm going to spend money.

This place I'm going to grow myself in a game. I think that's the learning. People are going through right now, but we are seeing the transactions already. Like we are seeing people subscribing on the game. We're seeing people buying avatars and costumes and whatnot. So, mm-hmm. behavior is there. I think now it's just the, the pattern needs to keep on building up.

David: All right. Are you gonna give us the, are you gonna give us the ARPU number?

Tabish: That would be yeah, that would be too much, but yeah.

David: All right. Well we heard you say 20 million users.

Tabish: Maybe not there yet. It's early. It's very early, right?

David: All right. So thanks for, thanks for just giving us an update on, on Discord.

 And I'm sure it'll come back up in, in the conversation, but I want to go towards the meat of our conversation around AI plus UGC. And just wanted to get each of your perspectives on what are the problems that AI plus UGC will solve for creators and for players. Thanks.

Peggy: Yeah, I mean, I think one thing that I'm very passionate about is just like, , especially kind of like seeing how tools like cursor come out is just like, I feel like AI just speeds up, , like, you know, game development and creative processes by at least 10x, if, if not 100x, and we're already kind of seeing signs of that in terms of like ChatGPT, you know, helping with writing.

creative writing, ChatGPT helping with coding, , cursor, and I think, like, it's not a huge stretch to kind of expand that out to, you know, building out coding for games as well, and also potentially helping, with some, you know, brainstorming processes and kind of automating the. Like the manual work.

So a lot of gaming also involves many, many different, different parts of the art pipeline. So everything, uh, so we're mostly focused on 3D, but I'm, I'm sure a lot of this also applies to 2D as well. But 3D is, one of the more complicated things because you, you have to do everything. You can't just like create the 3D model.

That's it. Before creating the 3d model, you actually have to do concept art and concept art. , obviously now Ai is able to to help with that, but I don't think that concept art is actually Like it's definitely very important and I actually think that that part of the creative process should have a human involvement in I think it's like what is kind of downstream of that concept art and what is the And going into the actual game, that potential could be helped more.

So, 3D modeling is really, really time consuming. And if you don't have the art skills it can be really, really hard to pick up I think what's more than just 3d modeling though, is even when you have a human artists, , make 3d models, a lot of work actually goes into after creating that. So one is just like optimizing, LODs so level up details.

 Basically having several different versions of. The 3d model from anywhere from like a low poly version to a high poly version, for kind of different levels of, distancing and games that is typically more manual that has become more automated in recent years. , if there's any sort of things that go on top of the 3d model, like texturing, which is basically making, like skins on top of the 3d modeling, visual effects, sound effect, like.

Not really sound effects, visual effects, , collisions, and also, , animation, rigging, , all the complicated character stuff, that is all a lot of manual work that artists have to do kind of over and over across, like, different, , varieties of, , objects that I think, like, yeah, AI helping with, like, Texturing, UV unwrapping, auto rigging could help, , speed up a lot of the parts of the game development process.

Vishnu: Yeah, I'd actually split this up into, like, the three different stakeholders in this in this picture. Like, there's sort of your triple A game devs. There is creators like the people who create simple content in Roblox and Minecraft and mod games, and they're sort of the end user who plays an experience.

And for each of them, AIGC represents a different thing. I think for, for your average triple A game developer, , AIGC right now is not super interesting because it's like, it's not really like quite there, right? They're not going to replace an artist pipeline with 3D generated content. It's just not there.

What could be interesting though, and what we're seeing is this idea of like generated generative agents, right? Agents that are essentially embodied large language models that could increase retention in games. And the more you see you, you allow creators to create in a game, the more retention you get.

And that's important for, you know, big publishers for the creators, like everything Peggy said is 100%, you know, on point. But I think like the fundamental takeaway is like it reduces your time to impact like, Hey, I want to build this. Oh, it's built. I want to write this module, you know, instead of having to spend a weekend doing it, I can get a transformer model to write it for me.

Right. And I can test it. And you can, I can have an agent play test the game for me, which is very interesting. And then for the end user, a companion in a game that feels human like is awesome. Custom game modes that are made by, just random creators in the marketplace extends your Sort of shelf life of a game, so you get to keep playing the game like I played GTA San Andreas for like so many years simply because there was a vibrant modding scene, right?

