China Steams Ahead in Gaming; MetaHuman Creator and the Uncanny Valley

Welcome to this week's episode of 16 Minutes, where we talk about tech trends in the news, what's hype and what's real, and where we are in the long arc of innovation. I'm Zoran, and today we have two brief segments taking up less than 16 minutes that touch on trends in gaming, online worlds, and entertainment. There are two news items this past week, both related to gaming, now and next. One is that Epic Games' 3D graphics platform Unreal Engine revealed an early peek at MetaHuman Creator, which aims to make creating digital humans easier. The other news item is that Steam, which is the cloud library and marketplace for buying, selling, and storing games that was launched by Valve in 2003 and is now the largest distribution platform for PC gaming, just released an official Chinese version. So what does this all mean? A16Z partner Jonathan Lai, a former product manager at Riot Games, joins both segments to share a quick take on these news items and the trends overall. So, Jon, you actually launched a competitor to Steam in China a few years back. Why the Chinese version of Steam and why now, given that the international version of Steam has been available for quite a long time and has been very popular?

Basically what's happening now is as the domestic games industry in China matures, it's by far the largest country just in terms of the gaming market. I think, uh, the last statistic that I read was that there were 660 million gamers in China, which is more than the entire population of the US, Japan, and Germany combined. But All foreign media platforms are required in China to go through a licensing process with the censorship bureau to make sure that all the games that they have on the platform comply with censorship policies, and Steam had not gone through that for the international version. But that largely didn't stop people from using the VPN to gain access to the international version of Steam because the servers were actually just offshore in Hong Kong and Taiwan. And so I think there was a risk that if they did not have an official version in China that they could just have all of their games either taken down or have the platform banned outright.

It's a very scaled-down version.

So the feedback over the official version of Steam China has been pretty mixed, and I think you identified the right reason. Steam China launched with only 53 games, which were the only ones that have actually gone through the censorship review process, and so this is the batch that they could launch with. And then on top of that, they also turned off many of the social features which made Steam sticky. The guides, the user-generated content, like all of that has been turned off. I think the open question is whether Chinese users will just ignore Steam China and instead just continue to use the VPN to use the international version of Steam, which has the entire catalog of all of the games in the world.

What do you see as the major impact for those outside China who want to sell into this massive market?

So if you are EA, Activision, you know, Riot Games or Epic, and, um, historically you've been reliant on Steam in order to reach users in mainland China, all of a sudden the presence of sort of Steam China, which is compliant with censorship policies, sort of calls your status quo into question, right? Like, can you still use the international version of Steam to reach your audience in China? Should we try to move our game onto Steam China? And just in case the international version gets blocked or banned, how should we change our policies regarding China? So I think that's probably more the concern.

What's interesting is that Chinese indie games have been a phenomenon in recent years. Just the latest example is Tale of the Immortal, which is a Chinese-only RPG that launched last month on Steam and took off, over 170,000 concurrent players. Many of these indie games are beating out much more established games.

Yeah, just in this past year we've had multiple high-quality games from smaller Chinese developers that not only were successful in China, but but also reached massive commercial success globally. The most well-known one of the last year is a game called Genshin Impact, which is a gorgeous sandbox action RPG game, which just won best game of the year from both Apple and Google. And the latest sales numbers on it is reported that that game had grossed over $400 million in revenue on mobile alone in just 2 months since launch. Like it launched very late last year. And this is a smaller Chinese studio. This is not Tencent or NetEase or Perfect World or one of the large AAA game publishers. Chinese game developers for the longest time had an unfortunate reputation as being copycats for the most part, who for most of their portfolio cloned more successful games from places like Korea, Japan, and the West. But I think that long-term, the rise of the Chinese indie dev scene will probably have pretty large implications on the global industry in terms of competition. To put it simply, if you have Chinese game developers that are working 996 for folks engineers are working 6 days a week, 9 AM to 9 PM. That's the sort of 996.

It's a version of 24/7.

That's right. And they're paid a fraction of the cost of developers in the West, and distribution is no longer a gating factor. And they have a sustainable cost advantage in terms of keeping their operational overhead low. And so I think all of that is a recipe for just a really tough competitive scene if you're an independent Western developer. They've got a much higher cost basis.

Bottom line, of course, Jonathan, what's the takeaway here?

I think it's just generally a sign that China is a really important market, and even the world's largest PC game company can't do without it. And so, um, you know, if you're a game developer or a publisher and you're thinking about ways of, um, building games for China, this is yet another option and something to think about.

Great. So our next news item is about Unreal Engine from Epic Games. Epic Games is of course the maker of Fortnite and other popular games. Unreal Engine is Epic's real-time 3D 3D graphics platform, and recently it unveiled a new browser-based tool called MetaHuman Creator that it says can let creators make a digital human in less than an hour. There was a lot of buzz about this in the technology, gaming, and animation communities. So let's quickly dig into why and what's hype and what's real here. But first, John, what is it and why does it matter?

