Tim Sweeney, the CEO of Epic Games and one of the game industry’s tech savants, has traveled the world to speak about his vision for a world of connected games with massive simulations and realistic digital humans who look and behave like real people.
Sweeney gave a talk at the recent Devcom event in Germany. I caught up with him at the company’s booth at Gamescom, the big game trade show in Cologne, Germany.
We talked about his vision for the Metaverse, the virtual world envisioned by Neal Stephenson in his 1997 novel Snow Crash. Sweeney thinks it’s possible to build such a world and the digital economy that goes with it, but he believes it shouldn’t be owned by a single company. He has been outspoken about ensuring openness in gaming in the past couple of years.
We also discussed the Leisure Economy, based on the notion that large numbers of people could be paid to play games in the future. And we talked about the launch of Hellblade, which uses Epic’s Unreal Engine 4 technology to create realistic digital humans, and Epic’s own Fortnite game, which generated several times more initial fans than Epic anticipated.
Here’s an edited transcript of our interview.
GamesBeat: What’s on your mind for the keynote?
Tim Sweeney: I’m just talking about expectations for the next 10 years of games and engines and the industry. I touched on themes we talked about at your event, like the importance of digital humans in the future, to re-create people naturally over circumstances both real and simulated. The importance of human photorealism, and the expectation that AR is going to be a huge platform. It’s eventually going to supersede smartphones and PCs and all other display devices. Hopefully that will lead to the development of the metaverse, hopefully on the basis of open standards, and not one powerful company controlling everything. You know me. My typical themes.
GamesBeat: Since May, all of that has been interesting. The blockchain part of it, that’s come into focus in the last few months. Everyone’s going nuts.
Sweeney: There are some incredibly powerful ideas there. I feel like they’ve only begun to be developed. There’s Bitcoin, but the more interesting one is Ethereum, which runs general programs in the blockchain. Instead of mining coins by verifying transactions, miners are running programs specified in the blockchain and computing the results of those. They can be anything from programs to effect a single transaction, exchanging money for a good, or they can be entire distributed autonomous organizations that execute governance rules designed by shareholders and can be extended over time by majority vote. You can implement these arbitrarily complex things.
You come to the realization that the blockchain is really a general mechanism for running programs, storing data, and verifiably carrying out transactions. It’s a superset of everything that exists in computing. We’ll eventually come to look at it as a computer that’s distributed and runs a billion times faster than the computer we have on our desktops, because it’s the combination of everyone’s computer.
GamesBeat: Jez San, his company FunFair built a poker game on top of Ethereum. They just run a random-number generator. They don’t have to take deposits because they know if the players paying in this currency, and they can pay out immediately.
Sweeney: A lot of things are now possible. But there are lots of limitations. The Bitcoin blockchain updates once every 10 minutes. If it’s going to run simulation code for games, it better update about 60 times per second. It stores 10 gigabytes of data from inception to present. You’re going to need terabytes of data produced every day if you want to scale it up for virtual worlds.
I think all of these things can be accomplished, and the tools for accomplishing them are well-known. People just need to go and build it now. It’s a wonderful model, to be able to build distributed systems on that basis, where everybody can contribute and nobody is in control.
GamesBeat: One company was proposing using it to replace the app stores, just doing transactions between developers and players. Developers get paid in 60 hours instead of 60 days, and get a higher percentage.
Sweeney: You want to be paid at 60 frames per second. I think that can be accomplished. Someone pointed out that there was a Bitcoin transaction on the blockchain, a $100 million transaction, and there was a transaction fee proportional to the computation volume of the transaction. It was 70 cents.
GamesBeat: We also have to worry about some kind of bubble here. Survive that and then it becomes useful.
Sweeney: There may be bubbles. There might be a bubble right now. But the underlying technology is solid. Something awesome will continue to evolve out of it. We all remember the tech bubble of the late ‘90s, but companies like Amazon survived. Wherever there’s strong, enduring value, it can last through that kind of turmoil. Ironically, almost all the startups that failed in that bubble exist now in different forms. There’s a new version of Pets.com. Uber was an idea from back then. It just took 15 years of people’s adoption of technology before the world was ready for it. I don’t worry about bubbles. Those are just temporary problems.
GamesBeat: It’s a tool to pull power away from centralized authority, which is something you’ve worried about. Platform owners have proprietary technology.
