Building the Open Metaverse

Boz and the Meta-verse

Andrew "Boz" Bosworth, CTO of Meta and Head of Reality Labs, joins Patrick Cozzi (Cesium) and Marc Petit (Epic Games) to talk about Meta’s vision for the metaverse, VR, platforms, and much more.


Andrew Bosworth
CTO and Head of Reality Labs, Meta
Andrew Bosworth
CTO and Head of Reality Labs, Meta





Announcer: Today on Building the Open Metaverse.

Andrew Bosworth: You have to build tools that allow these brilliant, creative people to thrive on the platform. Because that's a benefit to all.

Announcer: Welcome to Building the Open Metaverse, where technology experts discuss how the community is building the open metaverse together, hosted by Patrick Cozzi from Cesium and Marc Petit from Epic Games.

Marc Petit: Hello everybody and welcome to our show, Building the Open Metaverse, the podcast where technologists share their insights on how the community is building the open Metaverse together.

Hello, my name is Marc Petit from Epic Games, and my cohost is Patrick Cozzi from Cesium. Hey Patrick, how are you today and where are you today?

Patrick Cozzi: Hey Marc, I'm doing great for two reasons. One is I'm in one of my favorite cities. I'm in Tokyo, Japan, on business, but I also got to buy a bunch of video games last night.

The other reason I'm doing great is we have a very special guest with us today, especially coming off the heels of the Connect Conference.

Marc Petit: Yeah. And absolutely we're super excited to welcome to our show Andrew “Boz” Bosworth. Also known as Boz, right Boz?

Andrew Bosworth: That's right. Yeah. I go, I do answer to either name, but inside of Meta at least you'd get a lot of strange looks if you called me Andrew.

Marc Petit: Yeah and you're the CTO of Meta and the Head of Reality Labs. So welcome Boz to the show.

Andrew Bosworth: Thanks for having me. I'm glad to be here. I appreciate what you guys are doing, getting the word out on the open Metaverse and happy to do my part to try to bring what vision we have for it to your audience.

Patrick Cozzi: Thanks, Boz. We're going to have such a great time today and we'd like to start off the podcast by asking our guests about their journey to the Metaverse. In your case, you've been at Meta for 16 years, I believe.

Andrew Bosworth: Yeah.

Patrick Cozzi: You helped build the mobile ad business, News Feed from the beginning and leading Reality Labs. You've had an amazing journey thus far. Tell us about it.

Andrew Bosworth: Yeah, it's one of those journeys that doesn't feel as unusual when you're going through it as it does when you describe it later on. I think for us from the earliest period when I joined the company in, yeah, January, 2006 we were pretty focused on what do people want to do and what the people wanted to do, especially people like us at the time–young, recent college graduates–was connect, keep up with friends. And this is before digital photos were popular, right? Like digital photos were still like a relative rarity in this era.

And so it was, a lot of it was text and a lot of it was the pre mobile era of the web. And as different as it sounds to be working on News Feed versus then I went on to Messenger and Groups. And then even going into, yeah, running the ad business for a few years. Like there the threads are much more connected than anyone thinks.

Once you're like in that loop of things, you're like, oh, what are we, what are the, what's the service that we're trying to provide to people and how can we provide it? It all like, really makes sense. I think, I was probably a later convert into VR and AR, relative to Zuck. Mark obviously believed from a long time, ran the Oculus acquisition, and had shared from that moment that he always thought this was going to be a big social opportunity for people to spend time together.

And even to the point when he asked me to start working on it and move over to what was then called AR/VR, now Reality Labs, I was just a skeptic and I wrote him this: I was like, hey man, here's what I would have to believe in this thing. And of course, as often happens in the course of writing the document, I came to believe it.

I was like, oh wait, this actually makes a ton of sense. Like so much of our society is limited by physical proximity. We've all had, we know about in terms of relationships and what it is: my college friends who moved in my area, my college friends who didn't move to my area, and what are our relationships like?

Distance relationships, romantic relationships that struggle at distance, these types of things. Parents and family who are close versus those who are not close and how those play out. But also, and then professionally, the more I thought about it, the more I was like, our entire professional system depends on this. Gosh, Meta pays, I assume, an ungodly sum for real estate in the Bay area to have a bunch of buildings that are empty at night and all of our employees pay a lot of money for real estate to have houses that are empty during the day. It just seems like a bad use of energy. It seems like people are leaving their support structures to come here, right?

They're not near their family, they're not near their communities, and then they have kids. They need to get child support and the entire economy hinges on–and I've said to people a lot–that's not malfeasance, that's not economic misappropriation. That is the measured value of face-to-face communication.

That is how much better it is to talk to somebody face-to-face. And if you could uncouple that, if you could create that value and education in work, you have the potential to unlock humanity, because it's not just for the individual.

Sure, an individual gets more education, more opportunity economically. What about their community that gets them to stay there? What about us? And as a society benefiting from brilliant people who otherwise just by dumb luck of geography, of where they were born, didn't have access to these things. So I really wrote this whole document on what you'd have to believe. And of course by the end of it I was like, all right, I'll take the job.

Patrick Cozzi: Very cool. I love that story. So Boz, let's jump in and talk about, Meta’s been very vocal about moving into the metaverse and investing a lot of capital. Tell us more about the vision and where you're investing time and capital.

