DAVID BASZUCKI: Time and time again, I think we try to solve things or figure them out ourselves rather than take a step back and think through, could the community do this better than us if we gave them the tools and the functionality?

KEVIN SCOTT: Hi, everyone. Welcome to Behind the Tech. I’m your host, Kevin Scott, Chief Technology Officer for Microsoft.

In this podcast, we’re going to get behind the tech. We’ll talk with some of the people who have made our modern tech world possible and understand what motivated them to create what they did. So, join me to maybe learn a little bit about the history of computing and get a few behind-the-scenes insights into what’s happening today. Stick around.

CHRISTINA WARREN: Hello and welcome back to Behind the Tech. I’m your host, Christina Warren, senior developer advocate at GitHub.

KEVIN SCOTT: And I’m Kevin Scott.

CHRISTINA WARREN: And this is actually like one of our first video episodes. So if – if you want to give us a watch, I guess, you can follow us on YouTube.

KEVIN SCOTT: Yeah. Excellent.

CHRISTINA WARREN: Yeah, it’s good stuff.

Okay. So we’re super excited today because our guest this week is David Baszucki, who is the co-founder and the CEO of Roblox, which is a game creation platform that allows users to build their own games. And it has over 50 million users, which is just stunning to think about.

KEVIN SCOTT: It is, it’s a stunning number of daily active users on that platform. And, you know, two of the daily active users are my kids, who at 11 and 13 have been Roblox enthusiasts for a while. And it was – you know, honestly it was through my kids that I really got to an understanding of how big a platform Roblox was and the sort of impact it was having in the world. So I feel really lucky to be able to chat with David today and I’m going to be a hero to my kids whenever this show airs.

CHRISTINA WARREN: Yeah, no, I was going to say I was like, you now get to go back to your kids and say, “I just talked to the head of Roblox,” and they’re gonna be like, “Oh my God.” Like that’s – like all the other very important and famous people that you talk to all the time, it’s like, nah, not a big deal, but you’re like, I know the Roblox guy. All of a sudden, hero.

KEVIN SCOTT: Yeah, 100%.

CHRISTINA WARREN: All right. Well, let’s now get into Kevin’s conversation with David.

KEVIN SCOTT: David Baszucki is an entrepreneur, engineer, and inventor. He’s the founder and CEO of Roblox, an online game creation platform that allows users to design their own games and play different kinds of games created by others. Roblox has helped millions of up-and-coming game developers learn the basics of their domain and work on their talents.

Welcome to the show, David.

DAVID BASZUCKI: Kevin, it’s a real pleasure to be here. Thank you for having me on the podcast.

KEVIN SCOTT: Yeah. And thanks for being here, too. You and I have chatted about this before, but my kids are going to think I’m a superhero for having you on the show. They spend more time in Roblox than probably any other piece of software that they use.

DAVID BASZUCKI: Well, that’s awesome. I think they should think you’re a superhero because you design and build incredible infrastructure and think about technical things. So that’s really why you’re a superhero.

But, hey, it’s great to hear that, and yeah, we like to think when parents share that their kids are spending a lot of time in Roblox, we like to think that’s a lot like the time they might spend on the phone with their friends or the time they might spend creating a level in Dungeons and Dragons with their friend or spending time playing hide and go seek when they can’t play it outside. So we’re actually not always trying to optimize that engagement time, but we do want that engagement time to be socially connected.

KEVIN SCOTT: Yeah, I think that’s a – it’s an interesting, interesting thing. My wife and I had, for a very long while, been really sensitive to the amount of screen time that our kids were getting. And then, you know, we had this epiphany during the pandemic that Roblox was actually one of the places where our kids were able to get the socialization that they weren’t getting because of all of the lockdown stuff that was going on. So they would jump into Roblox and play with their friends. It was their digital playground, so to speak, and they’d be on, you know, Facetime on the side and like having a conversation. And like in a certain sense it was life-saving. I don’t know what they would have done to have maintained all of that social connectivity if they hadn’t had Roblox.

DAVID BASZUCKI: Hey, that’s cool to hear and that aligns a bit with our thinking about this type of technology as a utility that can host co-experience and connection, like the telephone system, like what we do in real life, even like the U.S. mail system at time.

And I think when we talk about social networking technology and social technology, I don’t think a lot of – you know, there’s – there’s a lot of nuance on it. Some social networking technology is more about sharing images or sharing video or consuming things statically. And then what’s cool about Roblox, I think is it’s much more of a real time communication connection type technology, and inherently real time connection technology I think feels different than, you know, solo consumption type technology.

KEVIN SCOTT: Yeah, for sure.

So you know, before we get into like all of the many interesting things that we can chat about, about Roblox, I’d love to just back up a little bit and ask you what I ask everyone who comes on the podcast, which is, you know, what was your life like growing up and how did you get interested in technology?

DAVID BASZUCKI: Yeah, I had a bit of a wonderful life growing up in the far out suburbs of Minneapolis, where we had sand pits and dirt bikes and, you know, things to build and go-karts and a lot of stuff to do.

