How Roblox Built a Digital Economy Beneath the Games

[00:00] In Q4 of last year, it was 35 billion hours, pushing 12 billion hours a month. That's a pretty interesting stat. In Q4 of last year, we were at 150 million daily actives. Global gaming market is emblematic of really a new way of people to communicate. When we think about this market, we think not just gaming, but ultimately what we might call the human co-experience market. Our developer creator earned about a billion and a half on the platform, [00:30] 1000 devs are averaging 1.3 million. The most underrated aspect of Roblox is how much deep tech and theories around economics and theories around systems sit underneath it. Underneath it, there's enormous technical complexity. There's two ways to think about the acceleration. One is what is the future of a game? And so if I was a user and infinite AI was available to a user, it's almost as if we can't. [01:00] I can't imagine what that game would look like. Imagining AI is super powered in 50 years, what percent of adults work eight hour jobs a day? [01:12] *outro music* [01:21] Are you ready to go? Hopefully we've already been rolling. [01:24] Like we already have some good material. Yeah, we're going to see how spicy this can get. You want what level spice do you like? Let's go like super flaming hot Cheeto spice. Okay. Did you know that some Cheetos products have various levels of spice on the individual chips? No.

[01:45] Yeah. And it's a Pavlovian response where you don't even know it and you're eating individual chips. No. And you're hoping one is hotter. And so like you eat three or four and they're not hot. And then you hit one that's hot and it's some weird mental hack. [02:02] that makes the chips viral. [02:05] So when you talk about spicy chips, like, yeah. [02:09] Okay, wow. It's like shishito peppers, but it's... [02:13] unpredictable shishitos yeah [02:15] How did you know that? I don't know. It's some weird Pavlovian psychology thing. [02:21] Wow. [02:22] Okay. [02:23] Well, let's just keep, you know, the hottest Cheeto possible. I guess. We'll try to keep it hot. Okay. We'll try to bubble a hot Cheeto to the top. [02:33] Well, thanks for having us today. Great to be here. Welcome to Sorcery. Thank you. [02:36] Your campus is huge. Thank you. [02:39] It's ginormous. How many buildings are here? So there are five buildings. We are sitting 20 years ago. We would have been watching horse racing. Like we would have been looking that way and there would have been horses. So we're in the the Old Bay Meadows area of San Mateo. This was a development of five buildings. I think we're in four of them. [02:58] right now i would say it's a vestige of what happened in covid [03:03] We have some very wise people in the company and weekly executive meetings. Are we remote forever? Are we in person? We are very much innovation culture used to white boarding, used to being together. So we never really said remote forever. We kind of kept it together and we have almost the vast majority of our people here now. How many miles around the campus do you walk a day?

[03:27] um depends on the day because some days um some days i go for walks for an hour with people so um well you guys also come in with a squad like even coming through here i think there's been like 15 people in this room oh really you like that yeah yeah yeah okay now we have like [03:46] Hey, squad, what's going on over there? No, we have a good squad. Yeah. And hopefully, yeah, you can pick up some energy. [03:53] So as you build out the headquarters, how is that going to facilitate this new development of digital worlds? I know you want and you have a goal of reaching 10% of global gaming. That's right. The external goal is 10% of global gaming. Global gaming is about $200 billion market. So, you know, we did $6.8 billion in bookings last year. [04:16] 10% would be like 20. So that'd be a reasonable first step. [04:20] at where we would like to go. But really, I would say the global gaming market is emblematic of really a new way of people to communicate. In a lot of games, people are in sometimes it's the holodeck almost, you know, a 3D place where they're playing hide and go seek, or they're playing a more traditional sports game or something like that. And so really, when we think about this market, we think not just gaming, but ultimately what we might call the human co-experience market, [04:50] hollow deck, you know, type space. So we need real people, real engineers, real product people, real designers to build this, this platform out. There's a lot of interesting magnification in our products. So for all of the several thousand people we have here, there are tens, if not hundreds of thousands of people working on Roblox, making a living on the platform,