Andrew Bosworth: Yeah, I mean it's, if you think back to the predictions that Mark made when we made the Oculus acquisition, that really started our journey here, they've really come. We said that VR would be social and we talked about this year at Connect how the biggest place that time is being spent in VR is social. That doesn't mean that people are just like staring at each other’s faces, they're doing things, they're doing games, they're having fun, there's an event, there's a comedy program. I thought Kashmir Hill wrote a great early summary of what the early communities look like in some of the Virtual spaces. And it's not about virtual reality–although that's the avenue that many people are- it's very closely associated with that.

Maybe that's a, maybe it's VR native. I think we've seen it with Fortnite. A battle royale style game where a lot of times teenagers are in there and no one's shooting cause everyone's just hanging out. So we've seen this right? In little glimpses. And so I think it's not about virtual reality. It is about the digital third place. It is about having, creating a place. And that can be for things as trivial as just socializing and things as profound as work. And I want to just call out, I always get so frustrated when I see people bad mouth those trivial use cases.

Hey, have you ever been to a bar? Have you ever been bowling? Have you guys been bowling? Marc, Patrick, you've been bowling. Of course. A weird thing, if we saw another, an alien species, bowling we would think they were out of their mind. Playing mini golf. If we saw an ant doing that, we would lose our minds.

But we, why do we do it? Just because it's fun to socialize and it's an excuse to do it. It's like the excuse to. And so, I do think that the social time in the metaverse is already happening. And so for us a lot of our vision is, listen, nothing is ever gonna be as good as being there in person. Nobody's saying that. Not one person is saying that. But I don't know a single person who doesn't sometimes miss somebody because they're not physically around.

Patrick, you're in Japan right now, so I assume that, you know what I'm talking about?

Patrick Cozzi: Exactly.

Andrew Bosworth: It's just, maybe it's just a generational thing.

I can't stand a phone call. A phone call, it's good to exchange information. I don't feel close to the person after we had a phone call. Video calling is better. One on one. Pretty good. Pretty good. You can do a lot. Read the face.

I love when I'm on a traveling video call. The kids–my wife puts them on the screen. They got a Portal so I can see them all. It's pretty great. But it's not the same as being there. You don't create shared memories when you do that. And shared memories I just think are profoundly important to how we think. Like when I think about my friends, I think about that–all the experiences we shared together.

And you can do that in VR and it's unique in that way. Society loves to take things that start out as games and trivialize them, and I think they do so at their peril. If you look at the history of technology, that's not really how it is. So for us, and then if you take that again, just extrapolate it out, you say, We've already had this experience, we've all spent the last 18, 20 months–those of us who are information workers during the pandemic working on Zoom–and there's some good things. And there's good things about hybrid work and people being able to work where they want to, and there's some tough things about it. And for some meetings, we get into the work rooms and it just is different.

You remember it better, the meeting feels more profound. It's just hard to describe and I think one of the challenges that the metaverse has–certainly a challenge that virtual reality and augmented reality have–but the metaverse has is it is a thing that when you've experienced it, even in part, you get it. You're like, oh yeah, I can see, like I can get it. But it does, it is hard to describe in the same way that, imagine if I tried to sell you on bowling and you'd never heard of it before. You'd be like, “What are you doing? What are you talking about?”

Patrick Cozzi: I think it's a great point, especially around the human connection that can be created here. So Boz, a big theme in every episode of the podcast is how openness plays a role in the Metaverse, whether it be for users or developers. I would love to hear about how openness plays into the vision at Meta.

Andrew Bosworth: Yeah, for us we really don't see this as being a thing that any one company could do even if they wanted to, and we do not want to. And this is going to play out in a number of different ways.

First of all, you already have a wonderful technological stack building up that has a lot of interoperability. You think game engines are a good example where there's, hey, that's not the most important thing that we have to agree on. There's lots of different ways. And likewise interfaces, I think we've seen great examples of worlds that are navigable on phones through touch controllers or through richer controllers on a console or through a keyboard mouse on a pc.

There are pieces where we have pretty good and healthy abstraction as an industry that we get the benefit of building on top of, everything all the way down to GPUs and streaming and the internet. There's just a wonderful set of standards that we already are all committed to and building on top of.

That's not saying that they're not competitive. Listen, I don't need to tell Marc how competitive the game engine business can be. They are competitive, but they're competitive in a way that really benefits consumers and I think benefits the whole industry and we're healthy of those.

Then there's some that just we don't know yet that are super important. So I think the three that we’ve come back to again and again as these cornerstones: Identity. How do I express myself and who I am, consistently? One of the things I've talked about for my team is the watch word for the metaverse is just “cohesion.” Does it feel like–you can't feel like you're getting ripped apart at the atomic level and then reassembled every time you go through from one room to the next room, that would just feel bad.

Does it feel cohesive? Does it feel like it hangs together? So identity- closer to identity -and, obviously avatars are part of that. Communication is part of that. Digital objects are part of that, just possession, ownership. Like you own things, that you have things, and you can reliably get them.

You're like, oh, like I have a chessboard. I can reliably get it and I can reliably share it with you. It's gonna behave reliably in lots of different worlds. Or there's enough cues around me that I know it will behave differently. Oh, this is a zero gravity world. Chess board's not going to work here. That's on me, that's not on the chess board. It didn't break, it didn't run afoul of some kind of provisioning system.

And then the last one for us is “travel.” Just thinking about travel and how you move between spaces. Again, to that point about cohesion, we want to be able to go from a place that is built by an app developer on one platform to a place that's built by a different app developer on a different platform.