But as part of that fun building thing, early on in my younger years, our school had a computer lab. The computer lab was a teletype with a paper tape. And we got to – I remember writing my first BASIC program, which was trying to sort five numbers and storing my work on paper tape.

Then my dad got us all an Apple II and as I was programing this Apple II and reading the manuals and it was the time when you had, you know, all these Apple II magazines were out and you’d enter this stuff, for some reason that Apple II had an assembler and a disassembler built into it. And there’s just something fun about, okay, my visual, my BASIC program’s running slow, but there’s this assembler thing.

And the Apple IIs also had this wonderful facility where you could hook up some potentiometers from RadioShack to the analog inputs and start getting those. So that combination, I think, got me interested in this and kind of that was the germination really.

KEVIN SCOTT: Yeah. And I mean, our technology that we have today is certainly amazing, but I do think this thing that you just mentioned, which is the fact that these early machines were so approachably programable, like when you – I just remember walking up to something like a Commodore 64 in a department store, for instance, and you would be sitting there at a BASIC command prompt. And so, like, the thing is inviting you to type code into it as your first interaction that you’re having, which is sort of amazing.

DAVID BASZUCKI: Yeah. It’s almost as if the user interface in the days was the command prompt and the user interface was entering code almost, which is it’s a lot harder to find a command prompt now on a modern day computer.

KEVIN SCOTT: Yeah, it indeed it is.

So did you have anyone around you when you were a kid? So you had this computer lab at school, but did you – was it mostly learning from these magazines and sort of teaching yourself?

DAVID BASZUCKI: Yeah.

KEVIN SCOTT: Or did you have anyone around you?

DAVID BASZUCKI: It was very self-learned, for some reason. There was a very rich up-and-coming PC ecosystem, a lot of magazines to read, to follow. I was a little bit of a – I remember we had a computer lab and I would make programs to make all the computers in the computer lab do funny things or, you know, try to play around with the people who are running the computer lab. But no, I was pretty autonomous. When I went out to university, you know, there’s a little bit more of that, but a lot, definitely a lot of personal self-discovery back in those days.

KEVIN SCOTT: And – and was gaming or writing games one of the early things that captured your imagination with computers –

DAVID BASZUCKI: Yeah.

KEVIN SCOTT: – or did that come later?

DAVID BASZUCKI: No. I remember writing, you know, an assembler with the very primitive shape rotation graphics the Apple II had, you know, versions of Asteroids and things like that, just to see what it felt like.

And there was one game on the Apple II, it was called Robot War, that had a very simple programing language. And you’d program your robots and they would go do battle. And I remember my brother and I just spent hours and hours on that. So there was an interest – there’s kind of an intersection of gaming, programing, user-generated content that was all interesting to me.

KEVIN SCOTT: Super cool. And how did you get from Minneapolis to Stanford?

DAVID BASZUCKI: Yeah. Pretty naively, reading magazines and what’s a good school for computer science kids. Where I was growing up, we didn’t really know. And it was Stanford and MIT, and luckily it was Stanford. My dad took me out. We visited and it’s an amazing campus. So it was a fairly non-sophisticated trajectory.

KEVIN SCOTT: And so you’re at Stanford. I’m guessing you and I are about the same age. So it was probably like a very interesting time to be at Stanford. And so when you graduated, there was probably like all of this super cool stuff going on. You know, the Internet boom was, I’m guessing, about to happen when you graduated, not quite there, but about. So how did you decide what to go work on after you graduated?

DAVID BASZUCKI: Yeah, I was pretty – when I graduated, I was pretty shy and pretty unsophisticated. And knowing what I know now, you know, there were all these companies I really would have liked: Microsoft, Apple. They were all coming along. But I was very much going through the standard channels and the college campus recruiting center and all of that stuff. So I ended up getting a couple, I would say, much less optimal jobs, where I wasn’t – like I wasn’t getting to build interesting stuff and things like that. So my first two jobs were actually a real disaster, and I thought, “Oh my gosh, like, what am I going to do?”

The things that finally started happening for me when I took some time off after two, two and a half years, and just started writing Macintosh educational software. And that’s – that was kind of, I had to invent it on my own.

If I was more sophisticated, I would have gone and camped out at Microsoft or Apple and just said, “You must hire me,” tried to meet all of those people, but I wasn’t. So I resorted to building this first version of Interactive Physics, and that’s what kind of got me back into the stuff I really like.

KEVIN SCOTT: Well, so tell us a little bit about that. So this is, you know, both you doing something interesting programming-wise, but you’re starting a business, too. And so like those are two very, very interesting things.

DAVID BASZUCKI: Yeah, it was driven a lot by the love of constructive software. And in those days, in the mid- to late ‘80s, educational software was very early HyperCard-y, templatized, pre-canned, slider type stuff.

And I was looking at the whole educational software market, and it’s interesting when you’re on the outside of something, it seems like it must be awesome and great. But then when you go and start playing with it yourselves, the ideas start flowing. And I say, “Oh my gosh, this is all pre-canned.” What would – wouldn’t it be cool if there is a way to build all of these things and try anything and build physics experiments?