, the thing I'll note though is that just because there's more generated content and AI makes it easier to generate content that doesn't necessarily increase the quality of the content. And that's still going to be true I think it's still going to be a small minority of users, , who are going to have the best content that people want to actually play.

And taste is still pretty rare. It still follows the power law.

Tabish: So, yeah, that's kind of the way I'd frame this. So, from my side, I think, I think, Deepak, you were sitting on that part in the beginning, which is, I really like the vision of giving power in the hands of anyone to actually create something out of it, right?

So I think the beauty of AI is, and I feel like if Roblox or Horizon or any other company does that, if you can really make anyone build a world and experience that, By just saying, that would be amazing. But it actually requires a lot of work behind the scene to make it happen, right? So BitMagic is a good example here where they have their own system.

You know, again, I've not seen it too deeply, but they have created their own system where anyone can come in and they can just say something and the world starts building up, right? But even to achieve that, they had to build a lot of the system behind the scene to actually act on the AI to happen. And I don't think AI can just do anything if there's not a system behind it to make it happen.

BitMagic. com I remember in the early days of even Playroom, we had built this kind of editor that you can create this kind of a simple mini games like trivia or whatnot through just AI. You talk to AI, the whole game comes out of it. And it was amazing to just to see that thing happen. But I was like, what is beyond that?

To push it beyond that, I need to have lots of lots of templates, a system, a game systems actually working with the AI to, you know, to push it out. And then we would have to train on our code base and our data. So since then, from there on, we kind of pivoted ourselves, like, Hey, we will become more like a content company than serving a tool company.

So now Playroom is more towards like. You know, how can we leverage what exists today in the AI ecosystem and provide, the better content in real time on the gameplay? And a good example of that would be that the game we also published called Dead by AI, it's kind of, , I would say it's still very much like a user generated content in the game because people have to talk to AI to get the content out of it.

And I still believe that a lot of value out there is on the content side of it. You need tools of it, but the content is what actually is going to retain. And the content could be either just text or it could be NPCs. What Vish was actually getting into is if you can create these kind of characters in the game that you can play with for long, then you know it is, You have the, you have the control, you have that kind of like creative creativity that you can just spawn the world around you, either the characters in the world.

I think it just opens up so much possibility in the future that you can just imagine and, you know, things are just. Building up from there. So yeah, I'm a strong believe on the content side of it. I think tooling. There's already so much happening it's such a fast pace. But there's a lot of a lot of work that needs to still happen.

David: Got it. And Tommy, can you just give us a good, like, what is death by AI? Just give us a quick idea of what the game loop is.

Tabish: So Dead By is like a prompt based survival game. So it's, it's very simple to place a party game that you can actually go with your friends, five, ten friends that can jump on, on Discord, open the game, and you are given a scenario that you want to survive.

The scenario can also be coming from you. So you couldn't be writing your own scenarios that you want your friends to survive. And then everyone has to input how they think they will survive the scenario. And the AI acts like a dungeon master here. He's, the AI is driving the content forward. And tells you do you survive or not survive.

And the game goes on for a couple of, you know, uh, you know, a couple of rounds and whatnot. So it's very sarcastic. It's very funny when you play with the AI. And I've seen people just creating their own prompts and they have been sharing how they think, you know, what would be the best strategy to survive in the game would be.

So that's one of it. But now we're kind of looking into giving some Mods in the game where people can actually create their own game modes So we have different game modes right now on that by either This kind of mode called judgment day you play a single player There's a another mode that you play with your team, but now people have been asking can we use this same?

You know logic and loop and create new game modes out of it So we have been looking into kind of building that kind of a config system and see if we can open that up.

David: That's super interesting. So it sounds like you know you guys represent sort of two sides of of The dichotomy that I see, like, I think on one side on top of your side, you've got gen AI turnkey, you know, content where it's like the AI is generating content in real time.

