# David Baszucki: Roblox, Minus One, Digital Economy, & the Metaverse

Markdown copy of https://realalpha.app/v/0tR7hcXRHXM/ on Real Alpha, the best minutes from long interviews with the people building AI. Cite and link the page, not this copy.

David Baszucki says Roblox’s growth comes from a creator platform whose tools, virtual economy and distribution reinforce one another.

- Show: Minus One
- Hosts: Dylan Itzikowitz
- Guest: [David Baszucki](https://realalpha.app/people/david-baszucki/) is chief blocksmith at Roblox.
- Published: Jun 4, 2025
- Length: 58 min

## The best minutes

- [10:34](https://realalpha.app/v/0tR7hcXRHXM/?t=634) [Roblox seeded its first viral loop with about $100 a day](https://realalpha.app/s/roblox-seeded-its-first-viral-loop-with-about-100-a-day/) (4 min)
- [14:41](https://realalpha.app/v/0tR7hcXRHXM/?t=881) [Roblox built its creator economy when growth stopped paying](https://realalpha.app/s/roblox-built-its-creator-economy-when-growth-stopped-paying/) (4 min)
- [27:02](https://realalpha.app/v/0tR7hcXRHXM/?t=1622) [Roblox is turning its platform data into creator tools and NPCs](https://realalpha.app/s/roblox-is-turning-its-platform-data-into-creator-tools-and/) (3 min)

## Also worth watching

- [48:09](https://realalpha.app/v/0tR7hcXRHXM/?t=2889) [Creator earnings are spreading beyond the top developers](https://realalpha.app/s/creator-earnings-are-spreading-beyond-the-top-developers/) (2 min)
- [21:27](https://realalpha.app/v/0tR7hcXRHXM/?t=1287) [Older users arrive through the creator ecosystem](https://realalpha.app/s/older-users-arrive-through-the-creator-ecosystem/) (2 min)
- [42:26](https://realalpha.app/v/0tR7hcXRHXM/?t=2546) [Roblox made email and meeting punctuality cultural norms](https://realalpha.app/s/roblox-made-email-and-meeting-punctuality-cultural-norms/) (2 min)
- [45:42](https://realalpha.app/v/0tR7hcXRHXM/?t=2742) [Roblox held off on NFTs because 3D assets lacked standards](https://realalpha.app/s/roblox-held-off-on-nfts-because-3d-assets-lacked-standards/) (2 min)
- [52:03](https://realalpha.app/v/0tR7hcXRHXM/?t=3123) [Roblox chose native clients for 3D performance](https://realalpha.app/s/roblox-chose-native-clients-for-3d-performance/) (2 min)

Skip the SPC introduction and early company history, the education and AR segments, and most audience Q&A on BCI, dogfooding, and future interfaces.

## Summary

Baszucki says he worked on Knowledge Revolution, a physics-simulation education software company, before starting Roblox with a team of four. He describes Roblox as a self-serve, user-generated gaming platform built around immersive 3D experiences.

Baszucki traces Roblox to a physics-simulation project where young users preferred making their own creations, then describes Roblox as a platform built to let creators and users sustain a reinforcing cycle. He says virtual currency and creator payouts helped turn that system into an economy, with three studios above $50 million in annual run rate and more than 100 above $1 million. He sees room to grow with older audiences, AI tools and social 3D experiences, while emphasizing moderation and connection. For founders, he recommends pairing a big vision with a clear niche, strong people and persistence; he says creator economies take many small design choices and should prioritize engagement over revenue.

### Minute by minute

- [2:03](https://realalpha.app/v/0tR7hcXRHXM/?t=123) **A creation platform can self-propel.** Baszucki says he spent two years in a pre-Roblox minus-one phase before a circuitous path led him back to an idea he loved. At Knowledge Revolution, young users turned physics-simulation tools toward building cars and structures, prompting a vision for an immersive 3D creation platform. He says Roblox chose a platform over a single game because it could become a self-running system; within hours of launching the development environment, creators were making unexpected experiences.
- [9:06](https://realalpha.app/v/0tR7hcXRHXM/?t=546) **Small ad spend sparked virality.** Baszucki says Roblox balanced deep tech with iterative improvements as it matured: early work was 95% vision-driven, while search and recommendations now get regular algorithmic refinements. He says buying initial traffic for about $100 a day helped kick off virality, after which Roblox grew from 100 to 1,000, 10,000 and 100,000 users without spending heavily. Chris Freilich found Roblox through his son, and Anthony Lee found it while looking for the next Club Penguin.
- [14:41](https://realalpha.app/v/0tR7hcXRHXM/?t=881) **Virtual currency powered creator earnings.** Baszucki says early Club Do membership revenue per hour fell as users and server costs grew, so Roblox chose a harder virtual-currency system over tactical fixes. Currency purchases, creator payouts and search and discovery created a feedback loop. He says three studios exceed $50 million in annual run rate, more than 100 exceed $1 million, and the money flowing through the system will hit $1 billion this year.
- [18:33](https://realalpha.app/v/0tR7hcXRHXM/?t=1113) **Roblox sees room beyond gaming.** Baszucki says Roblox began imagining an immersive 3D category alongside gaming, social networking and user creation, but describes the metaverse as early and uses terms such as human co-experience platform. He says $4 billion to $5 billion of a $190 billion gaming market flows through Roblox, or about 2.5%. He says education is already emerging on the platform, citing Boston Museum of Science’s mission-to-Mars simulation and arguing that consumer tools can become refined through broad use.
- [21:27](https://realalpha.app/v/0tR7hcXRHXM/?t=1287) **Older audiences need new experiences.** Baszucki says the platform’s future is for all ages, but serving children has required safety, civility and moderation systems; he reports that users over 13 are growing more than 35% year on year. He says attracting adults depends on an ecosystem where new experiences can emerge, citing Dress to Impress, made by a team of five or six in the UK and played by 20-year-olds. He sees potential for virtual fashion to connect to physical retail.
- [24:55](https://realalpha.app/v/0tR7hcXRHXM/?t=1495) **AR could open new interfaces.** Baszucki says digital and physical shopping could become social, letting people try items on with friends and connect digital looks to physical purchases. He is bullish on augmented reality and is waiting for lightweight AR displays; he also sees virtual reality improving and expects unknown form factors. He notes that Roblox is not yet on TVs and says consumer tools for designing homes could eventually bridge digital and physical spaces.
- [27:02](https://realalpha.app/v/0tR7hcXRHXM/?t=1622) **AI can support creation and moderation.** Baszucki says Roblox runs more than 200 internal AI models for safety and moderation, including models for images, 3D objects, audio and text. He says the company has begun offering language and 3D generation inside experiences, with a goal of letting people create assets by describing them and eventually generate full 3D experiences. He cites over 7 billion hours of monthly user activity as a possible source for training NPCs and says that data will remain on Roblox.
- [30:22](https://realalpha.app/v/0tR7hcXRHXM/?t=1822) **Immersive platforms can support connection.** Baszucki says immersive games are primarily ways for people to communicate and connect. He contrasts them with image-sharing and video categories that he says bring distinct risks, including FOMO, blackmail and overconsumption. He says Roblox must connect appropriate friends and uses moderation systems to measure civility, while also recognizing benefits in other platform types.
- [33:17](https://realalpha.app/v/0tR7hcXRHXM/?t=1997) **Creator payouts shape the economy.** Baszucki says Roblox values its relationships with Apple, Google, Sony and Microsoft while trying to route as much user spending as possible directly to creators. Asked what founders should focus on, he says Roblox found a niche with a big vision, kept asking what 10X would look like, chose people carefully and pushed through flat spots. He names an electric Dodge Ram 2500-style pickup, safe air taxis and computers a million times faster as things he would like someone to build.
- [37:02](https://realalpha.app/v/0tR7hcXRHXM/?t=2222) **Product loops signaled platform fit.** Baszucki says Roblox pursued Roblox Studio and a friend graph from the start, but lagged on social features and did not have chat for years. He lists user-generated creations, virtual currency, games running on mobile and console, and auto-translation as signs that the platform’s loops were working.
- [40:31](https://realalpha.app/v/0tR7hcXRHXM/?t=2431) **Culture shifted with company scale.** Baszucki says his leadership shifted from volatile to steadier and later back toward more detail as the team grew. He recalls shutting Roblox down for a day and a half after an early SQL-injection hack, and another outage lasting over a day about two and a half years ago. He says Roblox codified values around responsibility and punctuality, and that leaders still need to play and inspect the product even if hands-on use is harder.
- [45:42](https://realalpha.app/v/0tR7hcXRHXM/?t=2742) **Creator economies need patient design.** Baszucki says Roblox did not go all in on NFTs because 3D asset fidelity and interoperability technology were still immature, though he says a stablecoin could be worth considering. He says the monetization growth rate of creator 1,000 is higher than creator 1, and that creator economies take thousands of small decisions to get right. He says Roblox’s economy team prioritizes engagement over revenue.
- [52:03](https://realalpha.app/v/0tR7hcXRHXM/?t=3123) **WebAssembly could expand Roblox access.** Baszucki says Roblox chose a native client for 3D performance, but sees WebAssembly as a possible route to web play as performance improves. He expects AI NPCs to emerge from human-motion and interaction data, while saying Roblox prioritizes human-to-human interaction. The company’s stated ambition is to host 10% of global gaming DAUs and bookings; Baszucki quantified that target as $19 billion in bookings and 300 million DAUs, though not necessarily within two or three years.