[05:20] Last year, our developer creator earned about a billion and a half [05:25] on the platform and it goes pretty deep like top thousand devs are averaging 1.3 million like real people making a live on living so it goes way beyond the walls of this building [05:40] It's. [05:40] insane. When I was going through the numbers and I was getting them checked by your team too, I was like, how is this real? A thousand... [05:48] creators are averaging 1.3 million. That's pretty insane. So when you started Roblox to today, how did you form and how did you think about building the foundation for this digital economy? When we started the company, if you can believe it, [06:04] we looked at this sci-fi market because before this company i had an educational we had an educational software company we're making educational software for young people to learn how to simulate physics literally and when they were playing with this software in addition to trying to learn physics they were trying to build stuff they were building cars and they were building buildings and all of that [06:26] So when that company was sold and me and some of the founders of that company started looking at [06:32] This gaming market. [06:34] we we started to imagine this new category really it's that sci-fi holodeck category has elements of ugc you know video youtube was coming online so ugc was big gaming was was starting to show what multiplayer 3d was about social networking was coming online so we we literally have a business

[07:04] It'd be like talking about this category, arguably not as ambitious on the size of what this category could be, but that exact same vision. [07:14] And we, you know, now we're working to that. [07:16] literally that same vision we are you know at 6.8 billion in bookings we're literally under four percent of the global gaming market the power of providing a platform where all these other creators can work on it and make a living is very powerful get a super broad [07:36] tale of content they try to make everything new types of games so it's really a [07:42] interesting play it's a fun company to run [07:46] As you think about this as a CEO, what are the metrics that you use to measure and what are you trying to hit other than the 10% goal? [07:54] Yeah, I would say it's interesting because I would say our company, [07:59] You know, there's various types of companies out there. And I'd say all the way on one end, there's almost pure imagination vision companies like a Walt Disney. And on the other end, it's just pure metrics driven growth companies. We do like to be a bit in the middle on this. And I think drive the company both by vision. [08:29] video or do they go, you know, play grow a garden together on Roblox? Like what is the future of communication? So there's invention aspect.

[08:39] and a metrics aspect. I would say some of the metrics we look at [08:43] are literally raw engagement time. So I think in Q4 of last year, it was 35 billion hours. So that's pushing 12 billion hours a month. [08:53] That's a pretty interesting stat. I think in Q4 of last year, we were at just sub 150 million daily actives. [09:01] which is a lot of people as well so i would say those are the two we we kind of watch of the [09:08] What do you think the most underrated aspect of Roblox is given that? [09:12] I think the most underrated aspect of Roblox is arguably how much deep tech and theories around economics and theories around systems sit underneath it. And so when someone sees Steal a Brain Rock or 99 Nights or Adopt Me and just says, oh, that's really fun. I want to go [09:38] play that underneath it there's enormous technical complexity there's um there's tooling so a creator could open roblox studio and make a game uh there's there's the viral aspect of you know initially a hobbyist would start building something and then they would grow to be a big studio and start hiring a lot of people there is a cloud aspect to it we run on our own cloud it's [10:08] super cost effective. We have 40 plus data centers all around the world. We have hundreds of thousands of servers. At peak times when Roblox is hitting 20 or 30 or even more million people, that's a lot of compute complexity. There are over 400 AI models, many of which we built our own

[10:32] running on this so behind that you know very spontaneous let's just go play something together there's a lot of deep tech there's a lot of also systems theory like we're we've had to design an economy we've had to design thoughtful search and discovery we've had to design thoughtful systems there's a lot of that too [10:55] One thing that I was really surprised by, by even doing this video stuff was like how much data it consumes and how much file sharing and that kind of thing, but also the usage of [11:04] from gaming is just [11:06] astronomical to your point. Like you just have tremendous amount of data centers. Have any of the AI companies come to you for advice? [11:15] Well, I would say it's really interesting, right? Because we're arguably in one of the biggest [11:20] transformative times in history and we have llms we have text-based models we now multimodal we have image we're getting into video we're getting into interactive videos and now i would say people are starting to think about okay how do we train robots how do we train um what's called npcs or non-player characters in video games it's now there's a little bit of people are starting to [11:50] The classical thing right now is rather than just trying to find video data, people are trying to find video data with ASDW or user interactions so that you can try to marry the video data with those kind of things.