So that launched the idea of Interactive Physics. It was kind of a business, but I got to tell you, educational software is a hard business and – but I went down the path anyways. I remember meeting with a friend of my dad’s, who was very sophisticated, and he said, “Look, I think the overall size of this whole market is X,” and it was a very small market. And I kind of said to myself, “That’s impossible, it can’t be that small.” But yes, it did turn out to be exactly that size about four or five years later.

But nonetheless, yeah, we went and I did a lot of research on physics simulation. Back then, this was very early, you know, SIGGRAPH type stuff, 2D, simulation, was able to put together this physics simulator and build this thing that was called Interactive Physics, and it was really exciting. All the physics textbooks correlated to it. A lot of teachers around the – got to be millions and millions of students. So it was a great launching.

KEVIN SCOTT: Yeah. And so, it – I just sort of wonder, like, you know, again, you’ve got two things here at least that you’re doing, that at least on the surface sound like they’d be relevant to the founding of Roblox. So like physics is an important thing for videogames and then just whatever lessons you learned, you know, of succeeding or failing at business is sort of a great workshop for starting your next business.

DAVID BASZUCKI: That’s right.

KEVIN SCOTT: So was that actually the case?

DAVID BASZUCKI: Yeah, we with Interactive Physics and the company was called Knowledge Revolution, we made what I’d call a third mistake, which was rather than thinking about our end users, who we loved, and the market and consumer, we said, “Let’s take this physics engine and make it into something that mechanical engineers are going to use,” which was a little more logical. They’re a great market, too, but once again, not a huge market. And it takes a lot more technology and a lot less consumer polish for those types of products.

That, we finally were acquired by a company that makes mechanical engineering software and I worked there for a few years, but the notion was always come back to consumer and there was a notion that the best educational tools sometimes aren’t educational tools. They’re a really good word processor. They’re a really good movie editor. They’re a really good web browser. And that those types of things can go super high quality, super high volume, and be educational tools that tens of millions of people will use.

So the learning from Interactive Physics was, let’s think about going more consumer, let’s go 3D instead of 2D, let’s go multiplayer, let’s go cloud-based, let’s go avatar-based, so rather than a 2D experiment that you’re watching by yourself, you’re inside of a 3D experiment with your friends, you know, around the world on multiple devices.

So that Interactive Physics thought was still there, but the hope was if there’s thousands of engineers working on super high quality, free consumer software, the byproduct of that may be an educational product that’s even better than Interactive Physics.

KEVIN SCOTT: And so you started Roblox in 2006, right? So what was it like at the beginning? So in 2006, just to remind folks like it, you know, on the one hand, at 16 years ago, it seems like a long while. On the other, like maybe not long at all. But in ‘06, there was no iPhone, no iPad, you know, just for instance.

DAVID BASZUCKI: That’s right. We actually started writing code in 2003.

One thing that was there was Visual Studio, so cha-ching, really nice. And so, we started writing on Visual Studio on Windows. It was very early on. It was the days when Flash was big, when anything interactive has to be in Flash or no one’s going to ever install it. It was before the – you know, we took the controversial notion, no, we want an executable, we want the performance, all of that stuff. It was, you know, 3D graphics was somewhat standardized with DirectX and OpenGL. It was early on.

But at the same time, in the midst of all of this metaverse chat and discussion, even in 2005 and ‘06, we saw There.com and Second Life coming out. So, you know, it’s a little highlight that these ideas have been around for a long time.

Definitely no iPhone, definitely PC-based only, but a lot of the visions that we had, new category, immersive 3D co-experience, elements of gaming, social networking, toys, media, yes, yes, yes. Has to be cloud-based. Yes, yes, yes. Those core pieces were all there.

The early days were Erik and myself writing code in Menlo Park for about a year and a half or, you know, to get the first version of this thing going and doing a lot of the key stuff we have in Roblox now, some of that had that origin, you know, Roblox Studio, the cloud, all of that. We were one of the first S3 customers. So yeah, a lot of fun stories going back there.

KEVIN SCOTT: So what were some of the like big challenges just starting that business that you all faced?

DAVID BASZUCKI: So one challenge was this was, you know, we didn’t know anything about the gaming or consumer space. We – and I think that was good in a way. We were thinking way out of the box and not taking anything for granted. We were comfortable having our product function well, even if it didn’t look well. We were big fans of Craigslist because it was very simple and minimalist and high performance. And so, we’re very function based.

I think we picked up tidbits along the way. One was we learned to experiment from ground zero. We would buy very small amounts of AdWords traffic. We’d buy 100 users for $100, and we could kind of see them come on the platform and in – you know, it’s an interesting mathematical exercise, but you could probably prove that every Roblox player today has a word of mouth history going all the way back to some original thousand AdWords people that we bought that were like in some Darwinian evolutionary the chokepoint to the foundation of Roblox.

KEVIN SCOTT: That’s interesting.