## Transcript

Minus One, hosted by Dylan Itzikowitz, with David Baszucki.

Transcribed by machine (Whisper) from the episode’s audio, so a word may be off here and there.

[0:00](https://realalpha.app/v/0tR7hcXRHXM/?t=0) you started Roblox as a lifestyle company. It's now one of the biggest publicly traded tech companies. When did you realize that this was much more than a lifestyle business? Part of that circuitous path was actually thinking building something like Roblox would be really cool. The second we were live, though, the obsession started to kick in. One of your main investors found you because his kid was playing Roblox. That's a great articulation of build for your customers, not for investors. Chris Freilich, First Round Capital, his son Max found us. I won't say whether the partners approved it on the first round or a year or two later on the second round.

[0:37](https://realalpha.app/v/0tR7hcXRHXM/?t=37) That's like a funny story. But the cool thing is they found us through that. How did you get the first million kids on Roblox? We made the leap from friends and family to virality by spending about $100 a day. And we didn't do that too much. We didn't buy a lot of traffic. But that initial traffic was enough to get the machinery going and kick into virality. What do you look for when you're advising, investing, or acquiring founders? Wow, I would say... Before we kick off, let's give a huge round of applause to Dave and thank you so much for joining us. Thank you. I'll have to explain briefly for those who are new what South Park Commons is.

[1:21](https://realalpha.app/v/0tR7hcXRHXM/?t=81) South Park Commons has been around in the Bay Area for the last 10 years. We're now in New York and Bangalore as a community for people who are exploring what's next. Mostly technologists, researchers, entrepreneurial people who think that they want to figure out what their life's work is and want the time and space and community to pursue it. That's why we exist, and we've been fortunate to raise funds to accelerate the teams and the projects that come from it. and actually catching up with Dave before we started. I heard before Roblox, he had his own minus one journey that we're very excited to hear about.

[1:56](https://realalpha.app/v/0tR7hcXRHXM/?t=116) I guess we'll kick off by saying, giving you the floor to just explain to everyone what Roblox is, because probably most people here have heard of it, have played it, have friends or kids that have played it, but love to hear from your mouth how you would describe it. First, great to be here. And then I can't believe I did have a minus one as well. It was like a two-year minus one. And so when I heard that's a lot of what this group is doing here, it's like, whoa, that's really special. So I did have a two-year pre-Roblox thing. I also had a part of that where I was going in this straight path.

[2:32](https://realalpha.app/v/0tR7hcXRHXM/?t=152) And then I also heard you have this circuitous path thing, which is Roblox only got started because of then falling back to this random circuitous path where I really started building something I was in love with. And I thought it would be amazing. So wind back the clock a long time ago, social networking was new and interesting. User-generated content was new and interesting. Multiplayer gaming was interesting. And way back then, we had some notion that there would be this new category of immersive 3D simulated digital stuff that you could build games on, but ultimately go shopping on or hang out with your friends on or learn on.

[3:15](https://realalpha.app/v/0tR7hcXRHXM/?t=195) And so we were not gamers, but we kind of came up with this category idea, started very early, very primitively with four of us in an office in San Mateo for quite a while, actually, got a little bigger and bigger. And I would say we're really a UGC self-serve gaming platform that now almost 3% of the world's gaming runs on us. We've said we believe we could have 10% of the world gaming market space. And I would say along the way, there's been many of these, as you also said, little mountain up and down steps. Yeah, we call it the squiggle or find the right mountain to climb. And, you know, on your journey.