Sweeney: And they use it to exert arbitrary control. They tilt the market in their favor. They censor people whose political views they disagree with. It’s a big and growing problem, the amount of power possessed by Google and Facebook. President Eisenhower said it about the military-industrial complex. They pose a grave threat to our democracy.
GamesBeat: I heard a talk by Jon Jacobs, the guy who bought his island in the virtual world and turned it into a nightclub. He articulated something I thought was interesting about this leisure economy we’re heading toward. Why do we have to work? Why don’t we get paid to play? If you have the beginnings of things like esports, streaming, and user-generated content, you have all these people getting paid to play. How do we go from just a few to many more people getting paid to play?
Sweeney: There’s real value in that. If you look at why people are paid to do things, it’s because they’re creating a good or delivering a service that’s valuable to somebody. There’s just as much potential for that in these virtual environments as there is in the real world. If, by playing a game or doing something in a virtual world, you’re making someone else’s life better, then you can be paid for that.
We need to start rethinking the way we structure these large-scale game economies, especially as they get bigger and more complex. They should not simply be a means for the developer to suck money out of the users. It should be a bi-directional thing where users participate. Some pay, some sell, some buy, and there’s a real economy.
If you look at what Valve has done with CS: GO and other games, they have a vibrant economy. Lots of people earn a living in those games. When you buy something, you’re not just buying it from Valve. You’re often buying it on an open marketplace from some other seller. It’s a very complex economy. There’s a lot of potential in that. We need to move toward a model like that in order to scale up to something like the metaverse. As long as it’s just the machination of a single company, there’s not going to be this bidirectional exchange of value. We need to keep that circulating.
GamesBeat: What do you think of the notion that an entity like Facebook — they have their users. Those users interact with Facebook. That attracts more users to Facebook. Facebook makes money from advertising. Facebook should then provide a basic income to people who are engaged on top of it.
Sweeney: If there were not such strong network effects, economic effects locking Facebook in, I think you’d see robust competition. I understand the concept of advertising. It makes sense. Companies want to reach customers with information about their products. It’s a completely legitimate and valuable notion. But why should Facebook get all the money from that? Shouldn’t the user get some?
GamesBeat: We’re paying to access something when they should be paying us.
Sweeney: Imagine an economy structured around that. Say Facebook costs $30 a month because it’s running all these servers, but you can easily earn $100 or $1,000 viewing advertisements. What if everyone set their own price for ads? If you want to put an ad in front of my face, my time is valuable, so that’s five bucks. Somebody else might be willing to poke around with an ad for 50 cents. Everybody gets to set their own deal terms. You’d have an interesting information economy with something like that.
The same thing applies to all aspects where the internet is monetized through something other than users paying service providers. Look at Google ads. Shouldn’t the user have the knob to control that? What if each search cost five cents, but you got 10 cents back by viewing ads? Everyone could make a decision that’s best for them. I’d be very happy to pay for these services if they wouldn’t sell me out to their data brokers.
GamesBeat: That’s back to the danger of proprietary platforms. I wonder what triggers some of this to happen. It seems like the tipping monetization happening on Twitch — it’s a brand-new source of revenue. This company called Stream Labs says they’ve processed $100 million in tips for streamers in the last year.
Sweeney: It’s really popular. That’s not even a closed economy. You’re not actually getting any incremental value for these tips. Imagine if you actually got additional benefits for doing that – recognition or power or something like that. I think that could go much further than it’s gone before.
These technology revolutions come in waves. The social network could have developed 10 years earlier than it did. You just needed a generation of people that were okay with putting all their personal information into these services. The limiting factor wasn’t technology. It was social expectations.
Now we’re approaching 10 years of the blockchain. As these technologies become more widely accepted, we’ll see a whole new generation of distributed systems, which really changes the game — away from closed to open. Instead of building a business that extracts money from a service, you build an economy in which everybody can be rewarded for participating in many different ways.
GamesBeat: I finished playing Hellblade. That was outstanding.
Sweeney: It’s amazing. It’s one of the best-looking games ever created.
GamesBeat: It speaks to what you were saying about digital humans. I didn’t find any fault with those images.
Sweeney: There’s an incredible amount of talent out there. Recognize that most of that data was motion-captured in real time using this new hardware. Now extrapolate three years ahead. Maybe your typical VR and AR headset will be able to do that same level of motion capture, instead of a $10,000 rig for game developers. Imagine that technology is available to everyone. It’s happening fast.
GamesBeat: Were you satisfied with the results there, or do you feel like it still has some distance to go to be as perfect as it could be?