DAVID BASZUCKI: Yeah. So we would buy a few of these users. Then we started to see the virality. We would hang out with them.

The challenges I think were always around having to build a viral platform. We never – buying traffic was always way too expensive to scale a business like this. So it had to be socially viral.

I think some of the challenges we learned along the way, in those days, there were a lot of, quote, you know, “Web 1.0 viral techniques” of pushing links and doing this and, you know, getting people to check in and all of that. And we got distracted a bit by that because in the end, just true word of mouth, just come here to be with your friend was more powerful than a lot of those things.

And then I’d say a lot of learning along the way, right off the bat on safety and civility. Because we were in there with everyone on the platform, we knew within probably the first three weeks of launching Roblox, oh my gosh, like, we got to get all over this.

And my partner, to his credence, Erik, said, “No, like we have to build the first moderation safety system right now” and that was a really good learning. We actually when there were four of us in the company, we were on a daily rotation where the four of us were running the moderation and safety queue and, you know, kind of getting a feel for this before we hired other people. So that learning came very early.

And some other really good learnings we had early on, one of the learnings was very early on, back then there was this thing called forums which, you know, that’s very 19 – you know, early 2000s kind of stuff. We would – besides interacting with all the players, we’d interact on the forums.

And what was really a good learning was anything we did, half the people on the forums were going to quit Roblox. You know, I don’t care how good it was or how creative it was, when we added better animation, when we added clothing, there was just so much passion around the community, that anything, any enhancements from the retro original Roblox was like a big deal. So that was also good learning, to kind of balance the community.

KEVIN SCOTT: And so, like I’m curious about that in particular. So, you know, obviously when you’re making an improvement or enhancement to the product and like half of the people are threatening to quit or are actually quitting, like if that happens monotonically, you go to zero. So you obviously –

DAVID BASZUCKI: They never quit. They never quit. (Laughter.)

KEVIN SCOTT: Well, and I’m guessing that a bunch of these enhancements are like things that drove growth, right?

DAVID BASZUCKI: That’s – yes, they were always things that drove growth. It kind of highlighted the passion of the community and how much people wanted to be involved with that.

And that, I think, also taught us to respect the community and to start, you know, coming together with them, holding developer conferences, all of that. And we ended up with a really good balance where we have to drive part of the product with our own vision, and at the same time, we have to be listening to the community continuously, dev relations all over the place, because they’re very early warning thing around places where we may be going astray. So it’s a great balance. We have an amazing, passionate community and it has evolved into an amazing creator developer community as well.

KEVIN SCOTT: Well, so that’s a good segue into the next thing I wanted to chat about. The thing that I didn’t understand about Roblox when I was first watching my kids play it is I didn’t understand that Roblox was a platform for creating games. I thought it was a game.

And so I would see my kids. I’d ask them, “What you’re doing?” “Oh, we’re playing, you know, X or Y or Z.” And I would watch them and I was like, “Oh, this looks like Roblox.” And like, what – what made you all decide that the thing to do here was to create this platform versus like making a single game with a big, expansive world?

DAVID BASZUCKI: I think Erik and I inherently didn’t see ourselves as content creator capable. It’s almost like we had a bunch of ideas for things, but we trusted more our ability to build the technology and platform than those things themselves.

And I think we also loved the combinatorial possibility of building tools that if we were to build them right – it’s interesting, right? If we’re going to build this open-ended tool in Roblox Studio and a platform, it’s– there’s a such a wide range of what might happen. No one might use it. Okay, boom. Or many people might build stuff we’ve never seen before and we could never imagine. And so that variability of outcome and that risk was exciting to us.

And when we – I remember launching – when we launched Roblox, we at times would have 6 to 20 people online using an experience that we had built as a prototype place. It was called Crossroads. And we were not in any way viral. We knew we wanted to launch a Roblox Studio and see what other people would build, and so it was a bit scary. Okay, we’re going to launch Roblox Studio. If they don’t build really good stuff, like we’re not going to be viral, this isn’t going to go anywhere, but there’s a chance they’re going to build really interesting stuff. So let’s see what happens.

So the day we launched Roblox Studio, it’s kind of those days you remember in your company where, oh my gosh, five creations, 10, 20. This one’s pretty interesting. Oh my gosh, I can’t believe that. And so, you know, we knew that that bet had paid off. It is really good learning. Time and time again, I think we try to solve things or figure them out ourselves rather than take a step back and think through, could the community do this better than us if we gave them the tools and the functionality? And there’s been many things we’ve started doing that we stopped and tried to push back to the community from that learning.

KEVIN SCOTT: Yeah, that’s sort of an amazing perspective to have. And I think one of the really nice things about your platform is that you sort of have the interest of all of the parties aligned. So if you don’t have great content creators building on your platform, then the platform itself, you know, declines in utility to your users. And if the content creators can’t come to your platform and get a bunch of, you know, engaged, happy, excited users playing there or interacting with their content, then, you know, like they’re not going to come to your platform anymore. So are there other things other than this, you know, humility about what you’re great at versus what you’re going to trust the content creators to be great at philosophically that you have thought about in terms of nurturing this platform?