[12:05] And what's interesting about the data on Roblox, the 35 billion hours of data, it's not just video data. It's literally 3D positions. It's what avatars are doing. It's what they're looking at. It's literally data that ultimately can and does allow us to reconstruct. [12:25] anything that's ever happened. And that's very valuable data. So I would say everyone's starting to think about this video game data for sure. And what's been your strategic positioning for AI with all this? I think our strategic positioning, we have several. And in addition to all of the models that are existing on our platform, I think there's two ways to think about the acceleration. [12:55] AI was available to a user is almost as if we can't imagine what that game would look like. You know, in the movie Inception, when they go down to level three or something, they're like a dream architect and they walk around and they say, fold the road over and do this kind of stuff. Or in Westworld, sometime on TV, they've got like an AI console and they design things. I think that's all going to be possible. And so users are going to be more and more. [13:23] designing in game, designing clothing, designing their own worlds by talking about it, not by using 3D tools. So I think there's this whole we don't quite know where gaming can go,

[13:36] If users have infinite AI, [13:38] On the creator side, it's really interesting right now because we're just starting to see the very early edge of what's the future of coding and coding, you know, is a somewhat mature environment when it comes to the operating system or the C compiler or the JavaScript system. And so because of that, you go through the whole stack of all these coding tools that are coming on right now. We're just starting to seem to get better and better. [14:08] running overnight iteration and all the stories we hear about you know stories that blow everyone's mind like i built this in a week or i built this in a day [14:20] we see that same thing is going to happen to the much more complex, [14:25] thing of building a video game and building a game is super complicated because it's got art. It's got what is interesting for people. What is fun? It's hard to define. Is it fun or not? So there's a bit of art. [14:42] It's got a big mix of [14:45] 3D assets and, you know, code and 3D experiences and people building cities out of, you know, 3D tools and all of that. Sorcery is brought to you by Brex, the financial stack trusted by more than 30,000 companies, including one in three venture-backed startups in the U.S. Nearly 40% of startups fail because they run out of cash. Brex is literally built to help founders avoid that.

[15:15] idle, chipping away at it with fees, Brex's designs help you spend smarter and move faster. Their all-in-one solution combines checking, treasury, and FDIC protection into one powerful account. You can send and receive money globally at lightning speeds, get 20 times the standard FDIC coverage through their partner banks, and even high yield from day one. With same day and even same hour liquidity, access your funds anytime. Companies like Scale AI, DoorDash, [15:45] HIMSS, Anthropic, Flexport, Robinhood, and Plaid. Trust and use Brex. Start today at brex.com slash sorcery. That's b-r-e-x dot com slash sorcery. Turing is training the next generation of AI with tasks that require real expertise and real world judgment. That's why companies like NVIDIA, Anthropic, Salesforce, and Gemini partner with Turing. Turing builds realistic reinforcement [16:15] systems based on real operational traces, the kind of infrastructure frontier labs need to train superintelligence. Visit turing.com slash S-O-U-R-C-E-R-Y. Are you worried at all about... [16:28] rogue autonomous agents and bots and [16:31] for threats because of AI into the whole world that you built? Yeah, it's almost as if when someone's designing a system like this, [16:42] and you think one thinks about what are all the ways we would protect against rogue bots in a virtual world which could of course should happen right like if we can imagine ai bots with a prompt from a developer going to do interesting things what we're optimistic about is the ability in parallel unrelated to those bots to literally ultimately be able to scan everything in real

[17:12] And so really, AI is just as good for monitoring protection, safety, and civility as it is for making robots. So I think we're going to be in... [17:23] pretty good shape there but when you talk about rogue bots the um and having amazingly [17:31] high quality npcs there's a there's all kinds of edgy things to think about there's like the sci-fi stuff [17:40] of what's our principle about [17:42] NPCs. [17:44] probably they're always identified like if you if you and i are in a virtual space [17:50] It probably makes sense. We just always say that's an NPC. [17:54] like have any kind of environment where we would allow NPCs to impersonate people. So I think that's a principle you'll see. [18:02] I think we will have the ability for people if they so choose to have a doppelganger NPC that could like be their agentic proxy, go out and do some stuff for them if they want it to, which will be interesting. I think from a... [18:21] developer testing of a new game like say a creator wants to they have a brand new game i'd like to test it with a thousand users and it's hard to find those thousand users they may use npcs to test their game um so there's just like a huge opportunity for that it's really exciting yeah [18:41] Film, gaming, sci-fi has always kind of been...