But my friends all come with me and it feels like it's not. I've had to go all the way to the system level and then try again. Which is right now, and we always kinda joke with this, like the metaverse is here, just not evenly distributed. There's certainly games that have metaverse qualities of being in them, but what they lack is the, is any kind of cohesion to any–they're universes. They're just universes or they're maybe just worlds and they lack any ability to then move to a different place. And so those are the pieces that we're looking at a lot.

I think you guys probably know this, we're working with the Metaverse Standards Forum which I think is just such an important piece. And look, you don't, the Metaverse Standards Forum, it is a tricky problem. You don't want to get ahead of–we can all specify till our fingers fall off what it's going to look like, and then you have no adoption and then somebody comes over here and builds a popular thing and then all the adoption's there and now you've actually just managed nothing.

So there's a little of a two-step here. We all need to keep making progress and see what's popular with consumers and then commit to each other, which is what we have done as being part of that forum that we're going to fortify around that.

That being said, we're seeing early promise, right? glTF. Easy, an easier one. They're going to get harder. The more you want to script it, the more you want that line, those objects to be rich. Not just visual descriptions, not just textural descriptions, but really active descriptions. That's going to take more time.

And so you're going to write that, just that right balance to avoid specifying something that nobody implements or that consumers don't want. But you also don't want to wait so long that you just have a bunch of independent competing things.

Marc Petit: Yeah, no, absolutely. We've been, that's the balance we've been trying to achieve in the Metaverse Standards Forum.

And try to understand the benefits of glTF, what Pixar and NVIDIA have shown, and Apple too, what the part of USD that's actually working very well and could be the foundation of a standard. A follow up question on, on that, the Metaverse Standards Forum works on working groups. Is there one working group or topic that you think should arise quicker?

Like avatars or, digital clothing, or AI, or identity? What is your sense of priority you have in your mind?

Andrew Bosworth: Avatars feels like the nearest, like the one that's coming fastest. And I think part of the reason for that is that avatars just have a lot more outlets. Avatars have outlets in games today.

They have outlets in 2D surfaces through stickers and emojis. They have outlets potentially in real-time calling, like by virtue of just having more outlets, those feel like one that could be done sooner. And they're also, we're also drafting on a much longer technological history as it relates to avatars.

People who have been creating representations, whether they be highly stylized or not, for a long time. And so that one feels near term. Digital goods, including things like clothing is a pretty wide and varied space. I'm a little bit less worried than that just because I do think. It's very likely to have to lend itself towards descriptive languages.

The scripting part of that feels like the trickiest bit to figure out what you know, how much we're going to burden the object, how much you're going to burden the system, how much of implementation sits on the other side. So the scripting piece feels hard, but some things I think are pretty straightforward, creating ledgers of ownership.

You don't need to use the blockchain for that, by the way. You are welcome to, it’s one of the things that you can use. You can also use a database if you have a trusted company. There's, I think those things we could probably also do sooner. So I think it really depends on where the technologies and development travels farther off, because there honestly just aren't enough places to travel between yet.

We don't have a canonical sense of a party or who's with you or what identity they're expressing or what kind of agreements we need to make with the consumer so they aren't surprised when they get to a new place. And so I think that one probably feels like the one that's the farthest off and you just need to let these worlds develop a little bit more.

Marc Petit: So you mentioned digital goods and Mark Zuckerberg yesterday in the keynote talked about interoperable digital goods. How do we read  interoperable in this context?

Andrew Bosworth: Yeah, I think there's always going to be, like I said earlier, we have plenty of universe examples and world examples where when I go to certain games I can either, through grinding.

Or through an achievement or through buying a cd or buying something, I can acquire goods that really only make sense in that game. They only exist in that game and that's where they live. That's totally fine. That's always going to be a thing that happens. It's not going to be the case that everything is always transportable everywhere.

That's a little bit silly. But in so far as there are things that you want to say, like hey, the achievement itself, the trophy I want to take that with me. I'm able to, anytime I have, I get this thing. It shows me like, outfits I think are a good example. It's yeah, I just want to be able to have this outfit.

And whether I earned it or bought it, who cares? Like I, I want to be able to have it. Those pieces we, for consumers, it's a strict good for them to able to use it in more places, right? But it's strictly positive for consumers. In my opinion, if you can create a metaverse that is contiguous, like I said before, it has that cohesion.

It's really positive for the metaverse because the more somebody invests in any part of the metaverse. The more they feel invested in the entirety of that continuous space, right? And I want to pause there, there are analogs here. There exist clothing items that I can't wear in certain places, right?

I've traveled the world, and there's churches where you have to put on a more clothing to enter the church. That exists. There are items that I can buy in other states that I can't fly to California, right? Like that, those things exist. So it's not like the real, the physical world, has zero geographically limited things, but it's relatively few. And for the most part you feel like, yeah, I've got a good sense of what and why. And the continuity is, as soon as it's broken, you don't have a continuous world. You have a very discontinuous world. And it becomes really, the overall construct suffers from it.

For us, we want to be consumer friendly and we want to make sure that consumers, when they're investing in any part of the metaverse, the entire metaverse is stronger for it. And you don't see that happening if you don't have, try to build these modes where it's, “you didn't buy it in our stores, therefore, we're not going to let you use it.”

That's a bad experience. Now, if there's reasons like, hey, we literally just don't, we don't implement that primitive, like we don't have the ability okay? I'm not saying there's going be zero goods that have that problem, I do think there's quite a few.

Like digital clothing, which is, I can't imagine the scenario where it's “yeah, you can't have that shirt.” I don’t know, it seems fine.

Marc Petit: Yeah. congrats on the announcement of the avatar SDK.