[3:59](https://realalpha.app/v/0tR7hcXRHXM/?t=239) You started in education, and how did you get from education to building one of the world's biggest gaming companies? So before Roblox, there was a company called Knowledge Revolution that was also a passion project. And this was started in the early days of the Macintosh, actually, with the notion that with physics simulation, you could take all of these pre-canned physics educational software lessons and make a laboratory or a 2D thing where you could build projectile motions and pendulums and all of that stuff. And when we built this physics lab, we early on saw a lot of young people wanting to build stuff rather than physics experiments.

[4:47](https://realalpha.app/v/0tR7hcXRHXM/?t=287) So they wanted to build cars and they wanted to build buildings blowing up and they wanted to do all of that stuff. So as we were playing, we were seeing this notion of physics simulation, this notion of people wanting to build their own stuff. There's a lot of sci-fi out there already about jumping into your creations. Multiplayer gaming was already big. So the notion of UGC simulated platform kind of came out of that. In that we knew if we put this combination of things together, that's what we would want. And so it took, you know, selling that company a two-year, as you call it, whatever, zero to negative one to zero, and then finally starting Roblox.

[5:32](https://realalpha.app/v/0tR7hcXRHXM/?t=332) And when you started Roblox, why start as a platform for creation versus building a game yourself? We weren't really gamers. And we, I would say very early on, there was something about platform that felt like a perpetual motion machine. You know, it was the notion that if we could bootstrap this thing or get it going, we would tinker on the side, but the machine itself would keep running and running. And so we were really fascinated with that notion. And we actually, you know, very early on had a couple places where we weren't quite all the way to platform. We tried to ship it early. It did not go viral.

[6:18](https://realalpha.app/v/0tR7hcXRHXM/?t=378) So we had to go back to the drawing board. But finally, there was a time we had all of the elements, 3D, cloud, avatar, simulation, all of that. And there was a specific day when we launched the development environment where people would build experiences and literally we could see like within hours, oh my gosh, okay, look at that. I can't believe that. I can't believe that. You know, it was one of those first kind of feelings that that perpetual motion machine. We're starting to see that viral loop where we would work on the side on the tooling and we would see this content web kind of start to come.

[6:59](https://realalpha.app/v/0tR7hcXRHXM/?t=419) You use the word bootstrapping. And I think it's interesting because I also read that you started Roblox as a lifestyle company. It's now 100 million monthly active users, one of the biggest publicly traded tech companies. When did you realize that this was much more than a lifestyle business? Yeah, I would say part of that circuitous path was actually thinking building something like Roblox would be really cool and saying, well, that's very risky. Like, how is that thing going to ever get huge? Let's just build it and figure out could we run with four or eight people as a lifestyle company. And so actually that's a pretty good signal that we're taking a pretty big risk in building that.

[7:42](https://realalpha.app/v/0tR7hcXRHXM/?t=462) I would say the second we were live, though, the obsession started to kick in. And the second, you know, we started seeing the potential time and time again and fall. Followed that and essentially to where we got to be today. You use the word obsession. How long has it been and how do you stay so motivated to keep doing it? Definitely gotten to be more of an obsession. I would say definitely many times when we thought this was the most difficult thing we've ever done. So it's not all just, you know, roses the whole way. I would say many times is going, I can't believe how difficult this is and stuff like that.

[8:32](https://realalpha.app/v/0tR7hcXRHXM/?t=512) But also the bright spots would be the ability to design your job as you go. One way, you know, the longer you go and are having an interesting time, the more you can compound the growth of the company. So I've been somewhat fortunate in continuously redesigning my job in a way, like making it as interesting as I can. So it's like it's pulling me into work rather than something I have to do. So, yeah, it's a little bit more of an inventor job, which is somewhat compelling. Most people might not know here, but Roblox is like a deeply technical product. And it has been from day one. And this is a very deeply technical community that we have.

[9:17](https://realalpha.app/v/0tR7hcXRHXM/?t=557) when you're building and investing so much in tech, you guys wrote your own coding language for Roblox. How do you also decide when to invest in growth? We do both. So investing in growth, you could think of the more mature we've gone, the more we've started balancing what I would call deep tech, vision-driven stuff with more incremental refinements that makes the product much better. We probably early on were 95% deep tech vision driven. We're probably now, you know, there's various areas that are very much what, I don't know if you're using the growth term, like our search and discovery and recommendation system does daily, you know, ruthless improvements in algorithms, you know, very growth driven stuff.

[10:06](https://realalpha.app/v/0tR7hcXRHXM/?t=606) Our 3D simulation engine, really no way to A-B test it. The only thing we can do there is take big visionary leaps. But I would say along the way, we have very much looked like a young kid's toy product, but underneath it, very deep technology. cloud simulation tooling distributed simulation 40 data centers bandwidth so yeah we could go into the tech but there's a lot there's actually a lot of tech opportunity in what we're building and taking it back to the fun kids toy how did you get the first hundred thousand million kids on roblox i'm trying to think who told me this hack and i'm wondering if you told me this hack Jonathan Abrams. Um, so, um, but no, there was a really good hack somewhere where, um, we were looking at, um, we were live with about a hundred people, um, hitting the site every day, um, very early friends and family. And someone, um, told me, you know, here's this really neat hack.

[11:23](https://realalpha.app/v/0tR7hcXRHXM/?t=683) You can go out and just buy 50 users from Google, you know, just go do that or just go buy 100 users from Google. It's like a buck a user.

[11:34](https://realalpha.app/v/0tR7hcXRHXM/?t=694) **David Baszucki:** So we we made the leap from friends and family to virality by spending about $100 a day.

[11:42](https://realalpha.app/v/0tR7hcXRHXM/?t=702) And we didn't do that too much. We didn't buy a lot of traffic. But that initial traffic was enough to get the machinery going and kick into virality. So it was actually very fun and interesting. I think it's a little bit more expensive to buy. those users today, but. Yeah, arguably, I guess companies today need to hit 5 million users in like a month or something, right? So we were much more patient. We, you know, we went 100, 1,000, 10,000, 100,000, started hitting the big numbers a little later on, arguably didn't spend a lot of money doing it. But maybe the good thing is we had a year or two to refine what we were doing. Another interesting story that you told me back in the back before coming out here is one of your main investors found you because his kid was playing Roblox. And I thought that's a great articulation of build for your customers, not for investors. So very early on, there was a lot of savvy young players who found us. I think they found us because when they were going through traditional gaming, there was not a lot of user generated type platforms. They could build their stuff on Roblox and start sharing it and checking it out. There's a lot of really interesting new

[13:02](https://realalpha.app/v/0tR7hcXRHXM/?t=782) stuff showing up on our platform. So yeah, famous, fairly famous VC, Chris Freilich, first round capital, his son, Max found us, you know, showed it to his dad. They presented it to the partners. I won't say whether the partners approved it on the first round or a year or two later on the second round that's like a funny story actually but the the cool thing is they found us through that so um they found us that way and then anthony lee of altos um also found us in a very interesting vc way out um when we got started i don't know if any of you have ever heard of it there was this thing called club penguin which was like a kid's virtual world In retrospect, ridiculously primitive, flash-based 2D, orthographic, but going very viral.