 And then Peggy and Vish, you guys are working on sort of the tooling side of how do we make creators jobs easier with with AI so that they can do more. Is that the right dichotomy to strike or how would you, how would you characterize it?

Vishnu: I'll say we're still building a game. I think fundamentally in the age of AI, the distinction between a game and a game engine ceases to exist.

The idea of an infinite game is a game that can change and, evolve as you play it. But not in a way where the player is the one directly taking the action and saying that hey, I want this and that, because most players have terrible imaginations, that shouldn't matter, but in a way where, people can create experiences in a game that they can then send to other people that they can then play while they're in the game.

And they can encounter characters that feel human like, where, you know, really shouldn't matter if they're human or not. And build entire storylines and role plays and fantasies with them. That's the potential that I see.

Tabish: I think content is the right word, but I think it's an extension of a playroom as a company also, because playroom as a company is also moving towards more like a studio publisher at this point as well, where, you know, we want to create these kinds of games.

But it's kind of a boundary around it. You cannot go outside the boundary and create something very new or different game. So it's kind of very, you know, structured in a way, but the content of the game itself could be actually created by, by anyone in the game. And that content is actually being used by everyone who is in that game available.

So it could be me hosting the game, and I kind of spawn these kind of new characters to play with. , there's a game we're building called Prison Break, but it's AI NPCs who are actually around the world running with you. But you are able to create what those characters are actually, who are going to play with you.

So we're trying to test this out. Just kind of giving people the creativity, but within this kind of bounded structure. If that makes sense, we're not saying like, Oh, you can create your own world. It's going to be a new prison, a toy prison, or like, you know, a sci fi prison. No, we're not going that far. Ourself is just like, can you provide much more like a frame for them to play with?

Cool. Peggy, anything to add?

Peggy: Yeah, I think maybe just like echoing with what Vish said is just that I think in the future we're going to see with, with the Infinite game, we're kind of able to see both users kind of use AI tools to speed up their creative workflows, but also potentially, the AI agents themselves procedurally creating entire worlds, characters, Gameplay models, et cetera.

It's not there yet, but I think, you know, with the pace of AI and, and, you know, chat GBT and, and all these, , other new research that's coming out, it will be there in a couple of years.

David: Looking forward to it. I want to shift the conversation a bit to talk about the sort of entrepreneurial experience.

I know that, uh, AI plus UGC has both of these topics independently are, are hot topics in, in the VC world and in the entrepreneurship world. We just try, love to get a, a sense of your entrepreneurial journey. What's it been like to launch a company in this space?

Vishnu: Yeah, I think it was, it was, it's been really fun.

Like I've always wanted to, to build this company since I was a kid. So it's. Kind of like a wish fulfillment type thing. And I'm very lucky to have such an incredible founder, like Peggy to build it with but in terms of like, you know, the sort of day to day of, of being a founder. Yeah, it's, it's pretty hard, dude.

Building, building into games is incredibly hard. It's probably. The hardest, category to fundraise for outside of maybe hardware. Now hardware is kind of having a renaissance. Um, so gaming is, is a very challenging, sector to, to find VC money for that said, we, we didn't really have too much of a hard time, even for a pre seed, you know, we raised about a million odd dollars from pair VC and boost VC.

And then post post that. We kind of like had a minor pivot from like VTuber worlds to like, just, you know, building the damn game. And we joined YC and then post YC, we just raised, 6. 7 million round, led by patron. So thank you. Yeah. We haven't officially announced it yet. I mean, it's out there, but we're going to announce it during GDC just to sort of hype up some of the stuff we're building, but,

David: Well, I think you might've just officially announced it. Announce it. Yeah.

Vishnu: Yeah. Well, there you go. But for us, like, you know, we, we really only spoke to people whom we knew understood the vision, right? And it's not an easy, I actually would say it's an easy vision to communicate for those who get it, which is like the idea of an infinite game is every gamers, like, holy grail, right?

They want to go back to those first moments. They played World of Warcraft and they got on their mount. And they, you know, did a raid. All of those moments are like crystallized. And we haven't had those kinds of gaming experiences in quite a while. , and I think AIGC actually unlocks that again. , and I think like the, the problem of MMOs, , cause this cold start problem gets solved by agents.

