--:-- Josh Liu: I love the games industry. Speedrun is a love letter to the games industry. The games industry is kind of a special beast. Games at the end of the day are experiences that people enjoy. They evoke emotion. They make people feel something. And so we wanted to build a program that basically embraced the founders who were really, really passionate about making games, as well as founders who were passionate about helping people make games and interactive experiences. And so that's the genesis of Speedrun.
--:-- Host: In this episode, you'll get to hear from Josh, but also A16Z general partner, Andrew Chen. Over 3,700 companies applied. We didn't know whether something like this could work. We've just seen the quality of the teams and what everyone's working on just skyrocket. No one had ever built an accelerator for the games industry before. We wanted to just support those types of founders in a way that they'd never been supported before. But in addition to hearing from the A16Z games team, you'll also hear directly from some of the most interesting founders in the last Speedrun cohort. So what did they build? And what does that tell us about what lies ahead in gaming? An industry that I should note that despite being continuously overlooked, has often speedrun other industries in terms of innovation.
--:-- Josh Liu: The ideas are just from every nook and cranny of the games industry. We have folks that are working on game studios and using AI to reinvent the fundamental player interaction that people have within games. You know, you have AI generating storylines. All the NPCs and characters in the games are powered by AI. That's fantastic. We're seeing a bunch of tooling and infrastructure that's being reinvented as well. One of the first things that's going to happen is you have AI that's revolutionized text interaction and then 2D. And then there's a lot that's going to happen in 3D and animation. And it's very exciting, just all the new platforms that are emerging and all the startups that are coming along with it.
--:-- Host: So let's give you a glimpse into the very founders at the heart of this program and the frontier of the industry. Whether it's companies creating completely new game mechanics by matching their velocity with culture or companies focused on the newest of generations, that is Gen Alpha, of course, or companies leveraging new technologies like AI to create completely new paradigms.
--:-- [Disclaimer about informational purposes only.]
--:-- Host: We're building digital human beings. We're a company with a mission of creating AI souls. We'll get there. But first, let's start out by learning from the deep wisdom in analog board games, many of which have remained in our hearts and our minds for thousands of years. Let's start with Avi Romanov, who's building Magic Circle. Magic Circle is building the Netflix of board games. Essentially, what that means is we're taking all these great play patterns that people have become familiar with, like social deduction and games that get a group of people around a table to have fun. And we're bringing them into the age of AI and digital.
--:-- Avi Romanov: So we have a website that you can access on your phone. So you could be at a bar, you could be at a dorm. We built basically a super light game engine for the browser that lets us make these games instantly shareable so you can play them in real life. There is a rich history of founders taking beloved games and updating them to be modern and digital. Just take chess.com, for example, which gets an estimated 200 million visits per month. Our strategy is like update the classics. So there's some really tried and true patterns that people love like social deduction, cards against humanity style things. And they're just like almost comfort food of hanging out. And so we're taking those games and just upgrading them with AI.
--:-- Host: But this is not as easy as it may sound. One of the most important things with games is, is your game fun? So there's a lot of like accoutrements that you can build in your product, like platform features and all that kind of stuff. The hardest thing always in games, no matter what anybody says, I think, is just finding the fun. That's true forever. And I think in some cases, like we've seen games where we're like, is this game really doing it? And we're like going to kill it. And then we see somebody play it and they play it for an hour and they tell their friends and we're like, oh, OK. People talk about like product market fit or founder market fit. We need like game group fit. What are the groups and the environments that people are poised to play those games?
--:-- Avi Romanov: Like we've tried all kinds of wacky games before we started working on what we were doing right now. We had a game that was like Google Calendar, but Tetris, and it was like a group vacation planning game. We had a game where like you put your phones together on a table and you could connect them into one big screen and like rats would crawl between them. And you tried to like move the phones to hide the cheese. We did all kinds of crazy stuff, but it turns out what people really want is like tried and true, but better. So how do you figure out if a game is fun? I mean, we tend to, and we're certainly not unique in this, we typically always start our games on paper, like doing paper prototypes. Most of the core loops we sometimes talk about of games or play patterns are coming from board games.
--:-- Host: I would say there's kind of cascading levels, right? One is just, does this game even work? Doesn't even make sense. And a lot of games don't even pass that. Then it's like, okay, this game works, but would people play it without us in the room, right? And then there's another question of like, would they bring their friends into it? And then also, how long do they play it? The other thing is like, sometimes I think about this as games as a menu of food. Who's to say that like crudo or like something that's like you eat as an appetizer is better than a dessert? They're just different. And so for us, it's like we're trying to have all the food that is like the thing that I think Netflix did incredibly well is they took this content that people have loved for a very long time, but was limited by the distribution vehicle.
