你能仅凭提示词建立一家正规的视频游戏业务吗？Meta 认为可以

Listen now on Apple, Spotify or YouTube

In This Edition,
Meta Horizon VP Saxs Persson on…

• UGC vs traditional games

• AI games as a business

• Google vs Roblox vs Meta

Last month, Meta revealed Horizon Studio and Horizon Create.

They’re AI-powered game making tools where people can make games using text prompts and then publish them directly onto mobile and social feeds, including Instagram and Facebook.

“The developer vision is to break down the barriers [to game making] as much as we can using modern tools,” begins Saxs Persson, who is VP of Horizon at Meta. “AI happens to be an extraordinary accelerator for certain things that have been very difficult. A good game is iteration. If we can buy more iteration with powerful tools that make it faster, more precise, and so more people can iterate, we should expect more interesting games, because it doesn’t cost as much. That’s the developer vision.

Persson is a veteran of the user-generated content space. At Microsoft, he was executive producer on the game creation title Project Spark, before taking a leadership role on Minecraft. From there he went to Epic Games, where he was responsible for Unreal Editor for Fortnite, and enabling other creators to build games within that ecosystem (including LEGO Fortnite).

“‘User-generated content’ is not a great phrase for what this really is,” he says. “These are smaller developers making games that people play for billions of hours. There’s a whole generation growing up where this is just part of how you play games. And it’s a new generation of developers showing a modern take on how to make games and what a game should do to be successful.

“From my perspective, it still complements traditional game-making. I have the same respect for traditional games that are more crafted as a statement from the team, versus user-generated content, which is often more measuring yourselves in how people feel when playing your game, and about that direct communication between the community and the game maker.”

The Game Business is a reader-supported publication. Please consider becoming a subscriber!

Meta vs Google vs Roblox

Horizon isn’t the only operator in this space. There are numerous start-ups where people can build games via prompts, and then there’s major players like Roblox and Google.

Persson believes Meta has advantages over its competitors in that it’s been built from scratch for this. “We don’t sit on 20 years of very opinionated codebases and licenses,” he says.

It also has ambitions to go far broader. “Meta understands scale. It’s a company that doesn’t blink when you say a billion people. It just thinks, ‘Why not two?’ That’s the type of company it is.”

But one of the key differentiators between Horizon and something like Roblox is that the various games can look and feel completely different from one another.

“There’s not a particular look,” Persson tells us. “The character is your character, the environment is your environment, the style is your style. We’re very welcoming to ship to the widest possible audience exactly the thing that you want to make, in the style you want to make it. There’s nothing quite like that.

“Roblox is very opinionated from an artistic perspective. Minecraft the same thing, Fortnite the same thing. Fortnite on mobile is not at scale. Roblox is certainly at scale but it’s hard to imagine a rich artistic look compatible with how Roblox went about building its engine. We’re trying to fade into the background to provide the components for the games to be the center, and Meta to be the platform that helps the games get there. Similar to YouTube. YouTube is not opinionated about what content they have. They serve the content that matches up with the viewers’ interest, and gets out of the way.

“And from an IP perspective, that’s the other differentiator, this is a good place to be. You can make the thing that represents your IP accurately in the way that you would want to make it and distributed to the widest possible audience, with very little friction and cost.”

Is it a business opportunity?

I’ve told this story before, but a 12-year-old family member once showed me a Wipeout-like racing game he’d made. He coded it himself, but the assets were all AI-generated. He was so happy as we all played it together over Christmas, and it was a joy to witness. There are clearly deep concerns about the technology, but I can see the benefits, too.

“People that make games are like circus horses,” Persson says. “We smell the sawdust, and we walk in circles, because it’s so addicting to make a game, to make something out of nothing and have people find joy in playing it. It’s the most satisfying thing I can think of. Giving that feeling to more people, I have a hard time thinking that that’s nothing but a positive thing.”

Persson says this isn’t about one-shotting games. It’s about iteration, testing, trying and repeating. Google Playground is similar and even has a partnership with Unity, too, to add even more manual control over the games being built.

But we’re a business publication, and the question remains, are vibe-coded and prompt-based games a legitimate business opportunity for professional game teams?

“It’s absolutely a business opportunity,” Persson insists. “Is it a business opportunity for a triple-A team to switch to this platform right now? No. There’s a particular flavor to it that we’ll start with, but it’s still a business opportunity.

“I spend a lot of time with new indie teams. My pride lies in whether I can inspire people to take that first step, make that first level, and get a group together that’s going to make a magical game. Those are very much business opportunities. We should measure ourselves in our ability to enable people to be entrepreneurs on the platform.

“We should measure ourselves in our ability to enable people to be entrepreneurs on the platform”

“In many ways, it complements the traditional game industry. I don’t think it directly steps on top of it or anything. My controversial opinion is that this is the modern way of making games, which is faster iterations, closer connection to your community, and a higher tolerance for experimentation.”

Persson points to Spotify as a platform that simultaneously ruined the music industry, and opened it up so everyone could be a musician. He thinks video games “can do better than that”, by preserving the premium, AAA space while expanding the UGC sector.

But AI poses plenty of other concerns, such as IP protection. In my research for this article, one investor told me to check out AI game platform Spawn.co. And the first thing I saw when booting it up was an unofficial Pokémon game.

“You can’t operate a platform if you don’t observe all the basic safeguards for preserving intellectual property,” Persson notes. “There are things you can do above and beyond that. My goal is that this will be a first-rate place for IP, because I really think that matters from a player perspective. You want the legitimate experience that feels exactly like the thing. I worked a lot with LEGO and Star Wars in my past, and I certainly have an appreciation for how precise a [company] feels their IP needs to be represented.”

And the other challenge is making sure users, particularly younger users, are protected. This is something Roblox has been wrestling with in recent history.

“The industry is growing up, and you really have to be on the A-game to operate a global platform. The global regulatory landscape is not the easiest to navigate, but it’s there for a reason. For us to develop trust long term is directly proportional to the integrity of the platform, from a content and age perspective, and in terms of business practices.”

We’re entering an unknown future when it comes to game making. The optimists will tell you that AI tools will just unlock a new generation of game makers, potentially uncover new types of experiences that can grow the industry, and ultimately complement the traditional business. The pessimists will tell you it’ll flood the market with low quality content, make it significantly harder for traditional developers to compete, drive piracy, and result in a sharp decrease in the number of professional game makers.

Persson, perhaps inevitably, is in the former camp.

“Every year we take more steps to remove more moats, to democratize the act of making games,” he concludes. “I have no illusion everybody’s going to make a game, but I do agonize over the fact that the people that really have a game in them ought to have the option. So, ten years from now, I want to imagine that the making and publishing of games are democratized to a point that the best games can find a platform and an audience. Ten years from now humans are still making games like they’ve always been making games. Ten years from now a good idea is still a good idea. Ten years from now you’re iterating on your game for six months, but the tools surrounding it will just make it faster, easier, more fun to measure.

“And I think there’ll be a whole new category of games where the agent sits in the runtime, which really energizes me. I want to bring about not just a different way to make a game, but a different way to playit, and we’re just getting to that. From my perspective, that’s where we can really bring innovation to the industry, which is to allow reasoning to influence how the game develops.”

Thank you for reading! Did you enjoy that? Why not consider becoming a paid subscriber!

Subscribe now