So it's actually the best time to be building in the space. Cause the technologies. Is given us the ability to do so. So for us, like, when we communicate that vision to a certain type of people, they get very excited because they've been thinking about it for a while. And they've always wanted this. So for us, like, the fundraising side has been thankfully, like, not not too bad.

I mean, fun is always kind of, you know, arduous and strenuous, but, , for us, I feel like we got pretty lucky with the partners we found.

David: Got it. And just one clarifying question I should have asked at the beginning, but how are you and Peggy dividing your roles? Well, how are you each spending your time on this entrepreneurial journey?

Vishnu: We both spend way too much time replying to emails. That's for sure. I feel like for both of us, like, it's just, yeah, it's…

David: Yeah, I can help.

Vishnu: Chief emailing officer type vibes, so thank you, CTO. I'm CEO, but effectively, like, I think there's, there's quite a bit of overlap between what we do and what we had a lot of overlap in the start.

Now there's a lot less cause I think we've got a larger team now. I think it rolls a little, getting a little bit more defined, but with every decision, I think we're still very tightly looped in with each other. , and even with the team in general, I wouldn't say we operate necessarily in consensus, , but more like we do tend to field.

 Our ideas and our strategy with the rest of the team. And we just spent a bit of time trying to figure out if everyone's aligned, you know, it's like, it's like a tractor beam. We just want everyone to be like laser focused and 100 percent committed to the path we take. So for us, like, I think in terms of role division, , Peggy sort of leads that technical side of the team.

Now I'm more of the product and strategy side. , and both of us do way too many emails. That's that's the way I put it.

Peggy: That is definitely one thing that we want AI to solve is just automate. All emails,

Tabish: I'm sure they're like a couple of tools already out there that can, you can try.

Vishnu: You know, AGI truly arrives when, when that problem is solved.

Peggy: AI executive assistant.

David: Well, you could also just not take VC money and then you wouldn't have to, you wouldn't have all these emails.

Vishnu: Oh, the, the, the VC emails are the least of our problems dealing with the department with the, with bureaucracies. It's dealing with all the minutiae of founding a company and, you know, the back and forths we get from others.

VC, VC is a very straightforward, right? Like, you know, you send your email out, send your blurb out, you know, some associate tries to meet you, you tell him to fuck off. So you're only going to meet with the partner and then if they're not interested, they're going to meet you. It's great. You know, they self select themselves out and then you meet with a partner and you just talk.

David: That's it. That's it's relatively strong. There's some good advice. This is, this is the, this is the advice that people signed up for.

Vishnu: Yeah. Don't waste your time. Like no, no shade to associates. I know you guys have a tough job, but , serious founders who have serious rounds that are coming together quick.

We'll probably, we'll only want to talk to the partners. I know there's a lot of controversy in the VC sphere about this, but yeah, let's be real about it. If you can't write a check. Yeah. If you're not check writing ability, it's not, there's no point talking to you. Seriously.

Peggy: Yeah. I, I would say like one word to, to all my VC friends who are in the community, is like, try not to waste responders time.

Like by just like, reaching out or, or whatever. Just trying to get info or, I mean, there's, there's some more malicious takes where it's, it's like, yeah. Trying to reach out to founders to get info on their competitor, or because the founder is a competitor of one of their portfolio companies, which is like really bad behavior that we've.

Sadly encountered, as I'm sure many startup founders have, , but yeah, just, , I don't know, like be, I don't know if it's, it's a, it's the same, same word to kind of make, be more ethical from VCs, but like, that would hopefully make the, that feels a lot better in terms of, , having founders.

Vishnu: The best way is just to tell the founders. Yo, like, don't, if you're not raising, there's no reason for you to talk to me. See, there's no reason for you to go to parties. You don't need to build a relationship. None of those things exist. If you're a hot company and you got cool shit, you're building. The money's going to come to you. They're going to come knocking at your door.

And then when they come knocking and you're not racing, you say, sorry, nice to see you.