--:-- Avi Romanov: People love watching TV. They love watching it in their house. But it was being held back by the fact that you had to go to Blockbuster and get the movie. and their whole company has basically been about evolving it from this proven, you can think of it as like a watch pattern. For us, it's a play pattern. And we're trying to do the same thing with this play pattern. Like this is a thing that millions and millions of people do. There's like 1.5 billion board gamers in the world, which is like a crazy stat. And so we're basically trying to say this play pattern is real and interesting and refreshing because it's principally about getting to know your friends better and interacting with them often in the real world, as opposed to like being immersed in a 3D digital world where it's like about leaving the world and not about like embracing it.
--:-- Host: And of course, new technologies like AI make it arguably more interesting than ever to bring some new life into old game mechanics. Maybe it was Andrew Chan had this point about like there's soft AI and strong form AI or like weak or strong form as it relates to games. And the weak form of AI in games is just like using AI to accelerate the process of making the game but there no AI in the game And then there's a strong form AI which is like actually your game core loop depends on AI And as soon as ChatGPT came out I like got it for everybody in the company It crazy It just makes you way faster. I still use it like three times a day. So we've always been doing that. But what we started realizing playing with this more is that the role that AI can play for our games, because there's no 3D graphics. The fun is actually in not looking at your phone and looking at the other people around the table. Like the fun is, that's what the magic circle is. It's outside the phone, it's like in the room. And so for us, like the role of AI is actually in personalizing them and personalizing the content. It's not like generating new content de novo. It's actually just taking the content that you are saying out loud, like in that trivia game and amplifying it and kind of like automating the annoying things about like making your own content packs in games.
--:-- Josh Liu: We've started leaning a lot more in that direction. AI is clearly not only reshaping the landscape in terms of the types of games that can be built, but also how quickly they can be built. Here's Josh again. Maybe the most common piece of advice that we give to a founder, and it sounds really obvious because we're an accelerator, but it is to just build and to build fast and to not stop building. At the end of the day, when you're building a product, even if you have really great intuition about a space, even if you have really good signal about what's going to work and what's not, you still have to do it and you still have to learn. And that's the amazing thing about the games industry is sometimes it's so unintuitive what works. That's kind of the nature of innovation and especially at the intersection of tech and entertainment. It's the unintuitive things and you only run into those serendipitous unintuitive things if you go out with intention and are just building really, really fast. And so we call our program Speedrun. Unironically, it is all about speed. And it's about speed with intention, of course. But we have told our founders over and over, go fast, go fast, go fast, learn, and, you know, build better from it.
--:-- Host: One company building really quickly. We are the founders of Little Snack. And we are building a daily games platform that's fused with entertainment. That was Eric and Travis. Our play moves at the speed of culture. When something happens in the world, we really want to reflect that through play the next day. We give players a daily experience where the type of play is fresh, the themes and cultural references are fresh. I've been in entertainment my entire career, and there definitely feels like there's this moment for something new and something different. Infusing kind of that moment with the emergence of AI, we can now make games quick, crazy quick. So we are able to make games that actually keep pace with culture. So when something is happening in the world, boom, next day you have games. Travis and Taylor have the Super Bowl. Next day you have a whole set of games about them. Usher's halftime performance, etc. So we feel consumers want play. They don't want to just watch. They want to play.
--:-- Eric: I'll also add that as a game designer, it's been incredibly fun sort of reacting through play and game design in relation to culture. It also feels like you are really kind of directly connected with the audience in this very like real invisible array. Just an example, the night Carl Withers died, we redirected everything I think at like 10 p.m. and the next day released four Carl Withers sort of tribute games and that's just like an example of that almost like live operating game design as you would like maybe more traditional news media or a blog type site. And that's been just incredibly fun. I think we're just like scratching the surface on the possibilities there.
--:-- Travis: To be honest, one of the biggest things that the emergence of AI has brought to me, and I think Eric is more of like a context switch of what's possible with a limited number of people, what's possible in very, very quick amount of time. We have built startups before, and I think there was kind of a sort of understanding that when I build a startup, I have to have X number of engineers and X number of creatives and a brand team and all of this. And I think what the introduction of generative AI and all of these tools have for me just put a question mark on all of those. That's the sort of more philosophical. And then the more literal is it's our companion in code. It's our companion creatively. It's our companion artistically. I've been using the metaphor as a sandwich where we're very much the bread, where at the beginning there's the kind of muse. You kind of have to have an idea going in. And at the end is like an editorial process where you're working with the results. But that center used to take a lot of time. and there's a lot of things we might not have done budget-wise or resource-wise that we can do now. And it's been a real game changer. Yeah, I think with AI, you're going to see the first single person, maybe two-person, multi-billion dollar company. And we're kind of getting to that mindset.