Andrew Bosworth: We’re so excited.

Marc Petit: I noticed that you called it a “store. Is that going to be a vector of monetization? You think it could be a big business?

Andrew Bosworth: Yeah, I think, if we look at historically speaking, people invest in clothing, people invest in barbering, people invest in makeup, people invest in these things. Because we care about the image that we put out in the world and what it says about us and other people value that, other people around us value having a sense of being able to look at somebody and assess something about them.

And you can take it too far, and I'm sure that will happen, but I think for the most part, it's a place that you want to enable people to invest. And here's the problem - if you, as a single company, no matter how big the company is, oh, “I'm gonna go build all the fashions and all the styles that everyone wants.”

Yeah, good luck with that. It's never happened once in human history, I'm sure you're going to be the first. It's not going to happen. So if you don't create an economy, then you never get the creators who really I think, will have an opportunity? Give consumers the options that they want, right?

That's what's happened in the physical world. I don't see it happening any differently in the digital world. And so for us, I think the economy is such an important part of the metaverse, and it's so often framed and I'm sure you guys will appreciate this more than maybe the average person. It's so often framed in terms of business model or it's like it's not even about that.

Otherwise you wouldn't do it this early. It's about actually providing consumers 2ith the types of choices they want, which you just have to be skeptical that any one company is going to go build all on their own. Fashion has never worked that way in any human society that I know of. I don't see why this one would be any different.

So I do think that, for us, is building out economies is super important. Empowering creators is super important. That's true with the avatars. That's true with worlds. And thinking about the people who invest this time and energy to create a world, what systems and tools are they going to have access to, to monetize?

Can they charge for tickets? Can they charge subscriptions? Is it, can I just accept donations? You have to build tools that allow these brilliant, creative people to thrive on the platform. Because that's a benefit to all.

Marc Petit: Actually, you announced yesterday the ability to import 3D data from Maya and Blender into Horizon Worlds and connectivity with Sketch Fab as well. And also programming with TypeScript, which I think was very interesting. So what kind of experience do you expect people would build with that?

Andrew Bosworth: Yeah, Jason Rubin, who is our head of all content in our organization and has been since before I even started working in this space.

And of course he goes way back to Naughty Dog and is a tremendous figure in the industry. He's always given me this framework, which is if you, it's a pyramid, at the very top of the pyramid there are triple A devs who are capable of doing things that relatively few other people are capable of.

It's a team of people who've come together. They can build fantastic pieces of highly customized integrated software. Maybe they're using a game engine, maybe they're going straight to the metal when they need to. These are rare, There's not a huge number of them. There's a countable number of studios that can, that are capable of this.

And they're always going to build directly on your, on the base metal that they, cause they want to get every single last bit of performance and customization out of it. And of course we see that today in, in, for us in Quest, in the store. A lot of these titles are great titles built by professional teams.

And they're going to do that. I want to give them more tools for digital goods, for avatars, for social and communication connection. So we want to do more with our SDK, but they're going to do all the last mile work. Because they want it to be fully integrated into the universe that they're building or the world.

Underneath that you have highly capable creators who are indies or individuals who are motivated to have a vision of an idea, and they're going to build on top of a much more complete stack. They're going to use the tools available to them in Unity or Unreal. They're going to use the tools available to them to create, but they're still highly talented.

These are trained professionals that can make a living building content, which is a super competitive field. And so for these folks, some of them will build on the metal just like the triple A devs, but some of them actually are going to say, “Okay, gimme the higher order tools. No problem.” I want to work in these other tools for both these top two tiers of developers.

If you're not supporting Maya and Blender, you're taking the knees right out from underneath them. These are powerful tools that they use as part of their workflows. And you're saying, “No, you can't use that workflow, use like this other or non-existent workflow, or use a worse workflow.”

So we want to support them totally. If they're building direct, and hopefully we can coax them into our SDK by providing a lot of value or they want to build inside of Horizon Worlds. Great. Let's let them build, let's let them use all the tool chains that they're used to to do that. And then the, but the pyramid keeps going down, right?

There can, then there's this set of people who actually could, with basic scripting tools and basic world creation tools, create pretty fun, compelling experiences. And they're never going to go into Maya, they're never going to go and do modeling that way. But you can give them an object, you can shape it and they can use some of the tools that are readily available.

So that's a, it's an important layer because I think for a lot of places in Horizon Worlds. It's “Hey, I don't, I'm not trying to build a game necessarily. I just want to build a clubhouse for my friends. I want to build a place that I can experiment and do things.” And then it goes all the way down to the base level creator.

Like everyone's a creator in the modern world, like who's, who just shows up, what can they do? And we're looking at tools there as well. Okay. They're not going to be doing scripting, but they can literally put blocks together. Those blocks could do a thing or they can put blocks together. Those blocks can create an object.

And so we just think of this whole pyramid and we want to make sure we're building a tool chain that works for every single layer of it.

Marc Petit: So you also purchased a few years, months back, a game/platform called Crayta, which is like Roblox or Fortnight Creative, a place where people can create games.

How important is that platform in your strategy?

Andrew Bosworth: Yeah, totally. Crayta is another one of our creative Metaverse products, similar to Horizon Worlds. People can build their own virtual worlds, experiences, games, they can share them with other people. Now obviously they have, these platforms have different audiences today.

They're available on different device types. So Horizon Worlds today is in VR, although, obviously we've announced plans to expand that to the web. In the case of Crayta, now you're looking at web and mobile platforms. Having said that, they are very much building towards a common vision, so internally, like I said before, “identity, avatars, travel.”