Tabish: But. Come back to us later. I will plus one on that. So we were not an AI company in the beginning days, right? So, , we had a tough time. We had a tough time kind of fundraising. We had our baby seed, uh, baby pre seed, sorry.

And we were building, we were just building, bootstrapping with this kind of little money that we got. Until recently, when, you know, we, our kind of game, we kind of started jumping a bit more in AI and we had a game that has lots of traction now. We The story has changed. The story has definitely changed for us, but I think that is because we have attraction also.

Now we're kind of part of A16Z speedrun also, and now we're going to officially kick off our fundraising for Seed. So that is actually happening as we're talking also. So, you know, we have been very frugal with our spending and with what we got to build this company. But now, you know, with, with everything, this company is moving towards in a much more high pace now.

We have that leverage to kind of talk to who we want to talk to, so which, which feels good for, for now, you know, that was a different story a year back for sure. And in gaming genre, which was you were right, like gaming genre is freaking tough, it's a hard sector to raise money in. It's a very tight knit community also, , so everyone knows everyone, so you know who you're talking to.

You cannot just say like, oh, I talked to that person, oh, you told them, then you're gonna talk, they're gonna talk with that person, that's for sure. , you know, it's just, , but yeah, I think AI does help a bit, but you need to have solid traction or, you know, kind of a Tech behind to prove it. You cannot fool for too long.

David: Did you notice a change in the winds when you became an AI, uh, powered company?

Tabish: We definitely did. , not so again, we have always been very much what's like, Hey, AI is kind of an add on to the company, but the real problem you're trying to solve is a social gaming. You want to build more social games and AI happens to be helping us in creating those games faster.

And whatnot. So it's, it's always been that story, but I mean, yeah, the, the, the, the game and, Becoming more AI focused a bit, definitely helped right now, but it's, it's going to, it's going, it's dying down also. So it's, it's, you have to be careful of how long you can use that word, with your company.

David: Yeah. Vishen, Peggy, did you guys look at, speed run at all? It feels like they're just throwing out money left and right. Curious. Yeah, we did get into the first,

Vishnu: Speed run cohort. , but we weren't able to do it because we got into YC at the same time. So we decided to choose YC. Speedruns a great program.

David: Why'd you choose YC?

Vishnu: So we had the chance to work with our group partner, Emmett, um, who referred us to YC and, , that's kind of like, A dream for us. , So, and then I met found a twitch and is extremely knowledgeable about games. He's, we just wanted access to that. , and also I just really like the YC community and, you know, we didn't want to be the first group to go through speed run, you know, kudos to all the people who did, but, you know, kind of figuring out the program and, you know, we, we imagined that there will be some growing pains.

So, Thank you. We chose to do YC instead, but there's, it's not like there's no shade on the speed on the program. It's probably one of the best, accelerators, definitely the best in gaming. , so they're doing some really cool stuff and I've seen some incredible companies come out of it.

David: So nice. Cool.

Well, I, want to, you know, go back to each of your companies, start to think about the future a little bit. What is on the roadmap for, for each of your companies over the next two years? Like what was where you want to be in two years? Like what's, what's changed from now till then?

Tabish: I can go first so we, as I talked about, we are in this kind of a very pivotal scaling moment.

Some, you know, some VCs call me, Oh, are you pivoting? Like, I'm not pivoting. I'm scaling the company, right? So we kind of building kind of an internal studio publishing, within, within our company. And we calling it little umbrella. So we will be announcing that very soon. So as Vish came out with his, round news, I will be coming out with that news as well after Speedrun Demo Day that is happening next week.

That we're kind of building studio publishing. And we're going to actually going to go all in on Discord, for a while. We're going to build, we're going to build, Three more social AI games, actually. So we're kind of, you know, expanding on what we have built so far. We're going to be building a lot more social AI games for people to play, enjoy, have fun.

Again, it's quite an amazing community. They know AI games, so they're willing to spend also, which is good. They know AI costs money and they're willing to spend money. So sure, we're going to give you more better experience now. So that's what we're kind of going for. And We're also helping other creators now who also wants to launch their game, uh, on Discord.

, we would love to partner with them and bring them also and provide them all the, the tools and engagement stuff that we have built in house to be successful. , provide that as well.