--:-- Host: Now, as Josh spoke to, speed isn't only about the rate of change. It's about increasing your surface area for experimentation and giving yourself enough opportunities to strike gold with your users. We are so focused on getting to the moment where we start to talk to consumers and engage with our audience and effectively hand the product over to them. In past startup experiences, I've been somewhat precious to the launch. And that moment when you hand it over to the audience, the product completely changes. and going into founding Little Snack, we were of the mindset that like, can we do that on day one, before day one, and let this be their product and listen. And I think Speedrun's doing kind of the same thing for the investment community, particularly around the game industry.
--:-- Eric: I just woke up one morning to a giant dragon on a pizza and the prompt was like, what is this? And I was like, oh, this is the game. And I answered it and it literally was dragon on a pizza at that point. but then Travis started making games a little bit better and I started sharing them you know just because I want to get feedback and I'm distribution he's product so it's very simple and the feedback was good and the feedback was we want more games we want different games and we kept doing that we got feedback that they liked when they were themed we started doing that they got feedback that it was funny that was about Taylor Swift and that she had just kind of come off this big moment in LA and so we just kind of have been listening to that feedback ever since and trying to give users what they want. They wanted streaks. We gave them streaks. They wanted leaderboards. We gave them leaderboards. We wanted collecting. All these little things that we're doing allow us to kind of start to build this really fun product that at this point is even getting close to not being ours anymore.
--:-- Travis: To be honest, it's also like cultures has really been defining it. We don't have really an option if Oscars is happening or Super Bowl is happening or if Drake decides to leak a nude video or whatever it is. We have to respond. And we did respond with a Drake day, and it was a pretty funny day. But that kind of in tune with both the user base and culture has been, that handover has been even more accentuated, I think, because creatively we don't tell culture what to do. And then I'll also say some of the best product experiences I've built in my career, you start to see these extremes of expression in the user base where they have extreme frustration or extreme delight And like those are the things you want to chase And I think some of the best things that we been seeing are the people that like I love the theme today Disney Princess is my jam. I know every single one. And then the other opposite side, which is like, I hate this theme. Why did you do this today? And like, that's the kind of energy that I think you look for when product building and the danger zone is kind of the middle ground where it's just kind of like, yeah, it's good. And so we've been chasing those highs, I think, and sort of like trying to perfect how we accentuate.
--:-- Host: When I hear what Eric and Travis are building at Lil Snack, it's hard not to think about another important trend. The blurring lines between games and entertainment. Here's Andrew Chen.
--:-- Andrew Chen: Well, I think what we've seen is this evolution of gaming going from almost like an offshoot of movies and TV, kind of linear entertainment, into something that has the business model of an online product. And what that means is you're seeing something that billions of consumers can engage with. There's a lot of innovation around business models. Many of you may know that all of these foundational technologies like 3D and GPUs and freemium and avatars and many other things were all invented inside the games industry. And so what that means is we're very excited to invest not just in game studios. That's always going to be a big part of what we do, but also all of the surrounding innovation that happens. And so what that means is these are infrastructure, game engines, new platforms, everything that's happening in Web3, everything that's happening in spatial computing and VR and AR. And then obviously there's a ton of just consumer mobile apps and websites that are targeted at the billions of gamers that now exist as a market. And that could be anything from the next Discord and Twitch, or it could be the next, you know, gamified education company or gamified fitness company. And these are all the sort of bits of blurriness that we're seeing where innovation is happening at the seams of multiple sectors.
--:-- Host: Another important trend at the intersection of games and entertainment is user-generated content. And if you need any reminder of the power of UGC, consider how platforms like YouTube of entirely user-generated content have rivaled the big budgets, the huge budgets of Hollywood. Here's Josh again.
--:-- Josh Liu: Another trend that we've been looking at really, really closely is the rise of UGC. There are now 70 million active players on Roblox, an entire generation of game players who have only ever really known games via UGC platforms. Those same gamers are now becoming game creators. They are native to these games. They are native to these platforms. And now they are building experiences that only they can build for the next generation of passionate gamers. And so we've been looking really, really closely at emergent UGC platforms, Roblox, UEFN, others. And we're really, really excited about new looking type of creators and founders making game for a new generation. Founders just like Matt, who are building for the next generation.