DAVID BASZUCKI: I think the other one is, you know, we have a few key values. One is respecting the community and trusting the community. The other is taking the long view on this. And that goes hand in hand because whether it’s Pixar or Roblox or, you know, Disney in the early days or whatever, time and time again, we get confronted with places we need to move the technology where there’s decisions around very obvious short-term things that might be a little more tactical. And then looming out there, there is sometimes, whoa, that’s really hard. That could take us a year or two.

We at least want to talk about it like, whoa, well, how hard is it? Could today’s computer bandwidth, GPU, CPU maybe do that? Okay. If that really hard thing worked, though, in three or four years, would we be better off or not? Oh, yeah, we might actually be better off with that really hard thing than this stack of more obvious things. Oh, my gosh, but then we may have to, you know, may be a year or two out.

So I think trying to confront those things time and time again is super critical. And it’s super difficult to teach and learn from each other when to confront those things. But I think we’ve had several examples of those.

And I think trying to create a culture that does that, which is kind of what you might call a culture of innovation, is very, very hard, very difficult, but I think it’s a key component of the future value of any company. Like, how much can we be inventing stuff?

KEVIN SCOTT: Yeah, look, I think for any of the listeners who have ever worked at a technology company or like been a leader in a technology company, they will understand the deep truth in what you just said. And for everybody else, like I just can’t even address how hard a problem this is.

DAVID BASZUCKI: Yeah.

KEVIN SCOTT: Because there are like all sorts of things. You know, we – sometimes we call it Innovator’s Dilemma. You know, sometimes it’s momentum, sometimes. But there’s so many things that are trying to drag you into the short term when, I mean, we just sort of know it’s almost axiomatic that, you know, innovation is doing something different than what you’re doing right now. And like, we know that innovation is the thing that creates all of the interesting value in the world. So doing that well is just tough, man.

DAVID BASZUCKI: It’s so tough. It’s a lot of the reason why small companies with three to six people, who have had the leisure of, okay, let’s take six months off, let’s try this crazy thing, I mean, that six months off of hanging with your three or six other people is a place where deep innovation can happen, you know, where anything is possible.

And trying to make sure inside of a company with 2,000 people, trying to figure out how to have that happen as well through the feedback loops, how we review teams, whether teams feel comfortable proposing those long-term things is really interesting.

One of the principles we’ve kind of come across is, is it possible to both have that really long pull thing, which is going to be really hard, but break it down into small steps where you can actually build it, you know, in a hundred small pieces rather than drop shipping something two years out. And so, when those are possible, that makes things a lot more fun because at least you can see the progress to the big thing.

KEVIN SCOTT: Yeah. Well, I mean, another interesting thing that you all must feel is we live, at least in Silicon Valley, in an eco – and increasingly, I think this is like everywhere, we live in an ecosystem where if you’re really entrepreneurial and like really determined to get something done and you’re working for a company that won’t let you do the interesting, hard things, you just sort of quit, go raise some money, and build the thing yourself. And – but those are exactly the people that you want helping you find the next version of your company. And so, building an environment for them where they can do that hard, ambitious work, you know, for Roblox or for Microsoft or you know, for whomever their employer is, is like just so important.

DAVID BASZUCKI: Yeah. A really optimistic way of looking at it might be by doing the work to focus on innovation because we believe it’s so important, hopefully it actually makes it a more retentive place for extremely creative, innovative people.

KEVIN SCOTT: Right. And how do you get people to focus? Because, I mean, one of the things with innovation, you’ve sort of seen this in the crucible of being a startup, is not every hard thing that you can do is a useful, hard thing.

DAVID BASZUCKI: That’s right.

KEVIN SCOTT: And so there has to be some editing process or something in the ecosystem that helps you sort of say, okay, well, I’m – I’m getting signal here like, you know, this thing is good and this thing, not so much.

DAVID BASZUCKI: I think one thing to look for here is really hard things, when looked at from afar, even as a design document, I think tend to have some kind of a closed loop feedback system where you look at all the components of the really hard thing, and it’s not just hard for the sake of being hard. You can see like, if these hard things work together and we make sure we have all the pieces, we’re going to get a feedback loop. There’s either going to be more user-generated content or we’re going to see an economy that’s self-sustaining or the platforms are going to get safer, like and can see how that’s going to feed upon itself.

There’s other things I sometimes see that are very hard, but they’re almost like they go off and they don’t close the loop. It’s just like, oh, like this is amazing and it’s hard and it’s really cool. But does it end up being part of a system where something good is going to happen? And I think thinking in advance of the closed loop where this thing is going to grow upon itself, that’s one of the signals we try to look for in these things.

KEVIN SCOTT: Yeah. Well, so, you know, as we’re talking about innovation now, I’d love to get your take on the metaverse. And like I have this definition of the metaverse, which is a metaverse is a fully immersive environment that lets you connect with other people to express the fullness of your identity and to accomplish your creative endeavors.