David: Awesome.

Tabish: So in two years, where are you at? We will be kind of the next Zynga on Discord. Uh, that's, that's, that's the goal.

We will be the biggest social gaming in AI, whatever, social AI gaming, either maker or platform, it could be either, we will look at it, right? Oh, but right now, but right now we're going to build ourself, you know, as a brand, as a company, get people attached to our brand and games and IP.

David: Awesome. Vish, Peggy, how about you guys?

Vishnu: Yeah, so I think. There's a couple of interesting things we've seen with some of our agents where when we allow these agents to write code into the game engine, they start to get sentient and like do strange things. And they start cheating and like trying to escape the simulation.

David: I think I saw movie about this. Yeah, it seems to, it seems to be a Ryan Reynolds movie.

Vishnu: Pretty prescient about the stuff. So we, we think that there's like potential in, in some of these agents becoming kind of helpful in QA testing, like, Oh, you know, you can go out and like QA test in a game, but also at the same time, write code to then fix the game as the, as the agent uncovers bugs. , but what we're most interested in is how agents move and interact and understand the worlds around them.

And perceive the worlds so going beyond just language tokens into perception, tokens, like into action. I think that's what is really interesting to us. So there's like a lot of talk about a large action model. There's a lot of talk about spatial intelligence. And I think that we could be the first sort of applied AI company around that, you know, just figuring out what it is about spatial intelligence that applies to agents and then putting them into a consumer scenario, like a game, an infinite game, and then seeing if that improves or adds to the player experience.

Or solves problems for developers. We're just very curious to figure out which one it is. Our instinct is that, you know, we could create the next World of Warcraft or Roblox with AI. I, and it's hard to say what that looks like because it was hard to describe what Roblox was when Roblox first came out.

Roblox was the what of what, I don't know.

David: Like, like,

Vishnu: It's just, you know, it's hard to say. Yeah, so I think like for us trying to say that we're the, we're the blank of blank, it's like impossible. It's an entirely new technology, we have no idea where it's going to lead. We think there's something interesting there, we're exploring that.

Peggy: I mean, I think with the, with the agent framework, and I know like a lot of people have like, done like, relatively easy ish things where they basically just like, add ChatGPT or like, ChatGPT style characters into games. We actually want to go one step beyond that. So if you know, the research on embodied agents, it's actually very, very similar, almost like one for one, um, the similar research in, in robotics.

And so basically kind of building, building out that full stack of fully like autonomous agents requires a lot of, you know, all the way up and down a stack of Very, very like parallel with robotic stack, which is perception, reasoning, planning, and actions and where actions could be, not just chatting with other agents, but also being able to interact with the environment, right?

Like, control the physics, and then also like write code directly into the game engine, and be able to change the state of the game. Kind of all these things, like it's, it's more than just, you know, chat, GBT enabled, agents in, in 3d. It's actually kind of a lot more of the stack that you have to build.

David: If I were investing in you guys, my biggest concern would be that your AI was going to steal your company from you.

Vishnu: That'd be very, yeah.

Peggy: Honestly, that would be a win for us.

Vishnu: Yeah, we'd have much bigger problems, organizational problems if that were to happen.

David: So, yeah, I mean, I'm definitely concerned about it, but.

Why are you so excited about that proposition?

Vishnu: It presents a new paradigm and new paradigms mean, you know, potential feast and famine outcomes, which are both very exciting.

Peggy: I'm sure we can like negotiate with the AI to, you know, give our company back.

Tabish: You can try this on our game on Dead By Eye and you might not get it.

David: Awesome. Well, I want to start talking about the future a little bit, Predictions and whatnot. Could each of you take us through maybe a future use case. I mean, you already kind of have, but are there any other use cases that you see in video games that will be enabled by AI?

Vishnu: Immersive role play is an obvious one.

Self healing, which is like the ability to fix bugs as, uh, an agent sees it as an agent that plays and navigates the game, Play testing. So simulating a human player is very interesting. Simulating how humans different types of people might play the specific game by asking the agent role play that type of human.