[18:47] the place to go to [18:49] to imagine the future. It's all about imagination. So when you built out Roblox and as you build out the future of Roblox, [18:58] What are your personal... [19:00] inspirations and where do you pull from that kind of stuff? Is it certain movies? Is it sci-fi? Where are you [19:07] getting your inspiration. [19:09] I love sci-fi movies. [19:11] I like the ones that are the most realistic, actually. So it's really hard, I think, to make a good realistic. [19:19] sci-fi movie i was trying to get my kids to watch 2001 a space odyssey like it's so slow it goes whatever [19:27] but it holds up so well today. Just like, oh my gosh, I can't believe they made that in the 60s and 70s. So I do take a little inspiration from the realism of that. And it's interesting that the Hal computer in that movie is actually was viewed at the time of that movie as just like unbelievable. And now I think that's pretty much... [19:48] a solved problem i think the um the example i think for sci-fi that i believe will happen is any kind of holodeck or simulated environment type thing so star trek they use the holodeck all the time they go in there communicate with the holodeck see other people in that feels completely realistic and i think that's where we're going with our technology we've said publicly

[20:17] Like there's still a lot of work to do here to have [20:21] 10 000 people photorealistic simulation of a rock concert looks like film you're with your friends you can dance or you can chat so that's that's where we see this going and that [20:36] I think that really ends up being not just a gaming technology, communication, entertainment. [20:42] uh co-working type technology [20:45] VR and [20:48] AR [20:49] these different types [20:52] And even all the glasses that have come out have been big promises, but they haven't really existed yet. And I remember even during COVID, people were doing these online concerts and they kind of like, and then they went away. But I know that you're having a huge emphasis in sports and entertainment now. So I think Bruno Mars, that got like [redacted address] more than that. Like it was a Guinness Book of World Records simultaneous. [21:22] Thank you. [21:22] a lab, which was just recent. So I would say those types of concerts are continuing to blossom. I would say every major sport league is on the platform right now, like NFL Universe is doing super well. And I would say the when you go backwards to the VR, AR goggles, like we're in a really interesting time, actually, like those Google Glass goggles came out like 10 years ago.

[21:52] When are they gonna work? I think they're gonna start to work. [21:55] I think it's going to be AR before VR. I think it's going to be just speaker, microphone, camera, a little heads up display. I think the challenge is you need battery and a processor. So it's maybe a phone. And so I would say there's many entrants in the space. There's the behemoths Google and Apple. They happen to have a phone as well. They have others, you know, Snap and Meta that are trying to do that without the phone and many others. [22:25] it's ultimately going to come to pass. And one of the other emphasis that you have been putting on, you've been focused on 4D. So how have you been thinking about as you build out, [22:36] these different kinds of. Yeah. So 4D is what we use. Um, [22:41] There's a lot of right now work in generating worlds with AI and worlds to become games. [22:50] typically have to have multiple people in them have interactive environments games are a lot more fun if i push on a door and like the door opens or have a car and i can jump in the car with a friend and drive around um so those are objects that [23:06] In addition to 3D, we typically call how they would look in the real world. 4D, it's kind of a name we made up and other people use it now. We think of that as a functional thing or like in the real world, we're used to things that we can interact with. It's super interesting in that since the very first day we started building Roblox, for some reason, we were really into having all of the objects in the world be functional.

[23:36] first day we made roblox um when we started making cars in roblox they were actually running in a simulator like the the wheel was on an axle there's like a power to the wheel [23:50] And in those days, there wasn't as much of that in gaming. What it made happen is that when the car would fall, like a wheel would fall off the car, it would do what you would expect it to do. It's just not like some hacky gaming objects like, yeah, the wheel fell off the car. You can have emergent behavior. So we have been, I would say, 4D from the start. And that would be both simulated objects, objects with code inside of them. [24:20] paste, [24:20] you know, anything, a wheelbarrow, a car, a building, the function tends to be built into that. [24:29] I think most of the audience might know Roblox from their children or friends. [24:35] You have quite the demographic of young individuals. How did you get that mainstay with this demographic? And how do you stay relevant through generations? I think staying relevant is super hard, right? [24:51] Like, who knows what is gaming? [24:54] for young people and who knows what's the future of gaming. There's a certain robustness to building platform that is very powerful and arguably follows a very simple specification that I believe can create more resilience for the relevancy. And so our specialization,