And so, if I say that definition and I look at Roblox, like Roblox feels to me like a metaverse. Like I – even though it’s a 2D screen, you don’t have to put anything on your head. You know, like I – there aren’t many things more immersive than Roblox for my 11 year old, for instance.

DAVID BASZUCKI: I think that’s very close with what socially the definition’s going to emerge to. I like to think of it as an inexorable category following along mail, telegraph, telephone, video call, simulated 3D immersive communication. Definitely about identity, definitely about friends and connection and a social graph.

Definitely about immersiveness and that that immersiveness isn’t just pure 3D fidelity. It’s functional fidelity. It’s social fidelity. It’s device by device, going from phone all the way to immersive VR. I think it’s typically, I think it will more and more be about an infinite array of places and content and objects as part of that. I think these are evolving to always get into economic aspects.

And then I think metaverses will have various levels of safety and civility. Just like places on the web, we’re leaning in really hard on a civilized place for people together.

And I think it’s still so fun and so early, that every company is still trying to figure out their own view of this. Like, it’s fun because it’s still emerging.

KEVIN SCOTT: Yeah. Well, so thinking about this stuff as an engineer for an instance, for an instant, there are a bunch of components that I think any metaverse is going to have to have. So there’s, you know, some way for you to richly express your identity. There’s some way to do commerce with other people in the metaverse. Like there’s some, you know, some value store that you’re going to have to have. There’s some way to, you know, like, here’s what property means in this metaverse. So have you all thought about what those things are? Because again, I think you have a whole bunch of them.

DAVID BASZUCKI: Yeah, this is really interesting. And it’s funny because, you know, if we were building cars 60 years ago, this would be some biz school make versus buy discussion and what do you have to make and what can you buy and all of that stuff.

Some of these components aren’t yet invented and that makes it really exciting. So the components that might allow us to go together to a 50,000 person, photorealistic concert, with great audio and hang out and dance and wave across the stadium at everyone else and have them wave back, that’s a long ways off.

So I like the notion that it’s still early and there’s a lot of deep tech that needs to be invented to support this, as well as a lot of proven tech, whether it’s the economy, is it running on blockchain or a database identity. What’s the graphics drivers and all the machines? How do we do a social graph, all of that? It’s an interesting mix of some technologies that are mature and some that are a long ways off and being actively invented.

KEVIN SCOTT: So one of the interesting and like things that’s way more complicated than I think most people would recognize is like you having your own financial infrastructure. So you all have – have this currency called Robux. That’s a thing that players have. You can spend them across all of the games that are happening. Like they, you know, they let you purchase entitlements, you know, inside all of these games.

You know, my best friend used to run product and engineering at Linden Lab and so like they had their own currency.

So and it is complicated, right? Like having your own economy is an interesting thing. Especially you guys are, what, 50 million daily active users. Like that is bigger than some countries, right?

DAVID BASZUCKI: That’s right. Well, it’s complicated in many dimensions. It’s complicated in a reliability, anti-theft, anti-hack, SarBox compliant, SEC way, in that it just has to be run at a certain level of rigor and reliability and fraud detection and all of that.

It’s contemplated from a infra scaling standpoint, which, you know, many companies do really well, but that still we shouldn’t take that for granted.

It’s also, I think, complicated looking to the future where more and more, if we can see things happening in real life, we’re going to see them in digital life. And we’re still very early on this as far as advertising, as far as shopping, as far as collectibles, as far as a lot of other economic things that we’re used to, that have digital equivalents. So there’s a lot of complexity going forward in designing elegant systems that work well in the digital domain that we’re very used to in our real world life.

KEVIN SCOTT: Yeah. And you know, one of the interesting things about these metaverses is I – and you – you don’t have to offer an opinion here if you don’t want to, is like I don’t really understand or have an opinion about how things are going to net out, whether you’re going to have closed economies inside of each metaverse, or whether you’re going to have some kind of horizontal, like economic mechanism where you can sort of like have a thing that’s in Roblox and there’s a way to exchange for those things in other metaverses.

DAVID BASZUCKI: Yeah, I think this dovetails into all kinds of open versus closed discussions, and the analogy I make sometimes is today, we’re not talking AOL or CompuServe. Today, we’re talking a dial-up bulletin board with a 300 baud modem. That’s how early we are in the tech for these things.

And there’s going to be multiple stages, including the AOL, CompuServe stage, Prodigy stage, till finally we do get open standards, and there will ultimately be metaverse LAMP stacks and there’ll be metaverse, you know, open standards and all of that.

I think what the tension is, and this was tried with VRML 15 or 20 years ago is, you know, HTML with some JavaScript has gotten us very far for 2D and a lot of people have standardized on this and it’s been wonderful.

The raw fabric of 3D, multiplayer, photorealistic simulation, including a cloud across many devices around the world, it’s just such a big technical problem to solve, that I think it’s going to be how do we solve this on the way to open standards, and solving this may be 10 or 20 years. So, you know, it’s just very early and I can’t predict when the full openness of these standards starts to arise.