Sweeney: I think it’s ready for prime time. It’s proven out, many times now. It’s only going to get better over the years, especially as we introduce more logic into reconstructing facial motion from these video signals being fed in. Soon you’ll have the ability to do that without any preparation – without markers on on the actor’s face, for example. It’ll improve by leaps and bounds.
GamesBeat: You still need an artist to stylize it in some way. The original The Lord of the Rings had some of that, where they tried to make the orcs look real, but they still looked a lot like people wearing orc makeup.
Sweeney: Faces are the hardest problem in computer graphics. Your brain has such hardwired expectations of what they do.
GamesBeat: You guys launched Fortnite. How has that turned out?
Sweeney: It’s on track to be Epic’s most successful game ever by the end of the year. Several times more users than we expected came in. We had to really scramble to keep the servers running at that scale. It was great experience. This is just the very beginning. We have a bunch more game modes coming, which should be interesting and innovative. The whole aspect of building in Fortnite creates a really different dynamic compared to most games. Building structures, defending them, you can imagine all the multiplayer implications.
We’re very happy with the transition we’ve been making, from being a lone developer who works with a publisher and puts games on retail store shelves to running our games live as services where we can constantly improve the game based on feedback. It’s been a great step. Fortnite is getting better every week.
GamesBeat: The early access on that one at some point converts to free-to-play, right?
Sweeney: Yeah, at some time next year we’ll go completely free.
GamesBeat: It almost seems like a paid game, just the way so many people are piling in to early access.
Sweeney: It’s an area where developers have a lot of territory to explore. Early access is one option. Free-to-play is another option. There’s a wide range of possibilities. It’s expensive to build a game on this scale now, and it’s imperative that we find ways we can pay it off.
GamesBeat: Was Fortnite a lot more expensive to build than, say, the first Gears of War?
Sweeney: Much more expensive. What we saw over the last generation — the economics started out incredibly attractive. The first Gears of War game cost about $12 million for Epic to make and earned us more than $100 million. But by the end of that generation, the costs had blown up to almost equal the profit a game could make. At this point, in this generation, I don’t know how you make a game at the triple-A level unless you sell 12 million copies or more. The teams are hundreds of people, sometimes more than 1,000. Budgets are more than $100 million, sometimes several hundred million. The scale of sales needed to justify that is crazy.
It’s awesome to see so many different games of different sizes being built. Everything from Astroneer, which was number one on Steam and built by six people, to PlayerUnknown’s Battlegrounds, which is not a huge team. Indies can succeed with great games at any scale. We’re constantly reminded of that. The only thing all great games have in common is fun. Anything from Angry Birds to Grand Theft Auto, they all provide fun in a different way. There’s a 10-person version of a fun game and there’s a 1,000-person of a fun game, and they both work.
GamesBeat: Do you think some of these very successful free-to-play efforts might be useful to finance other big things? Say that 1,000-person project is kind of iffy, but you want to do that to push the state of the art. Something like Game of War can make a billion dollars. Why not use that to finance the big game?
Sweeney: Absolutely. We need to be constantly exploring business models that work. As the industry evolves, business models change. Epic’s survived through a number of different ones over our history, from shareware to PC retail to big-budget triple-A console and back to indie, publishing our own games for ourselves. You have to constantly change the rules.
GamesBeat: VR seemed to cool off a bit in the last few months, and esports investments are picking up. Do you see something happening there?
Sweeney: There’s definitely a lull in VR right now. After the initial wave of sales, Oculus and HTC don’t seem to be making a whole lot of new business. We’ll see what happens next. I hope we see a new wave of VR hardware that substantially updates the spec and improves the visual quality through higher resolution and better optics, and improves the accessibility by making things easier to set up. At some point we’re all going to need to move to inside out tracking so you don’t need to set up a bunch of cameras around the room, and that will be more user-friendly.
Every step we make is a significant iteration toward a better product. Miniaturizing the technology so it’s more toward being glasses and less toward being a helmet.
GamesBeat: People are saying we’ll have to get things like 5G in place before some of this is possible.
Sweeney: There’s also ARKit now, augmented reality that reaches more than 100 million iOS users. That’s going to be a huge frontier for developers of these types of games. That’s a huge user base, and we’re just in the early days. We’re going to see entirely new genres of games invented over the next year that take advantage of AR in entirely new ways.
Disclosure: The organizers of Devcom paid my way to Cologne. Our coverage remains objective.