[25:15] spec for building Roblox since day one is pretty simple. You know, thousands of people ultimately photorealistic, real time around the world, safely communicating text and voice communication, avatar of whoever they want. That's a very resilient spec, very technically difficult. [25:37] But because it's almost like the spec, [25:40] for roblox the the parallel would be in the video space [25:44] specs pretty simple like go from 2k to 4k pixels have stereo audio at high res and then whatever comes through on that video you could say is what's relevant right now so there's a little bit of a [25:59] powerful the spec is in capturing simulation of multiple people in the real world [26:05] the more whatever the trend is in gaming. And I think what we're going to see in gaming coupled with that spec, [26:12] is the ability one other thing behind the scenes people probably don't realize [26:17] We work super hard. [26:19] at going from game experience to game. [26:23] as quickly as possible no one notices it um you know they everyone on roblox is assuming oh i'm playing this now i'm playing this now i'm playing this like half a second but everyone has come to expect that from short form video and youtube so everyone you know part of being relevant is actually technology young people [26:43] Now expect video to load every two seconds. Most gaming architectures are not designed to do that.

[26:54] But I think we could imagine that will be the expectation of young people. It's the viral factor. [27:01] uh so the viral factor is also huge right so [27:05] Roblox has primarily grown virally. Viral means word of mouth, sharing links. I was just at GDC, the Game Developers Conference last night. I was talking to some representatives [27:19] from one of the big four short form video platforms. And they said a third of the gaming content on their platform is Roblox content. So so by having users and influencers share interesting things, that does create a lot of viral pull on the platform. [27:39] I think being viral is essential to a platform like Roblox. And when creators make something really big and interesting, whether it's dress to impress, grow a garden, adopt me five years ago, that does create like viral influencer pull and people come and try it out. [27:57] Maybe this is more of a philosophical question. Oh yeah, I like that. Because you're exposed to building economies, building games, [28:06] Building like [28:07] pretty much like careers for people because they're making a tremendous amount of money off this platform. What have you learned about human behavior? [28:16] I'd say I'm pretty optimistic about human behavior. So what I would say, [28:21] One thing I've learned is...

[28:25] enormous responsibility of, um, [28:28] and having all these people making a living, having 150 million people play, like that's an enormous responsibility. One of our values is respect the community. And really early on, we would share with employees, like just imagine one young person who lost their Robux. [28:48] like how big of a thing that is. When we started the company, the four of the founders, we used to do like the customer service and the moderation queue. And I think one of the things we learned is [29:01] um we have enormous impact on people's lives and it's very dangerous to clump it like 150 million million you know that's very different than imagining a single person who lost their robux like that's [29:17] pretty big so the other thing i've learned we would say we've learned about human behavior is that very i think simple economic principles very well designed very fair can create incredibly complicated [29:34] outcomes um you know roblox is pretty simple you come on board you hear about it you like it play with roblox studio [29:44] you're a hobbyist you make something you publish it you start making ten dollars a month [29:51] That's kind of motivating. Then you make 20. That's cool. Maybe you make a thousand and you say that's going to be like a full time job. So I think we before we launched the economy, we had a lot of discussions actually about should people even make money on Roblox. And I was an advocate for it.

[30:13] There's a lot of people that were against it, but the logic was ultimately people should be able to make a living doing this. And if they can make a living, they'll make higher quality content. So. [30:25] I think we've learned [30:28] Very simple economic principles can drive crazy, complicated results. People start a company, they hire people, they buy and sell their company, they want to learn more about us, they give us feedback, they have a company making $50 million. [30:48] I think that's another learning. [30:50] Well, one of the biggest scares with AI is that no one's going to have any jobs and then we're all going to be worthless and we won't have any. This is a key. This is like the key question. It's the key question. This is how did you know the key question? [31:02] I don't know. This is the question I was asking. I was asking everyone this question last night at a dinner. [31:11] with seven other ai expert people and the question was imagining ai is super powered in 50 years what percent of adults work eight hour jobs a day [31:26] and no one has the answer. [31:28] um but my i guess my belief is that [31:33] for many many people [31:36] accomplishing something is really a big important part in their life and so it may be very dangerous if