MetaHuman Creator actually sets out to solve one of the hardest tasks in 3D content creation. Which is constructing truly convincing digital humans. It is a step beyond the CGI that was being used in video games and movies just in the last couple of years. Probably the best way to describe it is that it looks right now like a motion-captured human. And that right now is the pinnacle of 3D digital human construction. When you have an actor or an actress that's putting on a suit with light bulbs and they're in a big studio and there's a computer system that's capturing their every movement and facial expression and then translating that into a 3D model. That production process is expensive, it's time-consuming, it has to be done on set with all of this elaborate heavy equipment. And if you can take that production process and compress it into just basically software, that it's just a software web-based editor powered by the Unreal Engine.

What do you think this will lead to in the short term when it starts rolling out?

So the near-term use case would be game developers that were already trying to make realistic-looking human avatars and beings in their games will probably sort of jump on this. It's very low-cost to experiment with, and if it works, we'll see more metahumans in video games that are coming out in the near term. I think long-term, this could potentially become a lower-cost replacement for real human actors. That could have profound implications on the filmmaking ecosystem. The majority of 3D-modeled human faces in video games and animated series, and even in some TV film where they use CGI, have basically been unable to sort of cross what we call the uncanny valley, which is where our brain detects that what we're looking at is realistic looking, but something throws us off, it makes us uncomfortable. Many movies today use lots of human extras on scene. If you could actually just take those extras and populate them instead of metahumans, I think that would be a pretty sizable change in the way movies are made.

Okay, so let's tease apart what's hype and what's real. We've seen movies that use CGI to populate crowd scenes, right? World War Z with hordes of zombies. But how do we get from there to what you're talking about, which is much more realistic looking characters populating scenes either in a film or like a gaming environment?

Yeah, there's still a lot of questions. Like we don't know how the animations will work, for example. It looks like this is a great tool for creating a 3D model, which from the way it looks from the trailer, literally looks like a character creation editor from an RPG game. This library of pre-created assets, everything from hairstyles to like noses, to beards, to tattoos, the different types of skin, and then you can basically drag and drop them together to create your ideal sort of human avatar or human being in this case. But then actually animating it is the next big step. How will all that work? You know, when the character is walking and jumping and running and firing a gun, there's a lot of detail in those interactions, which still remains TBD at this moment.

What does this mean for the next iteration of the internet? There's this idea, right, of like shared virtual worlds, Ready Player One, essentially, where gaming, VR, AR, social, they all kind of come together. In this almost totally immersive way of the metaverse, basically?

Right now, like, it's virtually impossible to actually build something that's metaverse scale, several universe-sized virtual worlds, right? You have a Star Wars world, you have a samurai world, you might have a cowboy world, and then the whole notion is that you can go back and forth between them. And then to create any of these worlds would just require so much content, especially if you do it sort of Ready Player One style, where this is a 3D immersive environment that people are able to sort of log into and participate in as if they were in real life. I think the potential for this is that it vastly opens up the playing field for smaller creators who previously would have had to work at large corporations of motion capture budgets and whole studios around animating these digital humans. But it enables sort of just a small creator to create a realistic looking human being and then deploy that into TV, film, animation, video games. That sort of paves the way for more content to be created that could one day lead to the metaverse.

Essentially what you're saying is the large corporations, they can't quickly enough produce all the content that's needed to create these rich interlocking universes that create the metaverse. And what's needed is like this sort of decentralized ability for many more people to do this.

Yeah, the holy grail when it comes to creating the metaverse is like we already have sort of these game design concepts and gameplay around how people interact in virtual worlds with each other that's still stuck in that uncanny valley phase. Like right now, some of the most popular games are these open-world sandbox games. So Assassin's Creed, for example, is famous for taking place inside historical real-world cities filled with what the games industry calls non-player characters or NPCs. They're milling around in the streets, they're the shopkeepers, they're guards, they're people like on the way to work. And most of those NPCs have pretty rudimentary sort of models and AI. They all look pretty much exactly identical and they're just sort of wearing different outfits. And they behave in very simple scripts. But I think if you had more realistic-looking humans, that would do a better job of not breaking the immersion.

Okay, so bottom line it for us, what's the main takeaway from this?

I think the main takeaway is that digital human creation has been a challenging area, particularly when it comes to faces and allowing creators to easily replicate motion capture work that would have taken months, if not years, at a dedicated film studio. We still don't know a lot about the technology just yet, but something to watch if you're working in a creative industry that involves creating 3D models of humans.

John, thank you so much for being with us today.

Thank you.