We're trying to figure out how we can make sure that these are stitched well together. So on the avatars front people in Crayta will soon be able to link a Meta account to their Crayta account so they can bring their Meta avatar to their favorite games. And we've done this, obviously we talked about the Meta Avatar SDK, so that's going to be something that we're hopefully seeing adoption of more broadly. On the creation side, both sides allow people to create these kind of cool experiences and games. Right now they're only available inside the platform they've been built into. But over time we do want to make it easier for people to kind of exchange creations between these different metaverse experiences. And so I, it's early still, but yeah, these are very much working towards the common vision and being operated by a common force.

It will be a good test around early things around travel and interoperability.

Marc Petit: Absolutely.

Patrick Cozzi: Great. And Boz, you mentioned the web in the Connect keynote, you mentioned Horizon on the web. What can you share about this?

Andrew Bosworth: Yeah, so I think again, to my earlier point about I don't want people to over index on VR and Metaverse, they're happening together and they're related in so far as the VR experience may be the most native way to capture that feeling of presence that we expect to characterize the metaverse, but they're not the same thing. And by far the majority of people’s, I think first experience of the metaverse will be on a mobile device or maybe on a PC or a laptop. And the web is the window to those devices. It's the best way to do it. Especially given, obviously, the limitations that have come up in some of these platforms over time. So the web is a really promising way to just make this available to everybody so they can get a glimpse of it, they can have an experience of it so they can understand a little bit more what it is.

I do think we all struggle a little bit because the metaverse is still too abstract a concept for a lot of people. Yeah. And so you got, you guys have a podcast, so you're helping, you're doing your part. I appreciate that. But I think people being able to say, “Oh, okay.” Because I know for us a little bit. Man, I remember when I first saw Fortnite running on a mobile device, I was like, “Oh man, like this is different.”

This is something completely different that I hadn't seen before. There were other things that I think people had that experience even before I did. But for me, I was like, “Wow, this was the social nature of it,” I was like, “This is a cool thing.” And for a lot of people it's hard to envision.

They think 3D has to happen in an immersive headset. That's not true. We've been playing 3D games since Wolfenstein II.

Patrick Cozzi: Yup!

Andrew Bosworth: We've been playing 3D games and our brains are really good at remapping from 2D screens into these immersive spaces.

And so I think that's one of the really important pieces. And of course it also allows us to start building a good connectivity of. You had this great comedy show happening in Horizon Worlds. There's no reason you have to be in Horizon Worlds, in a VR headset to experience that great comedy show.

Great comedy is great comedy. It's totally fine. TikTok is full of it. So it's totally fine to experience it on a screen. So I think for us, that's obviously just introducing more people to what this is and giving them a tangible value proposition around it.

Patrick Cozzi: Yeah, I think that's great.

Decoupling the metaverse from the device in which you access it. And then, yeah, thanks for the kind word on the podcast. We're doing our part, but at the same time, Marc and I think we're learning on each episode as well. And the way Marc and I work together, I say he's the responsible adult, and I'm the geek.

So I did want to ask you a bit of a geeky question. During the Connect keynote, you mentioned the Web XR support for the Quest, and, I wanted to ask about, what you think on the same hardware today, you can build for the web or you can build for native, what do you think about the gap between those today and maybe how that plays out in the future?

Andrew Bosworth: Yeah, I think history's instructive here. If you look at mobile phones is probably the most recent obvious history. Famously we had a project inside of Facebook called Face Web, which was an attempt to build, to use the web platform even on mobile phones. And honestly, it was, a little selfishly motivated.

We didn't want to, we had come from like a web workforce and now we are being asked to spin up an Android team and an iOS team. And it would've been great if you could have just had a little thin shim layer, on iOS and Android, and then a big layer that was the web still, and it was a total disaster.

Kinda one of the famous, I think, technical failures that we've managed inside of the company. I, again, the team working on it was outstanding. They did the best that could be done. But between policies which were prohibitive from iOS in particular and performance you had one set of problems and there was a second set of problems, which is just that the audience expected things to be native, to look and feel native, not native to your application across platforms, but to the device they were using.

And it's a lesson I remember learning when I was briefly at Microsoft after college and they talked about how at one point they had flattened the Mac and Windows office products into one single skew. And it was unpopular with both Mac and Windows users because it didn't look like either one of them.

And as a consequence it was an odd personality they had to undo that change. So it's a lesson that we've learned a couple of times in our industry. And so I think for a lot of experiences honestly Web XR is going to be fine in the same way that for a lot of mobile experiences, mobile's fine.

That doesn't mean it's not optimized, right? It'd be bad if I just had my website that was optimized for a laptop display on a mobile phone. No, I need a mobile optimized website. But a lot of services are totally fine being accessed that way. But for services that are really trying to take advantage of what the native platform is doing, I think they're going to increasingly want to be built native. We want to support both. We would love to have both a rich set of more casually and available services. Totally awesome. And also, things that are more like intense, performant. They need to have this tight cycle time. They struggle with the latency.

So you're going to end up with a little bit of both. There's a ton of stuff where we're working with developers on great Web XR-based experiences and we love it and we're going to continue to try to advance that. It's one of the really important things, I actually think, frankly the browser team that we have is one of our superpowers.

I think people underestimate the importance of the browser. It's a virtual reality. And it's been one of our, always been one of our top performers. If you look at the announcements we had yesterday where people have multiple windows up, they're mostly using browser-based tools and you don't need your spreadsheet to be VR native.