[14:00](https://realalpha.app/v/0tR7hcXRHXM/?t=840) And he was looking for the next Club Penguin. In retrospect, we were radical technical over the top, right? We're like 3D simulated, C++, game engine, Lua script, driven, developer tooling, all of that. competing in this kids' virtual world space. So Anthony found us, Chris found us, and they were our two first investors. Anthony and Dave are both fellow Canadians, so another connection there. And let's just say the pre-money valuations in those days were well under $25 million. And that was what, at Series A? Yeah. The good old days. Roblox has started with indie creators, but there's whole businesses built on studio, like Roblox Studio now.

[14:52](https://realalpha.app/v/0tR7hcXRHXM/?t=892) Can you talk about the rise of the whole Roblox economy from a studio perspective? Yeah, it actually goes to a fun metaphor for one of your ups and downs that you were sharing with me. So one of our ups and downs was we started maybe too much copying Club Penguin. So we had this Club Do membership thing. And here we were going along, growing rapidly, rapidly, rapidly. But as we were growing, revenue per hour, revenue per user was cratering, actually, with this old club membership thing. And so we were sitting there kind of revenue flat, a lot of user growth, more server cost, all of that. And we had this fun story where we thought we had broken something.

[15:38](https://realalpha.app/v/0tR7hcXRHXM/?t=938) and this is this goes to actually the best defense is a good offense we first went into defensive mode and spent about two or three months what did we break the magic is gone let's let's make a list of the 50 things we should fix um and then when that stopped working well what is the list of the 51 improvements we can And if we add all those up, does that get us 20? And it's just like, we did that. And then finally, in the back of our head, separate from all of that, we'd been thinking virtual economy. And we'd been thinking, you know, the system way to do this is have a virtual currency, have users buy it, have them spend it, have the creators be able to sync it for whatever they did.

[16:25](https://realalpha.app/v/0tR7hcXRHXM/?t=985) but that's like a really hard feature to build and like let's just do the easy thing but finally we

[16:31](https://realalpha.app/v/0tR7hcXRHXM/?t=991) **David Baszucki:** hit a point where it's just like whoops we just have to go do the big strategic hard thing that we think is right so it's like okay full court press none of these like fixes or tactical things

[16:44](https://realalpha.app/v/0tR7hcXRHXM/?t=1004) we're just going to build this thing we can't a b test it we just have to do it um and we need a full feedback loop we need to be able to sell the virtual currency you need to be able to sync it We need on search and discovery to reward some of it. So very similar to when we went live with UGC creation, we knew in about an hour when we went live, all of a sudden a couple of devs were syncing it. People were sourcing it. They showed up on search and discovery and kind of like got that full feedback loop of a perpetual motion machine. um this corresponded to a board discussion of would we ruin roblox if we made it you know a place to earn money and we actually um very early on had a lot of board discussion would this be a good or a bad thing i really pushed to do it because it it was the notion that creators could start earning a living so we got that machinery rolling that was also a perpetual motion machine. So that machine, when designed properly, just kept going and going and going. So there's now three studios that are making over 50 million a year run rate. There's over 100 making over a million a year run rate. The amount of money flowing through the system will hit a billion this year. That'll go up and up. And so we're starting to see really serious

[18:13](https://realalpha.app/v/0tR7hcXRHXM/?t=1093) studios now with hundreds of people. We're seeing M&A, we're seeing nine-figure acquisitions. We've predicted a 10-figure. No, we're seeing nine-figure. We predicted 10-figure M&A at some point. So yeah, that whole notion of building an ecosystem has moved forward. There's so much talk about the metaverse and you've basically built the metaverse. You have a virtual economy, virtual social environments, but you did it 10 years ago. Were you influenced by sci-fi? What were the things that made you think that all this would be possible? I think we were lucky when we started out. We have a business plan slide literally showing gaming, social networking, UGC creation, other things, and then showing a new thing we call the new category, you know, immersive 3D stuff. In a way, we got a little lucky with that, with the sci-fi was out there already kind of talking about it. In a way, we started building the metaverse, and I would say at the same time, it's very, very early because when we think about But the gaming market, for example, $190 billion market, right now about 2.5% of that is flowing through Roblox, $4 to $5 billion.

[19:47](https://realalpha.app/v/0tR7hcXRHXM/?t=1187) So the amount of potential upside is actually quite large. We're chipping at the edges of it. We don't use the term metaverse a lot. We use human co-experience platform, those kind of things. But and I would say also the metaverse in a way got connected a bit with hardware as opposed to a hardware agnostic approach. We've always arbitraged every device, phone, tablet, computer, console. So in that sense, I think it's a little more robust to be on all hardware. And do you think in this virtual world that you've created, you have an economy, you have with developer tools. Do you think education moves to a platform like this? Do you think that there's a future for that? I think it's already happening. Like we're starting to see a lot of interesting curriculum simulations. You know, Boston Museum of Science made a mission to Mars simulation, which is really cool. We're starting to see other things being built on the platform. But then another way to think about it is it's inexorable, basically. You know, Every new form of media, whether it's a connection form of media, whether it's books, whether it's video consumption, whether it's video connection, those all end up being used in education.

[21:11](https://realalpha.app/v/0tR7hcXRHXM/?t=1271) And in a way, when we started Roblox, that thinking was the best educational products sometimes are consumer tools that are used by millions and millions of people. They get they're free. They get hyper refined. They get super high quality. And then education gets to use them. You've had so much success building for kids, for young people. Why so much emphasis on building and trying to attract older users now? I would say it's, well, first, I feel the starting young was more difficult, but a huge advantage. Because starting young adds evolution of safety, civility systems, evolving into AI filtering, AI moderation, all of those kind of things.

[22:01](https://realalpha.app/v/0tR7hcXRHXM/?t=1321) I think, once again, the destiny of the platform is to build for all ages. That's also very difficult. Safety, civility, policy under 13, 13 through 17, a more vulnerable age, 18 and up. planning for ultimately experience where friends can play poker completely without moderation and the privacy of their own home and everything in between. So I do think that's the most robust architecture is planning for all ages. So we see enormous growth north of 13. Our north of 13 is growing over 35% year on year. And there's enormous amount of opportunity there for us. What's the strategy for acquiring the North of 18 demographic?