[31:46] everyone can just go do their hobbies like um so then some of the very smart people we were talking about [31:53] We may as a culture, like if we have infinite AI and robots and we can all decide that part of our culture may be inventing. [32:03] accomplishment so that people can feel accomplished if they so choose [32:09] You created a currency. Yep. You mentioned a couple values before our conversation and during. Yeah. What are your value systems? I know it's very important to you. [32:20] I don't know if you've heard the story of what's like a really good way for a company to create their value system. And it's a little bit like a good way to create the path system in a university. Like, you know, if you're creating the path system, you don't put down the paths. You let people walk around for a couple of years. You see where they're putting the trail. Then you build all the cement paths. And it's like perfectly optimal. [32:50] without a value system and then we had a [32:53] capture what is working. [32:55] and capture what has gotten us launched and what is working so far and we came out essentially with the values we have today like we um they're somewhat [33:08] orthogonal in that one is take the long view, but the others get stuff done. And so we like to imagine that as sometimes a fun chart where you're taking the long view and iterating every day. That's a very safe thing. Like, you know where you're going to go every day. It's just a going in the right direction. Every other combination is arguably super dangerous.

[33:35] taking the long view without getting stuff done that's kind of like the three-year research project that can never ship [33:45] um getting stuff done without taking the long view is a little bit like well we don't know exactly what we're doing but we're just going to worry and then there's the um radically dystopian quadrant of not taking the long view and not getting stuff done which [34:01] is like out of business in a second so i would say those two [34:06] um and then we um the other one is uh respect the community [34:10] We translate that sometimes into putting community ahead of company, ahead of team. Like that's what we're thinking about. [34:20] And then the final one is we are responsible. [34:24] like personal responsibility is is a way to scale a lot of culture so we just have four they're very simple [34:31] We also have a set of principles. One of our principles we were just discussing is as much as possible when we're talking about a team, we say our team, not my team. [34:44] Why is that so important? [34:46] It's really interesting, you know, saying if you if you were listening to me, [34:53] and I was talking about our executive team, [34:57] um it has a different emotional impact like my executive team's awesome [35:04] Okay. [35:05] as opposed to our executive team is awesome.

[35:08] I think that just as you use a little bit more we than I, I think we, it's literally one of our principles in the company. [35:16] VCX by Fundrise, the public ticker for private tech, allowing investors of all sizes to invest in venture capital. View the portfolio at GetVCX.com. That's GetVCX.com. Some of you may not have heard this yet, but our sponsor Public just launched something called Generated Assets, and it brings AI into investing in a way I've honestly never seen before. Here's how it works. You type in an idea like AI-powered supply chain companies [35:46] or defense tech companies growing revenue over 25% year over year. Publix AI then dispatches a swarm of agents that scan every single US stock, evaluates them, and instantly builds a custom index around your thesis. What really stands out is how clearly it explains why each stock is included. And before you invest, you can even backtest your idea against the S&P 500, so you're making decisions with real context, not just guessing. And beyond generated assets, Publix lets you invest in stocks, bonds, options, crypto, [36:16] They'll even give you an uncapped 1% match when you transfer your investments over from another platform. If you want to build a portfolio that actually reflects your thesis, visit public.com slash sorcery. [36:27] Paid for by public investing. Full disclosures in the description. [36:31] Enterprise AI runs on Merge, the AI infra platform for integrations, agent tooling, and model orchestration, so your teams ship product, not plumbing.

[36:40] Mistral, Dropbox, and Drada already trust Merge and production. [36:44] Start building at merge.dev. Founders ship faster on deal. Set up payroll for any country in minutes, hire anyone anywhere, get visas handled fast, and get back to building. Visit deel.com slash sorcery. That's deel.com slash s-o-u-r-c-e-r-y. [37:03] I'm curious, one of our sponsors is Brex and they're all about performance. For you, how do you think about your personal performance, whether it pertains to health, relationships, relationships? [37:15] What keeps you motivated? [37:17] I feel I've been very lucky to be dropped into this world and this life in a way. [37:25] and i think i'm lucky to have my job so i i literally try [37:31] more and more to treat being CEO [37:35] as a craft literally and it's um it's surprising i think it's a craft that a [37:41] like requires [37:43] like fitness and mental health and like a whole structure around it you know [37:50] figure skating is a craft being a chess player is a craft but i also think being a ceo is a craft so i am literally constantly trying to [37:59] think about how I can be better at my craft interpersonally as well. [38:05] I saw there's a lot of labels on the snacks. Do you like those labels? I did. They were very interesting. Really? They're implicitly educational.