That's actually spreadsheet's just fine as like a 2D panel. You don't need to have that be all kinds of crazy stuff. And so for me, the web, 2D and the Web XR are wonderfully important platforms that probably the long tail, 80% of experiences and content. By volume of experiences and content will be there.

Time spent, it's going to be inverse. It's going to be 80% of the time spent is going to be native. And it's not because we care about native. Native doesn't, Hey, listen, snow flies on us. It's just going to have the property of being performant and meeting the expectations the consumer has in terms of how it behaves.

Marc Petit: So do you think, like for example, you run Crayta on top of Unreal Engine, I think, and you run Horizon Worlds on top of Unity. So that's a lot of complexity. You're mixing graphics engines and operating systems and everything. So you think that's still a requirement? Because the Meta Quest browser seems to be a very good browser when it comes to 3D.

So it's an example of the browser that seems to be very performant on the device that you control. So is there hope that a browser would become a little bit more than what we currently see?

Andrew Bosworth: It's really good and we're going to continue to hope, like I said, we're investing on both paths full out, so we're not like, “Oh, we're favoring one over other,” and I think Web XR is critical.

I think having the browser be as performant as possible is important. I think there's also a bunch of things where like just the latency requirements and especially as we're unlocking, for example, if you've got a device that has cameras pointing at your eyes and your face, how do you make sure that you're managing the information correctly?

What are the platform policies that we're going to have to put in place to make assurances to the headset owner about their data? So there are going to be both, We're going flat out on both of them for sure. I think, this obviously raises a third question, which is around streaming, which we also believe in and think is important.

To your point, Crayta is absolutely a move in that direction. And we all, so there are a ton of use cases where streaming is actually great. By the way, social use cases are one of them. You can really do that. There's some tricky bits. If you wanted to do encryption on multiway conversations and have it be streaming now it's, you can't literally do it end-to-end.

So there's a kind of, there's new research that needs to be done on how to do encryption, in an, on an enclave, on a server maybe, so there's challenges there. And, but setting aside even those kind of, that's an arcane one I tossed to Patrick, my self-admitted geek buddy over there.

Patrick Cozzi:


Andrew Bosworth: You and I, we all know some of these experiences are latency bound. Even, I thought Stadia is a good example here. Stadia was a truly impressive build out of last mile infrastructure, and I think they don't get enough credit for how impressive the buildout was of putting that much compute at the edge, that much bandwidth to get down the pipe and listen, I played Red Dead on Stadia, and I thought it was great. Red Dead is a game that the latency is not really a major factor. If you're playing a fighting game, you're playing a first person shooter on multiplayer, these guys live and die on on milliseconds that you can't. And you won't, we don't have an obvious path through any combination of network technologies closing that gap for the median person who's going to use the headset.

So I think cloud streaming is also super important, and some of that will be Web XR and some that won't be depending, but it also has its own limitations on what kind of content it's going to be right for. And so I don't think for us picking one, we don't have to pick one. Like it's actually, it's great. Like we can do all these things,

Marc Petit: And one of the big partnerships you announced yesterday was Microsoft, and this is going to bring Xcloud gaming to the Quest. You think you see this as a viable alternative for people instead of buying those big monitors to play with a VR headset?

Andrew Bosworth: Yeah I think so. This is my, for a ton of popular games. I think it's great. And listen, I think working with Microsoft, I think Microsoft's vision for Xcloud is a cool one, as a gamer, right? If you're a gamer, you like this idea of, “Hey, maybe, maybe I can get a steadier expenditure of money in exchange for my gaming entertainment needs.”

I think it's a promising idea. We'll see how it plays out. Obviously, the ecosystem is big and there's always going be exclusives that are going to make that value prop a little less, obviously good. And maybe it's not tenable, I don't know. But I want in so far as I have people who are using our product, who are also subscribed to that, do I want them to have access to the games they're already paying for?

You're damn right I do. Of course I do. I want them to have access to that. I think, look, there's obviously over time business model questions that we have to answer. When we're trying to move these headsets and we're not making, as Mark has said recently on a podcast, we’re not oriented to make a profit on the hardware. And so you gotta have options for that. But for us, at the end of the day if it's, if they're already an Xcloud customer, they're going to play those games on another device, we might as well let them play on our device. So might as well give them full value for their money.

And so I'm excited to see that. I think there probably will be some titles that maybe struggle. But probably most of them won't. I think they've done a great job executing there.

Marc Petit: Yeah, no, we’re excited to see lots of people happy playing Fortnite on cloud. So even though latency is a challenge, it looks like users are pretty happy with it.

Andrew Bosworth: Yeah, totally. And I think, has always, we get, I think we get sometimes dinged in our industry, people think that we're out creating stuff for its own sake. We're all just following what we think people want. Like we're all just trying to build products that people love. That's all we try to do.

And so you follow them where they guide you. And we, this is one that if you've been in the early communities, like back in the day I was on, I wasn't on AOL, I was on Prodigy, I was on Prodigy boards. And then, it was early on, QQ early on, all these different things.

And yeah, it's got that vibe to it. It's got the feeling of this is a place that people are going to spend time. Yeah. I do think, as it relates to gaming, we continue to try to be as friendly as we possibly can. Obviously do want respect that we have business model questions and we're investing a ton here.

And at some point we have plans for that, but I think it's good for customers.