[22:52](https://realalpha.app/v/0tR7hcXRHXM/?t=1372) The strategy is primarily build the ecosystem very much like the very early days of Roblox. We've always thought, like, what would it take for emergent experiences that are really fun for someone who's 22 to play? It's the technology, it's the economics, it's search and discovery, it's having creators willing to take a risk.

[23:17](https://realalpha.app/v/0tR7hcXRHXM/?t=1397) **David Baszucki:** that's what happened last summer dress to impress you know a team of five or six people in the uk comes up with this and now all kinds of 20 year olds are playing on campus and vogue magazine is

[23:30](https://realalpha.app/v/0tR7hcXRHXM/?t=1410) talking about it today so i think it's very similar create an ecosystem platform where that can happen to dress to impress have which was a very big phenomenon did were you able to buy clothes real-life clothes with Robux? Or is that something that you'd want to see, like the interchangeability between the Robux ecosystem and real economy? So Dress2Impress was really interesting for a bunch of reasons. They took a future-looking approach. Avatars are not Roblox avatars. They're going in the direction we would like, where individual creators come up with their own avatar look and feel. Their gameplay was also very unique in the way they did animation and the way they ran that kind of thing.

[24:16](https://realalpha.app/v/0tR7hcXRHXM/?t=1456) There's an enormous opportunity, I think, for physical and digital to interchange in a place like Dress to Impress. And where Dress to Impress, for example, could become a place where you're both competing with clothes from your famous, your favorite designer, and also able to buy them physically. So we do think, you know, shopping is very 2D right now. There's some subset of shopping where when we go to Stanford Shopping Center and walk around with friends, it's kind of fun to go walk around with friends and try stuff on. And we think that'll start to happen digitally too. Are there other digital physical experiences on the Roblox roadmap?

[25:03](https://realalpha.app/v/0tR7hcXRHXM/?t=1503) Thinking through digital physical, there's going to be an interesting long-term digital physical as we finally start to get AR glasses that are really good. And both not just MR, but AR glasses, a lot of companies are working on them. That's going to create really interesting situations on physical digital communication and phone calls. You know, who's wearing VR, who's wearing AR, who's projecting into whose environment, I think is going to get very interesting. So that's a really interesting physical digital crossover, I would say. Trying to think what other ones besides shopping. I think ultimately there's some experiences on Roblox where people are designing homes, for example.

[25:59](https://realalpha.app/v/0tR7hcXRHXM/?t=1559) You could imagine a consumer tool like Roblox rather than an architectural tool, a consumer tool that gets really good at a design your own home thing, ultimately being something that goes digital as well. On that trend of like the technology improvements, you've seen so many platform shifts. You've seen desktop, mobile, cloud. Is AR VR the next big platform shift or is there something else that you're looking forward at? It's a little unpredictable. I wrote a blog post about 10 years ago. I'm very bullish on AR. I'm waiting for the first lightweight heads-up display in an AR device. I think VR continues to get better and we continue to have a lot of people on VR as well on Roblox.

[26:46](https://realalpha.app/v/0tR7hcXRHXM/?t=1606) I think there's probably some unknown form factors that we haven't quite thought about yet coming. You'd notice that Roblox isn't on TVs yet. That's interesting to think about. A lot of time is spent there. So that's a good place to think about as well. So I think there's more opportunity. We have to talk about AI. What have been all the ways that you've seen AI being used within Roblox? And then just more generally, how have you seen it impact the sort of creator gaming ecosystem? Yeah, I would say for us, we have radically reaped the benefits of behind the scenes, arguably boring AI, but for us, very non-boring AI.

[27:32](https://realalpha.app/v/0tR7hcXRHXM/?t=1652) **David Baszucki:** And so I would say we are running over 200 different models we built internally over the last four to five years, all on our own infrastructure, doing everything in safety and moderation.

[27:46](https://realalpha.app/v/0tR7hcXRHXM/?t=1666) So image, 3D object, audio, we now auto-moderate all the text on the platform, which is really exciting as well. We're getting we open. First, our voice model, because it's gotten so good and so efficient. So I'd say behind the scenes, lots of boring stuff. More recently, using that same infrastructure to start doing more interesting stuff, which for me, I would say is not just creator tooling, but AI as a service available inside any game anywhere. And so there's now TextGen LLM as a service inside Roblox Experiences. there is 3D gen becoming available in Roblox. We just launched that a few weeks ago.

[28:36](https://realalpha.app/v/0tR7hcXRHXM/?t=1716) And so the notion here would be, of course, we want to enhance game creators, but we also want to have enough technology so if you or I were building a fashion design game, we could get to the point where you or I are creating fashion by talking about it. And the AI that's generating that fashion is available in the cloud. essentially for free running on Roblox. We want to take that all the way to objects that are intelligent. So a lot of the objects on Roblox are physically simulated. They have embedded code in them. Being able to generate those by talking about it. Build me a drivable car that looks like this.

[29:20](https://realalpha.app/v/0tR7hcXRHXM/?t=1760) And I think the two more long-term things that are inevitably going to happen is one, full 3D experience generation. Just you can imagine a demo. Someone sketches a few things and says, hey, I'd like to make a fun game about this and this and generate that dynamically. I think that's on the way. And then I would say finally the opportunity to use all of the data of what people do and how they interact because there's about almost over 7 billion hours of that data every month. And so in a way, it's like, what do people do when they come up to a door? How do people interact with each other? What do people do when they're walking around?

[30:10](https://realalpha.app/v/0tR7hcXRHXM/?t=1810) Using that data to train NPCs is also on the horizon. And would that data and these models only be for the Roblox ecosystem? That's correct. That data shall never leave Roblox. So I want to, before we open up to the audience questions, I want to talk about something that I think is very important is safety and mental health and technology. So whether it's kids or the 18 plus or all ages, how do we balance keeping kids on technology but not over abusing the amount of time they spend on technology? This one's really tricky. I feel the category we're in is a bit of a fortuitous category. When we scan, you know, we work with legislators a lot, and they're all trying to figure out how to legislate social media right now.