[38:14] Tell me more about the labels. So we're [38:18] I don't know if you looked for all the combinations of the labels. I saw two combinations. Yeah, there's four combinations. So just so for everyone out there, right? So we have Whole Foods, [38:31] and good energy. [38:33] and so where this is coming from on the side uh family philanthropy we work on metabolic health and we work on research around how does diet [38:45] affects all kinds of things like bipolar schizophrenia how clearly we think so this is led into [38:54] hopefully in a positive way, I think so. Like using Roblox as a little bit of a trial site for some of this. Our employees get CGMs, for example, so they can watch their blood sugar, kind of monitor that. [39:10] The two parts of the snacks that we tried to use were everyone understands the whole food access. Some things are whole foods like an apple. Some things are not like some processed Cheetos that we're talking about. And then the other access is like good energy. And, you know, we pick them because that's been discussed in the press. Actually, the U.S. food pyramid, you know, we remember how it's been inverted a bit. [39:40] that was a little bit inverted towards whole food and good energy. So like that that boiled egg in our snack room,

[39:48] is both whole foods and good energy wow so you probably saw like two check marks yeah if you went in and saw that and so the idea is everyone can go in and see all the snacks some are whole foods some are good energy some are both some are neither so a can of coke would probably be neither right high sugar not a whole food um that's and then people can just learn about it and decide [40:18] diet a bit. I feel better, feel sharper, have a lot of energy at four o'clock in the afternoon without coffee, stuff like that. What's your favorite biohack? [40:29] Ooh, I, um, so I have a, thanks for asking. You're welcome. So, um, I got a couple, there's a new thing called a CKM, not a CGM. I was talking about on the Tim Ferriss podcast, which is a ketone monitor, not just a glucose monitor. So, uh, going into ketosis, um, when you eat fewer carbs, you know, sub 20, sub 10 grams, little more fat, you start burning fat, which is [40:59] energy pathway, your ketones start going up. And it's actually, I think, a sign you're getting good, consistent energy to your brain. It's part of what our Family Foundation researches is going into ketosis, [41:15] for bipolar and schizophrenia. And we're finding more and more

[41:22] people who tend to have these medical issues this can be like a a partial treatment for them um so it's actually pretty amazing um so i would say one biohack would be get a ckm [41:37] pretty interesting the other biohack i would say is [41:41] i um sometimes i wake up in the middle of the night and i would say younger years i would worry about it like oh i've got insomnia now i i have figured out no matter what [41:55] If I just let myself sleep, even if I wake up in the middle of night, I get a few hours there. [42:00] I've watched myself. I mean, like on five or six days, Hey, I'm okay. So then now if I wake up in the middle of the night, yeah, I don't know, like listen to something, read something, go back to sleep, just not worry about it. Did you ever monitor all of your health? Like, [42:15] really precisely and now you've kind of [42:18] gone into monk like yogi mode i i could you could argue that i'm a little bit [42:24] fastidious about what i eat like i have a phlebotomist come to my house once a month and do like a blood draw [42:30] Because just like I get my full blood chart every month, it's actually not a big deal. [42:36] Like it's easier than going to Quest, right? But what are you looking for? I'm just watching metabolic markers, you know, insulin resistance, glucose, all of those kind of things. It's gotten to now it's just pretty consistent. [42:51] But it's like a habit.

[42:53] It's probably the future, actually, because I'm in a unique position where I can organize to have someone and do my labs every month. I do think ultimately between devices... [43:04] and all of that like everyone on earth will probably have that data all the time. [43:09] I recently got a Prunuvo scan. [43:12] I'm interviewing Prunuvo CEO on Friday. [43:16] And so we're going to talk through that. We're going to go through it. It's just like phenomenal technology. It's so interesting. Have you ever gotten one of those scans? I've gotten all kinds of scans. So DEXA, you know, non-invasive, going in the big machine and doing those, all of those. So I get a lot of those. I monitor the radiation levels. I'm not like going in every day or every week. So it's kind of balanced, but it's super helpful, right? Like we're starting the, for cardio, you know, [43:46] comment on medical advice, but you know, CAC scan or calcium scan used to be the standard. There's more and more ability to now 3D image, like all of your arteries and make an estimate on that. So it's like, we're kind of in the midst of a whole new era. [44:02] Would you say that [44:04] metabolic health is the most important thing that people should watch out for. [44:08] I would say, um, [44:11] if people want to move the needle [44:14] the food we eat has a lot more power [44:18] than we would ever [44:19] imagine and then when we think about it why wouldn't it right we