Marc Petit: So let's try to take your crystal ball out and get a sense of timeframes. We had Michael Abrash on the podcast, your Chief Scientist, right? Yep. That's his title. He did a great job about setting expectations in terms of core presence and core experiences, and what's your take on the time it will take VR to go mainstream?

And the ancillary question would be, can you really lean on mobile as a way to reach critical mass? If you want to be relevant, Facebook is all about, and Meta is all about critical mass. So can you lean on mobile to get there?

Andrew Bosworth: Yeah, I so whenever I get asked this question, it's a super tough one.

I'm pretty sure VR is already over the initial elbow and the curve. I think Quest 2 did that. Quest 2 took it from a linear or a sublinear curve to now a super linear curve. Now we don't know the shape of that curve over time. We don't know how steep it is. But we've made the first inflection and so I'm feeling really confident in the future of VR as just being sustained as an ecosystem, sustaining medium, that is increasingly common.

The old saying from technology is that we always get less done in one year than we expect and more done in 10. And I have really found that to be true in my experience. I basically never bet against anything in a 10 year timeframe at this point. That's not totally fair. There's a few things that I do, but like 10 years is just a tremendously long time and we are investing a tremendous amount.

Not Meta, well, certainly Meta too, but like the industry, in this direction. And so for me at least, I'd be pretty surprised if 10 years from now, VR wasn't the kind of thing that they had the adoption of, laptop computers or desktop computers as they were rounding their apex in let's say the 2010s or something like that.

And I'd be pretty surprised if the metaverse wasn't a place that people spent time regularly. And again, it might be on a mobile device, checking out a show, going to a room, doing a hangout, something like that. But I think it's, I think it will be something that people will do commonly.

I think that, which means that I think the economy will exist and there'll be people who are full time, that's what they do, they create digital goods and digital experiences for the metaverse. A decade feels safe to me, a decade for sure. One year is way too soon/ And so, of course I was like, fine between two and nine years. What do you think? I don't have the answer to that.

Marc Petit: Yeah, the keynote was all about connectivity and social interaction. What's the role of entertainment in there? Don't you think entertainment could be the accelerator to get people on those platforms?

Andrew Bosworth: Forms of it really is. I don’t know if you guys had the chance to watch the keynote in Connect, but it was really cool, and the stage was dynamic and the avatars came out instead of the video. We actually showed the avatars and everyone rushed to the front of the stage.

And it was funny, cause I checked with people on tens of different instances and every single instance, everyone rushed to the stage as soon as the avatars came out. It was a cool experience. So that was a social construct. And I think that's, no surprise, that's what we're using to tell our story, but a lot of this stuff is just like, is it fun?

And that is, a lot of times, is about what's your individual experience of going in there? And this is an area that we're investing. Think Vishal [Shah] who's our head of VP of Metaverse, wrote this thing that got leaked, I think people made a lot of fuzz out of it. They didn't need to.

I thought his quotes in there were right. It's “Hey, we got, for the individuals going through this, we've still got some bugs. There's a lot of new user experience stuff that we haven't hammered out. And we've gotta make sure that the individual experience is entertaining in its own right.”

Otherwise, it's a chore. You gotta schedule it. It only really works if there's that serendipity, that sense of, “Hey, I'm going in. Oh, you're also going in, great, but I'm going in because I want to go in and that's that.” So I think for us, like the quality of it, putting the polish on it that people have come to expect, making sure that there's great content.

Like again, thinking of the content inside of Horizon Worlds, the same way that we think of the content for the Quest platform as like, hey, you gotta invest and you gotta make sure that there's the great experiences that drive people to have that behavior. Yeah, I do check with that a little bit because, people in our industry, it's yeah, you gotta dogfood more.

Teams are writing code, they're not using enough. It's a balance that you want to strike. Hey, we gotta do a quality lockdown. These are things that we know about that I think are a bit of intrigue for the journalists set. And so I don't worry about that as much, but I do think, yeah, just the entertainment, just the fun of it.

Like the stuff that we are seeing right now, which we have highlighted a little bit at Connect, is like comedy clubs in Horizon Worlds are great. They're really great, and it's like an open mic that's always open. And so it's, there are some things that are starting to come out and emerge as like early, really positive communities that are forming around entertainment concepts.

Marc Petit: Yeah. And I think one last question. When Mark Zuckerberg closed the keynote yesterday he made the case of the metaverse being a more richer experience. But the other side of that coin is that it exposes yourself more. Your interactions in the metaverse are going be more impactful and you want people to have real casual social interaction or even private interactions on the platform.

They will need a lot of trust in the platform. And there was a very interesting article this weekend in The New York Times about a journalist who spent time in Horizon Worlds, a pretty good one. So how does Meta think about moderation and how you scale moderation? How do you handle data privacy to build that level, that amount of trust?

Andrew Bosworth:

Yeah, that article is the one I referenced earlier from Kashmir Hill. I thought, I agree. I thought it was interesting and I appreciated that she really just spent the time and did it and got to know the communities and as a consequence, she really enjoyed it. It was interesting to see that report, which is so promising.

Yeah, the analog for unlike, so much of the digital spaces that we've been experiencing, which are asynchronous. Text or image based. What we're dealing with in the metaverse is largely synchronous. And in fact, I really, when I talk to people, I talk about this being the synchronous social network and that really changes things tremendously.

And frankly, it returns a little bit more to the expectations I think we have for the physical world. And so in the physical world, when I'm in the comfort of my house, there's a degree of privacy that I expect, which seems really important. I can say things in my house that I don't think would be appropriate to say outside of my house and so on and so forth.