[31:07](https://realalpha.app/v/0tR7hcXRHXM/?t=1867) And, you know, the social media, social networking starts breaking down into categories. The historical category of sharing images, you know, sometimes that's fun. It's our cat pics. Sometimes it creates FOMO. Sometimes it creates blackmail. So there's a whole category of that that needs a lot of thought. There's a whole category of consuming a lot of video. That needs a lot of thought as well, because in certain ways, I consume that on probably four different platforms every day, a lot. And at the same time, that can maybe be the way the younger people are getting their news today, so there needs to be a lot of thought in legislating that.

[31:53](https://realalpha.app/v/0tR7hcXRHXM/?t=1913) What's somewhat unique about gaming and 3D immersive platforms is they're very heavily a way that people connect and... communicate, just like being on the phone in the old days or being on a Zoom or a video. It's a way of communicating. And so there's something nice about building platform that helps people communicate with their friend. We have to be very careful about only connecting people with their appropriate friends. I mean, that's a huge opportunity for us. But lots of people get a lot of connection on platforms like Roblox and not just with their real life friends, but with new friends. So in that sense, it's both very, very important and a great opportunity.

[32:39](https://realalpha.app/v/0tR7hcXRHXM/?t=1959) I think we, given the way we run our moderation systems, have the opportunity literally to measure that level of civility and see if it's getting better from people on the platform. And I know you've spoken about why Roblox is better for mental health and that image-focused social media platforms? Yeah. Yeah, I think we want to, I'd say we think there's benefits of all platforms, so we wouldn't throw them under the bus. What we do want to make sure is, you know, with legislature, with what's happening in D.C., there's a lot of understanding of the mental health benefits of bringing people together optimistically for many, many people in the world.

[33:21](https://realalpha.app/v/0tR7hcXRHXM/?t=2001) Speaking of D.C., because you mentioned it, How does Roblox feel about the Apple, Google store stuff happening with Epic right now? And no pressure to answer. Yeah. We don't have to answer that one. No, I'm really happy to answer this. So the fun answer is we're really great partners with Apple, with Google, with Sony, with Microsoft, you name it. And at the same time, we're trying to move as much money as we can from users directly to creators, you know, not pass go, not go somewhere else, go straight to creators. So every movement to moving more money in that direction makes a more vibrant ecosystem, drives the economy, helps our creators.

[34:09](https://realalpha.app/v/0tR7hcXRHXM/?t=2049) So kind of to say those two things can coexist at the same time. Which it's consistent with the mission that you mentioned earlier about the dollars going to creators on Roblox is like unlike any other platform. The more we are efficient, lean infrastructure, the more users can route their money to creators, the more I think vibrant our economy is. So before we open it up to questions, there's a lot of entrepreneurs here that only wish to achieve what you've achieved. um what do you look for when you're advising investing or acquiring founders like what are some of the attributes that people can strive for wow i would say i'm probably a terrible person of sussing out companies that's why i'm not an investor it's probably super super hard um i would say um the thing that we got lucky on at roblox is um found a niche with a big vision I would say along the way, we were always trying to imagine what would 10X be like?

[35:15](https://realalpha.app/v/0tR7hcXRHXM/?t=2115) Like, do we have a crisp view of 10X? Is it very clear right in front of us? Can we imagine it? We were very careful in the people that we worked with. We had our own set of values. We ended up getting lucky with really good people. I would say we had several flat spots along the way where it was the feeling. Definitely the feeling of the most difficult thing ever done. But we pushed through that. How did you push through it? For me, I became much more stoic. I go, okay, let's run the checkmark. I've got food. Family's got food. I've got shelter. Family's got shelter. I've got... clothing. There's no war going on. Why am I worried about anything?

[36:06](https://realalpha.app/v/0tR7hcXRHXM/?t=2166) And if you were to do it again, or if you were to advise someone starting today, what's an area that you wish someone would go build and solve for? I would probably like a pickup truck that has all of the characteristics of a Dodge Ram 2500 and is electric at the same time. Those two categories don't seem to exist at the same time obviously would like safe air flight right safe air taxi that's very easy for me to take so that'd be cool um computing i would like a computer that's about a million times faster i know that's going to take a while so we'll wait for that if a computer was a million times faster roblox could use it so we literally could use a million times the compute and still not run out of gas so um those are a few things you guys gotta go build uh so we're gonna open it up to some questions from the audience there are a lot so we'll do our best i'm a member here and i'm working on brain computer interfaces and like how to you know sort of interface with computers with your thought um i wonder if i could chat to you later about how would you like use thought as like an input for roblox like you know the stuff you were saying like oh you know imagine a thing that appears in the game um seems like a really cool sort of input to building stuff

[37:34](https://realalpha.app/v/0tR7hcXRHXM/?t=2254) yeah yeah i mean i've seen some of the research of ai enhanced images by looking at the probes it's kind of freaky so i do think based on what i've seen of the research at least like that's going to be a fertile interface. So, and love to hear what architecture you're using, number of bits, probe count, all that. Super cool. Thinking back to the early days of Roblox, I'm wondering for a UGC platform, how you balanced improving the creator tools and also focusing on the social tools for people to connect with each other and which you found most valuable during like the early stages of growth? I do think it's a balance between them. I would say, you know, we, we, from day one, we were thinking Roblox studio and a friend graph. We were thinking have both of those. We arguably probably lagged on the social stuff. Like we didn't have chat on Roblox for a lot of years. So people would use their own chat. So we leaned a little more on the tooling. And I think people found alternate social. So in retrospect, we might have done social a little faster. You mentioned a lot about building these perpetual motion machines.

[38:54](https://realalpha.app/v/0tR7hcXRHXM/?t=2334) And again, I guess, harkening back to the early days, what were these early signs of life that you were really listening closely for as you saw Roblox starting to work and the virality starting to take off? Yeah, well, we were always more tool makers and, you know, those types of people. And so, you know, we would have this vision of this machine that we could take a day off and the machine would keep running. Then we go tune the machine up and it would get better and better and better. So sign of life, sign of life one, UGC tooling, the system kind of becomes UGC and all of a sudden, you know, seeing UGC creations, sign of life number two, you could arguably say virtual economy, like virtual economy, sign of life system goes viral, kind of really interesting.

[39:51](https://realalpha.app/v/0tR7hcXRHXM/?t=2391) I would say sign of life three would be the notion that what we had running on PC could go mobile and console. with all of the exact same stuff. And so all of a sudden, everyone's game's running on mobile. Oh, my gosh. So that was also, I would say, hyper-leveraged. Sign of Life 4, auto-translation. So, wow, we're going to, because we know all of the words, we're going to auto-do this. So all of a sudden, every language. That's kind of cool. So every one of those that would multiply that, you know, it's just like gratifying. Hi, Dave. Thanks so much for sharing your insights about Roblox. I'm Rob.

