This article is part of a VB special issue. Read the full series here: The metaverse – How close are we?
Brendan Greene invented the battle royale genre for video games. His PlayerUnknown’s Battlegrounds (PUBG) has sold more than 75 million copies since 2017. Instead of making more sequels to PUBG at Krafton, he spun out his own studio, PlayerUnknown’s Productions, and is working on the most ambitious game he could possibly make.
He calls it Project Artemis, and the plan is to make a planet-size open world where players can define the rules and contribute their own content. I talked to Greene about it at our GamesBeat Summit: Into the Metaverse 2 event.
Greene’s small team will first work on a proof of concept, dubbed Prologue. It will be a Unity-based single player game world that is quite massive itself, with a play space of 64 kilometers on a side. The object is for the player to get to an airfield and exit the map, overcoming obstacles such as freezing weather. You have to create warm spaces and move on in order to escape the increasing cold.
That map will be regenerated every time players start the game, and Greene hopes the same will be true for Artemis, which will take perhaps five years to make. So far, Greene hopes that Artemis will enable players to set up their own towns and create games within the borders of those towns. But he won’t script a narrative for Artemis. And in spite of his ambition, Greene doesn’t really like the word “metaverse.”
Three top investment pros open up about what it takes to get your video game funded.
Watch On Demand
“I’m not sure if that’s what I’m building,” he said. “I first set out to do a concept what I want to do next. I just knew I wanted to build a big open world. And as we developed the ideas internally, that big world got massive. And right now I just want to find a focus on building this digital place. I don’t even know what the metaverse is meant to be yet. I mean, is it a place? Is it a set of standards? So I’m very hesitant.”
When he first thought about the idea a few years ago, the metaverse wasn’t really on the radar, and it has only lately picked up steam.
“It’s more important to build this digital place first. And if that is something metaversal, that’s great. But for me, it’s more important to focus inward first to build this world and make sure that all the systems and utilities that I wanted to have are in place,” he said.
Distinguishing between the projects, Artemis is the end goal, and it will require Greene to build his own game engine. His idea is to use professional game design, user-generated content, and machine learning to build all of the content needed for the world.
“It’s my dream to build a truly massive open world,” he said.
But rather than just build a large game, they decided to go much bigger.
He added, “Prologue is using Unity to run a tech experiment. Basically, we have a 64 by 64 kilometer world that generated a runtime every time you press play. And it serves to kind of as a proof of concept that look we can do, we need to build our own engine because in order to do a planet-sized world, we have to build it with scalability in mind.”
Prologue is a stepping stone to Artemis, he said.
“To provide the full framework for this planet-scale idea, we need to build an engine, and that’s going to take time,” he said. “Prologue is like a testbed. We’ll try some systems there, we’ll see how gameplay is, and how the world works at that scale. And then when the the engine is ready, we’ll move over to the first demo, or the first alpha or beta of Artemis, which is the planet scale world.”
The indie metaverse
I took a moment to pause and ask how such a small startup could succeed in a world where Microsoft is buying Activision Blizzard for close to $75 billion. How can such a small company pull this off?
“I don’t know. This is a again a question about the metaverse, right?” Greene said. “This was the news about this acquisition where they’re betting big on the metaverse, and to me, it’s like, well, is the metaverse just a portal for games is it? For me, it’s more it’s like a digital experience. It’s a digital place where you can go and create your own space if you want to or experience other people’s spaces. It’s not so much about playing specific games or just being a list of games.”
Greene believes that any metaverse being created should be open, but he sees the big players trying to get their catalog of games into the metaverse.
“That’s not the metaverse, or at least that’s not what I have in my head,” he said. “I think maybe I am one of the people who can create something like this because I’m not really doing it for profit. If that makes sense. It’s more that I really believe in this kind of open space where people no matter what device or hardware you have can access it. And it’s a space for everyone. And I think it would have to come from an indie rather than from a big multinational.”
He added, “They all have different goals. Battle royale, Dayz, Dota, and CS:GO all came from modding. It came from someone’s crazy idea, and they had this passion to bring it forward to much more people. And I think it speaks a lot to where these great ideas come from. It comes from some nutcase in the corner working for years trying to kind of prove an idea.”
Greene is definitely counting on the creativity of players themselves. In fact, Greene will create a world full of systems that will enable gameplay, but he won’t put gameplay in the world itself, or any kind of narrative. Instead, that will be up to the players to create, once they discover what they can do with the systems in the open world. If they want to rope off an area, build a town, and then create a battle royale game in that space, they can do so.
“The creativity of players inspires me daily. So I’m not really worried,” Greene said. If you give them systems, they will make use of them. I was thinking about just different layers. If we can make the world, then we can start adding stuff on top.”
I noted that battle royale had intense gameplay, where you put a circle on a map and make it smaller and smaller, forcing the players to battle each other as the space got too small so that only one survived. That was a narrative.
“I’m a big believer in emergent gameplay,” Grene said. “I’m trying to create a world where emergence comes first. The fun is what you make it. What you see in the world, rather than what I’m telling you, is where the fun is.”