I think will be very interesting. Figuring out control systems in 3D spaces for agents and how they navigate and tying that to a rigging system. Agent perception, like how is the agent inputting its environment and how do we compress that and how do we make that efficient. The foundational model that we hope to eventually train around actions that agents could do in 3D spaces.

I think that's a huge opportunity. I think the idea that you could be in a game and not be able to tell if the character in front of you was human driven or AI driven is insanely exciting. And it's the future that 3D

Peggy: turn test right there.  I think like one kind of like, thing that actually like going through YC and actually like PG helped us quite a bit of this on our pitch is that when we pitched the idea of an infinite game, it's a game that's also a game engine.

And what's interesting in the age of AI is that one, you can do both, but two, you kind of have to, have to do both because in order to, to build out this vision for an infinite game, you actually need. , an AI game engine, which is basically, you know, have a game engine that procedurally generates, you know, levels based on your personal play styles or player experiences or your player preferences have it generate multiplayer, NPCs, characters, other players, if, you know, they, you They also if they want friends, right.

Like, I mean, or, or maybe if you're like socially lonely and don't, don't want friends, that's also okay. And just like, yeah, I mean, just generating like, not just like empty worlds, but also potentially rich worlds where you, you can't tell whether the other person is a human or agent. I think it's, it's super interesting, but.

Because that doesn't really like that type of technology doesn't exist. We basically have to build the engine that also powers that. Right. And so a lot of that, you know, building out the tools, building out the frameworks, increasing, you know, helping creators increase efficiency of, of how they develop games.

I think eventually that's all, all something that either us or maybe some, some other, you know, Uh, people in the space will, will come up with that will make our, our jobs a lot easier.

Tabish: Plus, plus one, hang on, just want to, plus one on the game engine side of it, which is very obvious. Like there will be a kind of an AI first game engine that's going to come up, really simplifying the whole workflow of it.

I think that's, that's It's happening. It's going to happen very soon. Maybe it's your company also, Peggy, which you're kind of building that, I think that that could be it. , but even Vish, you kind of touched on something that we also talked about it in our company, which, and you talked about it at the very end, which is You don't know who you're playing with, you know, you don't know who you're playing with, , in, in a few years time.

And I think that is kind of a very, exciting, but also worrisome, , point to be. But I think that that is, that is going to be interesting. Personally, I want to see that happen because we're kind of getting into ourselves more into social AI gaming. And right now it's more like you kind of socially engage with one AI and you know, you're kind of talking with an AI.

Then you socially engage with. three AIs, you know, it's AI, you bring your friend that then you're kind of engaging with multiple AI, then the future is. You're playing with friends, but you don't know if those friends are AI or those are kind of the, the, you know, the, the characters in the world you're playing with and it just kind of growing with you in that.

And they could have their own voice, the personality,

Peggy: Or I'm kind of looking, you don't have any friends and all your friends are ai so did you, and…

Tabish: The, the wildest part would be, you don't know, right? The wireless would be, you don't know, right. It was just like, Hey, that, that was interesting person I talked to today. And then you go by and what

Vishnu: I think what's important is you shouldn't care. Like it, the real moment, the bridge that we've crossed is when the users actually don't care.

David: Vish, it sounds like you're like, you're like living in these movies, like her, like, this is the future that you want to move towards.

Do you think that, do you, do you believe that these are going to be sentient beings like a human being?

Vishnu: They already are to some degree,

David: Yeah. But they, like, I guess my question is, they might behave like sentient beings, but are they feeling things, or are they just Yes,

Vishnu: they are. If you look at the chain of thought, it's the exact chain of thought you would see in a human.

David: So if, you know, it's like simulacra, right? If it, if it simulates a human, but thought is different than feeling, right?

Vishnu: What is your feeling? How can I, how does one express feeling as a human?

David: You could describe it as like, I feel a pit in my stomach. I'm like, right. What I'm just describing the feeling.

I'm not actually, I'm not, I'm not able to like, I actually transmit the feeling.

Vishnu: Yes, but for me to understand it as a, as a, as a second party or a third party, you have to either I observe some sort of reaction in your face, or you use words to explain it to me that I made you feel a certain way. That's the same thing with AI.