[40:34](https://realalpha.app/v/0tR7hcXRHXM/?t=2434) And I wanted to actually ask you a little bit about leadership in the early days of Roblox. Maybe one of those darkest times you had to make a really tough decision. Is that maybe something you would be comfortable sharing with us? Something you encountered and what you learned from that? Yeah. I mean, you learn a lot of this as you go. I would say there's a definitely ebb and flow in my leadership, both emotionally as well as detail oriented. Emotionally in the early days, much more volatile, right? Like it's life or death, like you're yelling at people, burning bridges all over. I would say over time and kind of a little smoothed out.

[41:15](https://realalpha.app/v/0tR7hcXRHXM/?t=2475) And then I would say another ebb and flow was very early on, hyper detailed. A little bit in the middle of my tenure, less detailed. Now, arguably, as our team is so strong and everywhere, I'm actually able to do more detail. And I actually feel that's super satisfying. So it's been like back to detail here. I would say unexpected events are very difficult. So I would say we had two events, arguably both our own making. event number one, SQL injection hack early on, like the whole economy, you know, just like fans out. There's no floodgates. There's no like friction in the economy. So we just shut down the whole site within 20 minutes and had to be down for about a day and a half, which is very bad. Then arguably about two and a half years ago, we were down for over a day with an October event, arguably our own making as well. Those are very difficult times, but figured it out.

[42:25](https://realalpha.app/v/0tR7hcXRHXM/?t=2545) In addition to leadership, can you talk about the importance of culture and team building? Because I've heard you talk about this as well. And I know you have an in-person culture. There's lots of other things. Why is that so important for you guys? I think we captured our culture, our value system pretty early on. When we had about 12 people, we wrote it down. It's interesting that a lot of things you might have as a CEO, as an intuition, you back off a little of it. And then as you reaffirm your culture, you lean into it and get really crisp. I would say one thing early on is we started having a lot of people from other companies. And you would hear like in a meeting, well, I'll, you know, don't send that email at seven because it might put too much pressure on someone if you send it. And when we, when we thought about that, that's a very fragile system because it puts so much onus on worrying about that rather than a responsibility system that look, everyone should figure out how to deal with

[43:37](https://realalpha.app/v/0tR7hcXRHXM/?t=2617) **David Baszucki:** an email, even if it's from the CEO at midnight on Sunday. And so we said, you know what? No responsibility culture, send email anytime. And kind of crisping that up was really helpful.

[43:51](https://realalpha.app/v/0tR7hcXRHXM/?t=2631) There's another thing where, you know, a few years ago, I'm kind of going into meetings and they're starting at like three after the hour. We're in five after the hour, and it's just like, I don't have effing time for that, right? It's like, you know, so we just said, hey, well, let's do an experiment and say we're an on-time culture. The meeting starts on the second, even if the CEO's late. And so now everyone just knows that meeting's going to start. If I walk in five seconds late, the meeting's already going. So there's these fun cultural things you can, you know, make more expressive of those early 20 people.

[44:29](https://realalpha.app/v/0tR7hcXRHXM/?t=2669) And maintain them through to today. Hi. So when you're building a fun and creative platform like Roblox, you can spend a lot of time heads down building these really complicated features. How do you maintain a balance between building heads down and dogfooding your own product and staying connected to the actual games people are making and the things that you can create on this platform? How do you balance those two? Early on, that was a lot easier because we were literally playing roblox six hours a day you know we were we were building shipping playing constant feedback knew all the first thousand people on the platform much more difficult now especially in my role like to literally have fake usernames have the specific time to go on and suss things out but it's super valuable and um there is you know having a lot of your leadership team logging bugs sending things being on slack all of that it definitely creates an attention to detail vibe we have a you know we have other principles around that kind of stuff that we try and enforce as well.

[45:44](https://realalpha.app/v/0tR7hcXRHXM/?t=2744) So for me, the Robux token and also the asset sharing community, they both embody a very strong Web3 spirit. So have you also thought about launching your own stablecoin and then combine it to the real world? We have thought about that. I would say we did not go all in on the NFT sitch three or four years ago because we felt we did not need to. I don't know if we need to right now. It's worth thinking about. It could be interesting. We have started thinking about that. And so I think the one thing we have started thinking about, it would really be cool if an eight-year-old who was on Roblox had like a carbon Roblox card and like pulled out their Roblox carbon card.

[46:45](https://realalpha.app/v/0tR7hcXRHXM/?t=2805) So yeah. In terms of platform shifts, did you discuss Web3 and NFTs? And why do you, I mean, you made the right call, obviously, but why did you decide not to go all in on it? I think there was definitely several board meetings where things were super hot there. And we, you know, things were like super frothy. And I think there was also a vision or a hope, which is very altruistic, that this pair of shoes would go to any platform and have value and all of that. But I think we came to the notion that for 3D assets particularly, the technology is still very, very early. The fidelity of these assets, whether they're physically constructed, whether we're simulating the elastic of the shoes, how they fit on an avatar, it's still very, very early.

[47:42](https://realalpha.app/v/0tR7hcXRHXM/?t=2862) And for quite a while, the technology is going to be moving.

[47:45](https://realalpha.app/v/0tR7hcXRHXM/?t=2865) **David Baszucki:** There's not going to really be a standard way to share these shoes.

[47:50](https://realalpha.app/v/0tR7hcXRHXM/?t=2870) So that made us pull back from that notion. Hello, hello. So my Roblox account turns 18 years old this year. Probably very valuable. probably people on roblox actually look at your user number and like there's value in low numbers i've seen um so thank you for introducing me to programming and starting my career but my question is you mentioned uh top few creators making a large amount of money um but you see the roblox creator economy being a winner take most market like actually i feel it's going in the opposite direction yes and so we do watch the growth rates of creator one ten hundred

[48:32](https://realalpha.app/v/0tR7hcXRHXM/?t=2912) **David Baszucki:** thousand and the monetization growth rate of creator thousand is growing faster than creator one so the curve that curve if anything is continuing to flatten i think the thesis we

[48:44](https://realalpha.app/v/0tR7hcXRHXM/?t=2924) would have as as wider ages wider um presence around the world occurs that there's just a lot more opportunity for a flat, flat curve. So we, a long time ago, we were worried, you know, could developer number 1000 support themselves? They definitely can now. So yeah, it's going in a good direction. Thanks again for all the insight. I have two questions. First, I feel as though we should see more creator economies. And I feel like you've kind of seen firsthand what's worked extremely well. And I think you touched on the perpetual engine, some specific thresholds that you cleared with Roblox. What advice do you give to founders that are constructing creator economies?