Just before I get hit on that, I don't actually say anything differently between my house and outside house, but like it seems important as a matter of like civil jurisprudence so that I would be allowed to do that. I think we'd all be a little bit alarmed if we had read a story that like, hey, somebody said something privately in their home, and then immediately the police busted into their private home and arrested them for thought crime. That seems like a dystopia that we don't want as a physical society. It's also one we should not want as a digital society. And likewise, if I'm out at the park somebody can harass me and there's not a huge, I can call the police and somebody may come or not come.

In the digital world, we can do strictly better than that. In the digital world, unlike the physical world, I can cause that person not to exist for me. Like they can keep doing whatever they want to do in their world, but like they cannot exist. To me, that's pretty powerful. You don't have that in the physical world.

We can certainly still summon authorities and say, “Hey I've recorded this. I'm asking for somebody to intervene. I think what you've done is against the rules.” That's a possibility. We certainly have a better response time because we don't have to physically travel from a police station somewhere.

And we've got guaranteed an ability on your party. If you want to record your experience, you can do that. Which you may not always be able to do in the physical world, We understand it's an immersive medium and we're trying to build those tools

I just, I do think. How to manage these things in the metaverse is an open question. It's a hard one that we spend a lot of time on,

Sure. And so I, I just, I think we've gotta set expectations much closer to the physical world and understand that the good news is that the floor for the digital world is strictly better than the physical world.

Marc Petit: Just last word about scale, it looks like, when Kashmir Hill was in the metaverse, I mean she met moderators, you think that, that could scale to having billions of people on the platform?

Andrew Bosworth: Yeah. So there's two different levels to this, right? The first one I think is the level that we're coming from, where in places like TikTok, Facebook, Snapchat, YouTube. Like literally there's only if you want to escalate something, you have to go all the way basically to the company. I think the metaverse is going to be different.

I think, you're going to be in a world and you're going to choose to be in that world and you're like, there's somebody built that world and is trying to maintain that world for a certain purpose. And if that world is one where the rules of the world don't align with what you want to do, you just don't go to that world.

And if the rules do align, then you've got some kind of recourse inside of the channel. Does that mean, is it the admin? Do they want to hire moderators? It depends on what type of world it is. Then of course there will be a separate relationship that you have with the provider of what? the platform, maybe? the hardware maybe? depending on how much those things line up.

And yeah, there's going to be a grievance to that level too, for things and behavior that is explicitly not allowed, or illegal, or require review. But you do, I think, want to reduce the degree of flattening that seems to have happened in asynchronous spaces where,- and nobody's happy with it, by the way.

Nobody's happy with it. Like people don't want the only recourse being going to a company and the company doesn't want to have to adjudicate every disagreement. It's unavoidable, in my opinion, and it's an important problem that we've invested more in than anybody at Meta. On asynchronous platforms because it's just the nature of them to be, and nature of the internet that we all inherited to be completely flat, transparent, everything can be shut up everywhere.

I think it's very different when you have a synchronous social network and all the, it's not a bunch of artifacts of previous statements, It's all happening right at this moment. It's a real-time thing that we're experiencing. There will still be artifacts, digital goods that have to have a certain degree of content moderation.

Everything's a content moderation problem. But it will be just very different, I think, than the internet as a whole, as we see it today.

Marc Petit: Thank you.

Patrick Cozzi: So Boz, we've covered a very impressive amount of topics. I really admire your passion and depth. To round out the episode, we have one last question for you, which is the shout out, if there's any person or organization you'd like to give a shout out to.

Andrew Bosworth: Oh, man, there’s so many things happening. This is a tricky one.

Patrick Cozzi: You could do more than one if you like.

Andrew Bosworth: Yeah, I think there's a couple. I think, for a shout out. To Guy and the work that he's done with Virtual Desktop. I think that was a use case that has really been plumbed through thanks to just, really one tremendous effort.

I think, shout out to all the companies working in fitness, FitXR, Supernatural these pieces. I love all the creator tools that we announced, and I'm going to get myself in trouble if I don't talk about all of them. Every time you start naming things, you're in real trouble. Let me tell you that right now. Just, what I love is people who are seeing the opportunity, whether it be metaverse, whether it be virtual reality, to take a niche and just and just go deep on it. And they find out that they're loyal, the people who are like them, follow them. They come with them on that journey and they get rewarded for it with these kind of cool markets that grow up around them.

So I think there's a tremendous amount happening that's really positive and that's my advice to anybody in the community is not to pay attention to all the hype. You don't have to understand all the stuff. You don't have to understand. Do you have something that you think would be popular with a community of people that you know?

The cool thing is there's just never been better tools to build it than there are right now. Never in history have there been better tools, more readily available to build them right now. And what we were finding increasingly is that there is an audience and there's a business around those tools.

Patrick Cozzi: Well said.

Marc Petit:  Yeah, agree. So Boz, you are the CTO and the head of Reality Labs at Meta. It was a pleasure talking to you today, I think, on the heels of a very rich and dense Meta Connect. And for you, we could not cover, scratch the surface of all the announcements that you guys did, but that was not our purpose, but we're glad that you were with us today. So thank you for the depth and the breadth of this conversation.

Andrew Bosworth: Yeah. Thanks for having me guys.

Marc Petit: And Patrick, thank you so much. Have fun in Japan. And to all of our listeners, thank you as well. Keep on hitting us on social. Let us know what you think. Let us know what you want to hear, and thank you very much and we'll talk to you next episode.

Patrick Cozzi: Thanks everybody.