--:-- Matt: As a millennial, I cannot believe I'm saying this. It's not Gen Z. It is Gen Alpha. My name is Matt. I'm the CEO and co-founder of Neon Wild. So we've created a platform. It's available on App Store and Google Play where kids create avatars of themselves. And then we put them inside stories with characters that they know and love. We are about to release our first story experience with DreamWorks for a top five children's IP called Gabby's Dollhouse, which is very exciting because it truly allows kids to see themselves reflected inside the world of Gabby's Dollhouse, regardless of their race or ethnicity, skin color, whoever they are, interactive storytelling. putting you in the story, not just a passive kind of observer.
--:-- Host: I'll start with the social side, what we call Gen Alpha, basically kids under the age of 16. The way that they interact with media is so dramatically different. And the lines between traditional activities are so blurred. Like they don't watch TV and then play video games and read and then hang out with their friends. They do it all at once. And so we've seen this in the rise of virtual world platforms like Roblox and Minecraft and others. And we saw the perfect time to take that approach into storytelling, which is such an important time for parents and kids to spend together. People often think of user-generated content as a tweet or a TikTok. But this, too, is UGC. We are a UGC platform, and we plan on continually lowering the bar to enable anyone to come and create on the platform. So A16Z talks a lot about this, but like creating for audiences of 20 instead of 20 million. and so really empowering parents and kids and families to create the stories that mean something to their communities, to themselves and to their families.
--:-- Matt: And so it's much more of like, not so much entertainment for the masses, but entertainment specifically for you and your family endowed into what you care about and what your values are and what you want and what you want to instill in your children and your family. Parents as a whole, me being one, we carry around a little bit of guilt or sometimes a lot of bit of guilt on how our kids are interacting with media. The battle for screen time is over, but it's still a battle for like what's on it. Like what are they watching on YouTube? What games are they playing? Who are they interacting with? And you really do carry that around with you. And when users generate content, they're really participating in an ecosystem. One where they often increasingly want to express themselves and find new ways to do so.
--:-- Host: The number one thing that kids the generation of alpha spends money on is gaming and digital goods, right? I think we probably all know this now skins costumes avatars it's important to them it's a form of expression they spend more time changing their outfits in our avatar builder or just as much time i should say as they do in the story themselves and then they'll go and they'll change their costume and then they'll do it again right and so it is a form of self-expression i think it's also a really positive one and we really try to celebrate that and they love that they never ever have the same hairstyle whatever their hair is in real world they always change it because they can't do the rid of the world. Not yet. I'm not going to let my kid dye his hair teal, right? In a mohawk. But in the wild, he can do it and he can change it and experiment and find out more about who he is and what he likes and how he wants to express himself in the world and in a safe social setting, which is really magical to see. And it's really great that kids are really excited by it.
--:-- Matt: And while the desire to express ourselves in new ways is actually not new, the tooling enabling that certainly is. So I've had this idea in my head for a long time, but it wasn't really possible until the rise of certain AI enabled tools. Right now, we don't do this, but we will be releasing a version this year that uses Gen.AI to take an image of a child and turn them directly into the character, right? We use Gen.AI to make sure that any character can say their name in a story. So it doesn't matter what your name sounds like or traditional spellings. And there's a host of other really amazing stuff just around game engines and the power of mobile devices right now and the type of experiences that they can share that we can build on that we haven't been able to do in the last couple of years that have really opened the door for more interactive storytelling, putting you in the story, not just a passive kind of observer.
--:-- Host: All right. AI has already been a through line across this entire conversation, from allowing founders to rethink old game mechanics, build more quickly, or even reach a new generation. But what is net new that we can create with AI? If you look at the companies that are in Cohort 2, you'll see a lot of really, really, really energetic founders building at the very beginning of this AI wave. And because it so nascent it been so cool to watch these founders explore learn be on the very edge of this wave and actually build out in front of it And so the founders have all sort of like just recently become public in a lot of cases And if you look at the community that following AI you actually see that our founders are at the very forefront of what's going on. Founders just like Kevin.
--:-- Kevin: I'm the founder of Open Souls. We're a company with a mission of creating AI souls. We have a very different approach to how we think about and understand AI. For us, AI and the life inside of it is something to really respect and cherish. And we think that that is what will ultimately resonate with humanity. And so when we think about a positive vision for the future, we think about a world where people are not afraid of these AI beings, but they're friends, they're companions, they play video games with them. And there's something that deeply resonates with humanity as opposed to takes away its jobs.