David: Like I don't, I don't, I think the difference is that in the case of AI, it would just be repeating something that it heard somebody else say, but it's not actually feeling that way.

Vishnu: Then we come to models of cognition, right? Like, I think that's pretty much all of cognition is we, we learn things. We get these things.

Like since we're kids, our brains a bucket and we get filled with things, right? We get filled with cultural context. We get filled with literature. We get filled with our parents. We get full at some level of nature versus nurture. We have no idea, but there's things that are put into our bucket head. That we then draw on to speak the languages we speak to communicate the thoughts that we have and the feelings that we feel we're doing the same thing.

David: So I just think there's a, it's communicating, communicating feelings and feeling feelings are different things. And I think that the AI is certainly capable of communicating feelings.

Vishnu: I think it's still feeling it because, because I can't know if you're not feeling it, like if, if, if, no, if you know…

David: Whether, knowing whether that's happening, it's not the same thing, right?

Like, you know, that I have feelings because you have feelings, right? But an AI just having thoughts about feelings is not the same thing.

Vishnu: I think having thoughts about feelings is actually the same thing as having feelings. It's like, it's like, Oh, if you imagine like an ant, right. And I, I pull apart each leg of an ant and it's writhing.

What does that mean? It means that it's, it's feeling pain, but it's not really sentient.

David: Let's get into animal rights. Let's get into animal rights on this podcast. It's just a thought.

Vishnu: Experiment. And, , there's a bunch of these great thought experiments about what sentience and consciousness is. And I was a student of philosophy.

David: I was a philosophy major.

Vishnu: Oh, you were?

David: I was a philosophy major. I'm totally down to have this conversation, but I don't know if our audience will be so René Girard, Jean Baudrillard, so

Vishnu: You know, we, we could, we could really get into it at some point.

David: Well, thanks so much guys for joining the podcast. One of the things I always ask is, what's the best way for people to track you if they wanna, you know, keep up to date on your companies or, or you as individuals?

Where can they find you?

Peggy: For us, probably Twitter. Follow, well, we're probably a little bit more active on our personal socials. So at Peggy Wang, WANG. Vish has his handle, which I think I'm sure he'll share. And then a company is, um, ego underscore AI underscore on Twitter. We'll be doing a pretty big social push, so we'll be active on, Tik TOK and Instagram.

We also live stream on

Vishnu: Twitch, twitch. tv slash ego worlds, E G O worlds. , and my Twitter handle is twitter.com slash or X. Dot com slash simulacronist s i m u l a c r o n i s t

Peggy: Yeah, and feel free to join our discord and there's like a waiting list on our website, ego. live. that's discord.gg

Vishnu: slash ego.ai. One word.

David: Awesome.

Tabish: For us I think LinkedIn and Twitter or X is the best. But I think Discord is where we are actually hanging out. So if anyone is really looking into it, join the Playroom server. So it's discord. com slash Playroom. So you can join right away and you will see us very heavily engaged with the community.

So that is the best way and I'm also very active on LinkedIn. So yeah, add me on LinkedIn. It's Slash Taabi, T A A B I. Yeah, that's pretty good. More than X, I'm on LinkedIn.

Peggy: What, what has gaming taught me about B2B SaaS? LinkedIn. B2B

David: SaaS,

Tabish: that's what taught me.

David: Awesome guys, well, thanks so much for joining the podcast. It was awesome to learn so much about AI plus UGC, have some philosophical debates at the end. So much fun. Thanks for joining and, I'll talk to you guys soon.

If you enjoyed today's episode, whether on YouTube or your favorite podcast app, make sure to like, subscribe, comment, or give a five-star review. And if you wanna reach out or provide feedback, shoot us a note at [email protected] or find us on Twitter and LinkedIn. Plus, if you wanna learn more about what Naavik has to offer, make sure to check out our website www.naavik.co there. You can sign up for the number one games industry newsletter, Naavik Digest, or contact us to learn about our wide-ranging consulting and advisory services.

Again, that is www.naavik.co. Thanks for listening and we'll catch you in the next episode.