[49:29](https://realalpha.app/v/0tR7hcXRHXM/?t=2969) And then you also mentioned your board construction, or rather, you mentioned conflict within the board. I'd love to hear how you approach the construction of that vehicle for your company. Yeah, I would say it's not easy. I would, you know, every year, there's probably 1,000 small little feature decisions we make to make this work. And we do that year after year after year. So it definitely, we were looking, we're talking actually today in our staff meeting about certain trends. And traditionally, it has always been the culmination of 100 small things or 1,000 small things over time, rather than one just big breakthrough. So I do think creator economies take an enormous amount of craft to get right. On our board, we have very small board, seven people on the board. I think every board vote we've ever had is unanimous.

[50:31](https://realalpha.app/v/0tR7hcXRHXM/?t=3031) We have very thoughtful people. So I would say we've been fortuitous to handpick our board members based on kind of a, you know, superstar. Do we have the finance person, the marketing person, this person, this person? And so for me, small has been better than big. One of the things that you mentioned that I thought was super interesting was about how essentially you guys built your own virtual economy. I see a lot of like actual gaming companies trying to embed things like crypto type incentive mechanisms into their games, right? And it doesn't always work super well. So I was curious in terms of like if you had any lessons about how you built a really good virtual economy.

[51:14](https://realalpha.app/v/0tR7hcXRHXM/?t=3074) I would say we have always been very sensitive in two directions. One is weird mechanics that are short-term good, long-term bad. And then the other is over monetization pressure. So one of the biggest things we have done in search, in discovery, everything is literally the economy team's primary goal is to increase engagement, not revenue. And so if being very careful on over monetization, driving engagement, DAU's hours more than the bookings has turned out well for us. So I'd say that's been one. of our things. Hey, I'm a gaming founder. I'd love to hear about how you think about expanding to new platforms.

[52:07](https://realalpha.app/v/0tR7hcXRHXM/?t=3127) In particular, sort of a two-part question, how you think about the web? Roblox is a giant of UGC, and the web is a major UGC platform. I'm curious if the decision to still not launch Roblox on web is a technical limitation or a monetization form of payment readiness. So when you're thinking about a new platform like VR, how do you decide? Yeah, I would say 15 or 18 years ago, not being flash or not being web was a huge friction point for VCs. And, you know, you got a client, no way, you know, everything's flash, like there's no way.

[52:44](https://realalpha.app/v/0tR7hcXRHXM/?t=3164) **David Baszucki:** Because we were so into 3D performance, there's just huge native benefits.

[52:50](https://realalpha.app/v/0tR7hcXRHXM/?t=3170) And so we sucked it up and had the client install issue, which was, I would say, less an issue today. And now, you know, on PC and Macs, we use Spotify's app more than, you know, web. And so it's kind of come full circle. That said, web is super interesting for a spontaneous play, IO gaming. There's environments where clients are blocked and having a web version of Roblox would be very interesting. And also the performance is getting there now that it's super interesting. So I would say WebAssembly is super interesting. Figma runs pretty good on it. Google Maps runs pretty good on it. We've had people in Hack Week do interesting things with Roblox on WebAssembly.

[53:42](https://realalpha.app/v/0tR7hcXRHXM/?t=3222) And it's also very interesting that the same C++ code base that we target everywhere, iOS, Android, PC, Mac, whatever, just redirect to WebAssembly. So I would say there's some interesting possibilities there. It's interesting at SPC, we have a lot of people looking at the gamification of browsing. How do you turn every browser into a gaming experience? I think if we see high-performance gaming browsers, I have seen some companies that are like, we're building the best browser for high-performance WebAssembly or whatever these other things are. If Roblox was ever to have a WebAssembly version, it would work really good in those.

[54:23](https://realalpha.app/v/0tR7hcXRHXM/?t=3263) I'm just wondering, how are you thinking about AI NPCs or AI companions? This is really interesting. We are talking about this today. I do think I'm not saying Roblox is going to do it there is going to be this really unpredictable AI companion situation going on and whether it's mobile companion whether it's a robot like who knows right if you can imagine it stuff's going to happen there so I think that's really interesting I think on Roblox we are very focused on real human-to-human interaction and accelerating that. That said, as we start to train on all kinds of human motion data and what people do, I do think you'll see over time the ability for amazing NPCs on Roblox.

[55:17](https://realalpha.app/v/0tR7hcXRHXM/?t=3317) Do you think we're far away from having NPCs that are indistinguishable from humans? I think long-term, especially in gaming environments, this is going to be a really interesting challenge. You can already see it on chess.com probably, right? A simple client, who knows who's human and who's not out there, right? It's probably very, very difficult. I think what you see in the chess community, ultimate in 3D immersive gaming, it's probably going to be hard to distinguish humans from robots as well. And I would add, that said, we're thinking very thoughtfully about that and have quite a few plans.

[56:00](https://realalpha.app/v/0tR7hcXRHXM/?t=3360) Thanks for that. On a related note, I'm a 3DA IPM. I'm curious, what are the three wins that you'd like won from Roblox in the next three years? I think the win we're all going for is to see 10% of the DAUs and the bookings of the global gaming market run on Roblox. So that's what we said. Like we've been saying that for a year. That would be 19 billion of bookings and, you know, 300 million DAUs. Not saying it's in two to three years, but like the whole company knows that's the target. So we were speaking about Chess.com. So on the hackathon recently, we built AI Companion for Chess, which also speaks back to you.

[56:44](https://realalpha.app/v/0tR7hcXRHXM/?t=3404) So I'm guessing what is the voice as an input before BCI comes in in gaming? So if we thought we were a gaming creator, there's probably a lot of interesting services we would like. We like LLM as a service, we like speech to text, we like text to speech, we like translate text, we like translate speech, we like moderate text, moderate speech, stylized speech, all of those. That's like an interesting roadmap to imagine for Roblox, like give all of these pieces at high quality in the cloud and creators will start putting them together in interesting ways. So, yes, speaking to someone else on Roblox, having it be used in certain ways, like, yeah, that should happen.

[57:32](https://realalpha.app/v/0tR7hcXRHXM/?t=3452) I want to thank you so much for your time. This was excellent. And just give a round of applause. Thank you. Thank you all for coming. That was another episode of Miners One from the team at South Park Commons. Make sure to subscribe to our show wherever you listen to podcasts and find us on social at South Park Commons. And thanks to our friends at Atomic Growth for their support in bringing this episode to life.