KEVIN SCOTT: Yeah, and I think you’re totally right. I mean, the most interesting standards that we use right now to run the internet co-evolved with product. It wasn’t that they like came out in whole cloth and then we built everything on top of them. They sort of come along with each other. And I do think you’re right, we do feel pretty early to even have the idea that we’re going to go standardize all of this stuff.

But I tell you, I have been interested over the past couple of years with, you know, what people are doing with NFTs and like what, you know, property can look like in the digital domain, like how you go from a thing that had zero marginal reproduction cost and like you can still figure out how to make it scarce and rare and, you know, collectible and, you know, like a thing that, you know, a set of social contracts give value, you know, not that it’s like physically rare.

DAVID BASZUCKI: Yeah, I think this is a huge opportunity and we’re seeing the natural evolution that in the digital domain, just as we have things of value in our physical world, it makes no sense that some of these things have so much value. Rare items, collectible items, nostalgic items, tradable items, they do have value besides the value of them as food or shelter. I mean, there is extra value there.

KEVIN SCOTT: Yep.

DAVID BASZUCKI: And so, I think when people see this in the digital domain, sometimes they go, “Oh my gosh, that makes sense” and they lose track of the fact that in the real world, we already have had this forever.

So for us, it’s a really interesting opportunity. There’s – there are items on Roblox right now worth $20,000. You know, a Dominus Crown is very valuable. And when having that economy be resilient around that, it does logically start to open the opportunity that the trading doesn’t need to be on Roblox. There could be off-platform trading of those same things. We want to have a healthy economy that’s user generated with scarcity. So there is a big opportunity to make the value on Roblox liquid on and off of Roblox, which is something we’re looking at.

KEVIN SCOTT: So looking forward a bit into the future, like what are you most excited about right now?

DAVID BASZUCKI: Wow. We have, you know, some projects we’ve shared and then some that we haven’t that are all about keeping trying to innovate, really. I love some of the stuff we’ve shared. You know, we want everyone’s avatar to be fully UGC. People lose track of the fact that Roblox has gotten pretty far with a lot of the avatar tech still being controlled by us, unlike the 3D experience tech. So I think that’s just going to be very exciting when our creator community has full control of bodies, heads, faces, all the clothing, all of that, really excited.

Our behind the scenes, you know, our core engine team and developer tooling team, we just have started shipping more and more technology where the Roblox cloud backend is accessible not just through Roblox studio, but just through any kind of web API with auth.

And so, an example we would like to get to is, you know, the data in your place or the persistent data, our devs can build their own tools to access it rather than using Roblox studios.

So we just – I was just this reading about this yesterday. All of the key value stores that our developers use in their experiences, they can access programmatically now. They can start building stuff on top of that rather than just through studios. So super fun stuff there.

I think we’re really getting into how do we do voice in a safe and civilized way because it’s ultimately a big part of what we’re going to do. Syncing facial animation, either automatically with the camera, with lip synching, is going to bring a next level of life to these platforms, as long as it’s once again done in a safe and secure way. So that’s, super interested about that.

Interested in ROS scale. Interested in our internal version of Lua, which is Typed. We’ve open source that. And so we’re interested in efforts to have people use that in education, and nudge, nudge, of course, help our developers be very productive in the creation of that, which is something you guys are really good at. So just a lot of fun stuff.

Because we built such a high quality Lua engine, I mean, the feature I want to see is cloud compilation of our Lua. So, you know, behind the scenes you have both the flexibility of an interpreted language, but once that thing starts running in the cloud, it’s running even ten times faster. So that’ll be very powerful. So just a lot of fun stuff.

KEVIN SCOTT: Yeah. I mean, it does sound like you have a pretty special environment where you get to work on interesting problems across the full gamut of computer science, which is just a ton of fun.

DAVID BASZUCKI: Totally. Totally.

KEVIN SCOTT: So we are almost out of time. And the last question that I like to ask everyone, you know, you’re the CEO of a public company. Like you’ve got thousands of employees, you know, tens of millions of daily active users. So you’re a pretty, pretty busy guy.

But I like to ask everyone what they do in their spare time. Like, so when you’re not doing Roblox, like, what do you do that is interesting to you?

DAVID BASZUCKI: Yeah, it’s, it might sound actually really boring because the things I do when I’m not at Roblox involve family, friend connections, social balance, personal health, fitness, decompression. And I’d say, whereas I try to keep Roblox as the hobby I would want to do, so, you know, I try to – it’s almost like I try to keep the rest of my life organized enough and happy enough and keep my head on straight enough so that I can then make Roblox kind of my hobby and my work at the same time.

KEVIN SCOTT: That is a really great way of looking at life.

Well, so thank you so much for taking the time to chat with us today. This has been a really awesome conversation, David.

DAVID BASZUCKI: Thank you. Kevin. Thank you for having me.