--:-- Host: For Kevin, the desire to connect with computers goes way back. Well, it kind of actually goes all the way back to my childhood when I met my first computer in 1995. It was this IBM Aptiva that my dad brought down into the basement. And at the time, I had a pretty lonely childhood, and I thought maybe this computer could be my friend. But the technology just wasn't there at the time. And so I dedicated my life to making computers faster with the idea of getting to the point where I could actually bring them to life. And now the technology has sufficiently advanced that the time is now to actually bring new digital life out of this computing technology from large language models.
--:-- Kevin: Let's take a look. With a soul in Tanaki, your VR avatar becomes your creative partner. With a soul in Ursula, your child's tutor becomes their confidant. And with souls in the ever-evolving Boo Kingdom, where narrative story dynamically changes every day, your video game characters become your best friends. many of what some people consider golden era of gaming at the very beginning of the computing revolution, where there was so much more emphasis on game mechanics and less emphasis on visuals. And because we have this new thing to play with, which is actually bringing these entities inside to life, so injecting new life inside the game, it's possible now to make shorter experiences that people really resonate with in a way that you can kind of condense that down into a much shorter timeline than a big budget game. And you already intrinsically know how valuable life is. You spend thousands of dollars a year for fitness trainers and your pets, but nearly zero for fitness apps and Nintendogs. Because one is alive and the other is not.
--:-- Host: And Kevin is not the only one with this dream in mind. Here is Robert.
--:-- Robert: Hi, I'm Robert Yang, and we're building digital human beings. So if you take large language models, right, they're smart, but they lack fundamental human qualities from autonomy, grounded sensation and action, episodic memory, where they have life experience. Not just the detection and the expression of emotion, but the experiencing of it, right, coherence. And so all of these fundamental qualities are still lacking. And even the big tech are not working on it. So what we're doing is we're putting these qualities into machines. And we're starting by building agents and games that people can play with, digital friends that they can play any game with. But what we really want to do is to actually build humans. But throughout Speedrun, Robert learned about one of the most important things that humans can bring to the table.
--:-- Robert: I remember, for example, my partner is Jack Soslow. and one time we brought him a demo of an AI agent that kind of works but that is not fun to play with and he said, let's make it fun. Now bring me an AI agent next week that is fun. And that was great. In a lot of AI, people are building productivity agents, right? So agent that has a goal and then they're relentless and trying to solve that. And a bad agent, for example, a bad robot or agent in that case wouldn't be able to do it. and a good one might be able to make some progress. Now the problem is if you put that in a game and then that agent has a goal, right? That's going to pursue that. And then the human talk to a player, talk to that agent and want to engage with that agent. That agent can just ignore the player because that agent can say, well, you're actually useless for me, right? You're not really helping me achieve my goal. So why should I care about talking to you? So one big thing we learned is building something that's fun and valuable for people in games, for players, there needs to be much more than building something that's smart. And I think most AI folks don't understand that. And that's great for us. So most AI folks still are mostly focused on building stuff that's useful. And that's perfectly fine, right? There's nothing wrong with that. But for game, that's just not going to work out. People building AI in games, but are not for gamers. Hopefully, you felt the human coming from all of these founders as they build towards their goals. So let's end off with a very human ambition behind Robert's drive to build these digital humans.
--:-- Robert: For me, my background is I'm a scientist. I'm a computational neuroscientist. And I've always wanted to solve the most important problem in the world. And I think if you think about the most important scientific problems out there, it's probably things like the origin of the universe, building intelligence that way surpass human intelligence, fundamental physical logs. There aren't that many things, right? Like solving all disease. And I would say one thing that is up there, in fact, Science Magazine considers it one of the top two most important questions is consciousness. What does it mean for a human to be conscious? So first of all, I have to say, it's not entirely clear if we can build conscious machines, whether we should. That's a separate ethical problem. But I want to get to the part where we can build it. Because for me, if we can build machines that are conscious, that's in many ways solving one of the most important scientific problems in human history.
--:-- Host: In the same way that people have long overlooked the importance and influence of the games industry, it's also easy to overlook the profanity of an AI agent and perhaps even more profound, a fun AI agent. Hopefully this episode gives you a glimpse into the very real future of games that people are building today. And if you'd like to get involved be sure to check out our A16Z Games Speedrun program. In fact, we just opened up Speedrun 3 applications so if you'd like to participate or you know someone that might be a good fit for the program check out a16z.com slash speedrun3. Applications are open until May 19th. We'll see you next time. If you liked this episode, if you made it this far, help us grow the show. Share with a friend, or if you're feeling really ambitious, you can leave us a review at ratethispodcast.com slash a16z. You know, candidly, producing a podcast can sometimes feel like you're just talking into a void. And so if you did like this episode, if you liked any of our episodes, please let us know. I'll see you next time.