I asked Greene to drill into the technology better. He noted how much time it would take to manually build out the worlds. So his team looked at machine learning because procedural generation is interesting, but it gets a bit noisy and a “bit samey.”
But if you give machine learning a unique suggestion every time, and then you generate a world in real time around the player, then it becomes to big a challenge. You would have to ship a hard drive with the game. That’s just not feasible, Greene said.
“So we have to figure out a way to generate the world around you as you move through it,” he said. “We have a low res version of the world. And then we use various agents to optimize it, add population, add locations, and add all the other gameplay that you need to find in the world. And that’s been our breakthrough, and we just have to provide the right data to this agent, and it will generate the world or the population trees, or the locations.”
Greene wants to be able to fill the game with systems and add hundreds of thousands of players. He knows that will take time, as it is beyond what current games and technologies can do.
“It just requires fundamentally rethinking how we’re doing things,” he said. “That’s what we’re doing with our engine. We’re a component system and data-oriented design where we’re going for scalability first. So as hardware gets faster, our engine gets faster. As new software comes available or new ideas become available, we can integrate these things with great speed instead of being held up by something.”
His team has been researching the metaverse and blockchain. He thinks that blockchain will be an essential part because the metaverse is a set of standards and blockchain and smart contracts are the easiest way to enforce the standards in the digital domain.
“Blockchain is of course necessary for the metaverse,” he said. But we’re taking things slowly and we’re not doing things in a usual way because that’s how we achieve a planet-scaled world at a fidelity that I want which is real world.”
I asked Greene why he was being so transparent. He said that this is one of the things that blockchain enables, as the technology is both secure and transparent as a digital ledger. Everyone can see it, and it can be used to set up organizations where everyone has a voice in building something.
“It’s really essential in building a world for everyone,” he said. “I’m building this digital space that should be accessible by any device you have. It should also be open and I think players should have a voice. We have to set up the world and make sure it’s balanced. But the players can have a voice. Blockchain gives players this kind of vote. But I am very much about open development. I want to share what we are doing. I think sharing this tech with the wider public is the way that we all make a bigger game.”
When players move in between worlds in the metaverse, Greene thinks that machine learning will be able to translate what they have in one world so that it works in another world.
A digital twin for free?
I noted that Nvidia has some very ambitious plans to build a digital twin of the Earth, using its Omniverse simulation technology. It can create the assets for the world and it is standardizing a way to share 3D data via Pixar’s Universal Scene Description standard.
Nvidia plans to create a digital twin of the Earth with meter-level accuracy so that they could model climate change, use their AI and other supercomputers, according to CEO Jensen Huang. And once they create it, I wonder if they can just hand it over to Greene.
“My first thoughts were always to go to NASA for data because NASA has cameras have generally been pointing down at the Earth for 90 years or so,” Greene said. “they’ve been looking down rather than out. And so there’s lots of data about how the real world works. And I always wanted to make a massive, realistic world. So why not use the best reference we have, which is the Earth?”
He added, “I would hope and I believe that our tech is built in a way that we can take new tech like this and integrate it in a relatively seamless way. Because we’re doing everything in a modular fashion. It’s all data driven. So I would hope this kind of tech when it comes along, we can use those. There’s so many interesting things being done, especially with machine learning and procedural generation of the world. It’s exciting to me, because it gives me hope that maybe my ideas aren’t so crazy.”
If he could snap his fingers and take the digital twin of the Earth and procedurally generate a new planet, and have it come out very realistic, that would be pretty cool. That is the goal the company hopes to achieve, he said.
As for making sure players behave in this open world, he said the topic has some big questions.
“How do you moderate? And how do you lower these unsavory elements so they don’t get a voice,” he asked. “It doesn’t happen as much in the real world. If I go to a supermarket, I’m not worried about someone screaming at me. But this is tied closely to like this idea of digital or decentralized identity, where anonymity encourages this, right? It emboldened these people because no one knows who they are. But in our world, your identity has a significance. Maybe it will curtail what you’ll do because you won’t be able to access the world.”
He said it triggers a debate about how much control to give the users. At the outset, the team will be inward-focused as it figures out the right systems and makes sure everything is stable.
I asked him what he would like to see happen once he finishes this world and sets it loose.
“This is the beauty of emergent behavior,” he said. “I don’t know, and that excites me. I’ll put the systems in place and how players will use that is going to be eye opening. Like, I remember when we were first discussing about putting flashbangs into PUBG. Someone at the time said, ‘Oh, my god, I can’t wait to throw a flashbang at someone who is driving and see them crash into something. It took two years. But I finally saw this happen. And as I said before, it’s their imagination, their creativity. I just want to give them the tools to allow them to just go wild.”
As for problems to solve, he noted that multiplayer is a huge problem. He said the team would do its own research and see what it can do, but it will leave multiplayer to the last as it is the hardest to solve. He hopes to take advantage of new ideas that emerge over time.