[ANIMATED TRANSITION]

CHRISTINA WARREN: All right, that was Kevin’s conversation with David Baszucki. And again, I mean, that was such a great conversation and so interesting to think about all the things that go into basically building, you know, what goes into building the metaverse, which is basically just kind of like building, not just like kind of an online community, but like whole world, you know?

KEVIN SCOTT: Yeah, it is sort of nuts. I mean, 50 million daily active users is bigger than a whole lot of countries.

CHRISTINA WARREN: Yes.

KEVIN SCOTT: And inside of Roblox, like you’ve got some of the – or many of the same things that you would have if you had a country. There’s an economy. There’s identity, you know, system. Like there’s just all of this creation and activity and interaction between the, you know, sort of the citizens of Roblox. And it’s just it’s a crazy thing to think that, you know, you’ve built all of this from scratch since 2006.

CHRISTINA WARREN: Yeah, no. I mean, that was the thing that I couldn’t stop thinking about as you two were talking was just the enormity of the challenge of scaling and managing something that is this large and is only getting bigger because as you say, there are all these other parts of it, right.

I mean, when David was talking about just the challenges in having to have, you know, an abuse team and having to have, you know, people in place to make sure that things are running the way that they need to be running, like these are challenges that a lot of, you know, global communities face, but I think become even larger when you’re talking about having a virtual world where not only do you have users interacting, but you have the users building in that world as well.

KEVIN SCOTT: Yeah, for sure. And, you know, it’s an interesting, just like you do in any sort of global community, you know, setting all of those community standards, like the, you know, the codes of conduct, the standards of civility and decency that you have. And then sort of trusting by and large that, you know, the vast majority of people on your platform are going to, you know, conform to that, but still having enough, you know, mechanism in place to protect folks from the chaotic folks or the bad actors or whatever you want to call them, who – who want to cause mischief – you know, regardless.

CHRISTINA WARREN: Right. Yeah. I mean, and I have to think, too, and this was something that he touched on a little bit and certainly that you’ve seen with your kids who are huge Roblox fans, but what’s also interesting about this is I have to imagine the balance of having your own features and kind of your own direction, but also really leaning in and allowing the users and the community to really define what the future of the platform and of the metaverse, I guess, if we want to use that term for it looks like.

KEVIN SCOTT: Yeah. I mean we talked about it at some point in the conversation. It’s like I think particularly what we’re talking about is having the incentives aligned between, you know, the platform creator and the people who are creating on top of the platform.

But like you also have to trust one another.

CHRISTINA WARREN: Yes.

KEVIN SCOTT: Like that’s sort of at the core of it. You know, they have to trust that people are going to come to their platform and do things that are sort of net constructive for the community and the platform that they’re trying to build. And like the people who are coming onto the platform have to trust that it’s going to be, you know, safe and interesting and, you know, sort of serve all of the, like, multivariate interests that everybody has there.

CHRISTINA WARREN: Yeah, no, exactly. And I think trust is a real core word that you said.

What also kind of got me thinking, and this goes along with that, too, is when you were discussing with him whether or not, what would it look like to maybe have, I guess, open standards between the various metaverse that exist?

And – and I love the analogy that David used of, you know, that we’re currently at the 300 baud BBS stage. We’re not even at the, you know, online network stage. But I think that’s an interesting thing to think about and it goes alongside, you know, the idea of a trust with users, which is, okay, we have this role here, but we recognize this isn’t going to be the only world. How do we make these worlds interoperable or do we?

KEVIN SCOTT: Yeah. Well, look, I think eventually we’ll have to make them interoperable, right?

CHRISTINA WARREN: Yeah.

KEVIN SCOTT: Just so the same way that we learned as a world of sovereign nations, like in the physical world, we all benefit from global trade.

CHRISTINA WARREN: Right.

KEVIN SCOTT: You know, it’s just we’re a richer world when the, you know, the 8 billion-ish people on the planet can do things for one another. I think that’s also going to be true in this confederation of metaverses that we’re about to have.

CHRISTINA WARREN: Yeah, United Nations of metaverse or something like that.

KEVIN SCOTT: Yeah. Correct.

Or the, you know, global trade organization of –

CHRISTINA WARREN: Exactly.

KEVIN SCOTT: But, you know, it is just going to be the case that like all of these metaverses that will emerge and the people who occupy them are going to have a richer existence and be able to do more interesting things if there is some way to have, you know, trade and other sorts of interchanges between them.

CHRISTINA WARREN: Most definitely.

Well, I’m really glad that we have, you know, people like David who are thinking about this and steering some of the largest communities around this space.

KEVIN SCOTT: For sure. I think, you know, no community is perfect, right? And – but – but I think they have done a very good job of being very thoughtful about how – how they’ve cultivated this community and – and how they’ve dealt with some of its, you know, nastier problems.

CHRISTINA WARREN: I totally agree. I totally agree.

All right. Well, that is it for today’s show. Once again, thank you very much to Roblox’s David Baszucki for joining us. Fantastic.

And if you have anything that you would like to share with us, please send us an email at [email protected].

And don’t forget to follow us now on YouTube, you know, for all of your great conversations in the future. See you next time.