[44:24] we walk around all day we sleep we get up but what we eat is can be very different there's a little bit of an argument that what we eat now is very different than 10 000 years ago you know there there were diets 10 000 years ago there was a little bit of an agricultural revolution [44:41] where we discovered agriculture and started upping some of the things we eat relative to that i would also say in the last hundred years there's been a little bit of a sugar and seed oil revolution as well so the things i personally [44:56] watch out for his sugar and seed oils. [45:01] I think it's really cool. And I was talking to your team earlier about this because the gaming community doesn't have the best rep for health, but you're a very health oriented gaming company. Oh, yeah. [45:13] What's really interesting is, I don't know if it's a coincidence or not, [45:18] but you could imagine my family is a little like they've seen roadblocks for 20 years so my kids are like [45:25] yeah cynical whatever um in a good way like in a powerful way but what's been fun is even on the creation side um [45:36] a year ago was dressed to impress like think of it that's not a traditional game it's a fashion game competing on the runway whoa [45:46] Um, grow a garden. Oh my gosh. Like there's a top game on Roblox about growing a garden. And in our backyard, we have like a big garden that one of my daughters has planted and they're really into that. So it has been kind of fun, not just on the health side, but the diversity of the content that's popped up. What's the craziest thing that's ever happened here?

[46:08] Oh, man. The crazy, I think the good crazy or bad crazy. I'll just share the good probably. Um, [46:18] the craziest things that have ever happened [46:22] You know, we've taken the company to Disneyland. We had a certain thing. We marched a marching band through the whole campus down the halls just because we hit something fun and always had the fantasy of leading a marching band. You? Oh, yeah. Really? Why? I don't know. Just like. [46:41] You want to be a marching band leader? You know, you have like that thing. The... [46:46] baton and all that so i got to do that that was fun [46:50] I would say that one of the craziest things was very early on in the company. When we started the company, we had launched Roblox, but we had not launched Roblox Studio, where people can build games. So we were a very small team building stuff internally. [47:09] And we're like, what's going to happen when we launch Roblox Studio? Well, like anyone use it. [47:15] And then within like two hours, just hundreds of mind boggling things happened. That was pretty crazy. And another one was when we launched the economy. There used to be a time when there was no economy on Roblox. And when we launched Robux and the ability for creators to make a living, had that same thing. What's going to happen? And we knew within six or eight hours, like, okay, this is going to

[47:42] take off. So those were really good crazy. As you look forward to the next 10 years, what is the vision of Roblox? What are you most excited for? It's fun because it's a very technical vision. [47:57] like what we're trying to build is really hard so there's a research aspect i would say not just on the ai but the ai intersection with cloud with 3d with rendering with physical simulation i'm excited about that i'm excited [48:13] I would say it's [48:15] always exciting for us to see some new [48:19] type of game that we've just never imagined that someone comes up with some new mechanic, which is fun. I think it's always fun to see. [48:30] esoteric, niche games like about model railroading or things like that, that are fun hobbies. So I think it's that's all very rewarding. [48:41] Amazing. Thank you so much. Thank you for coming in. Appreciate it. [48:45] It's an awesome HQ and huge campus. Good. Thank you. I can't wait to find the secret underground tunnels and simulation rooms. We're discussing either underground tunnels or sky bridges to connect the buildings. We need some slides, too. Okay. From building to building. Yeah. Cool. Thanks. [49:05] Thank you. Appreciate it. Of course. Hey, it's Molly. If you enjoy our interviews, check out our newsletter, sorcery.vc, where we deliver a once a week top deals and tech headlines email, and also go deeper on our podcast interviews. Subscribe to Sorcery today. And don't forget to subscribe to the podcast on YouTube, Spotify, Apple, or wherever you listen. Link in description to sign up.