This AI startup envisions '100 million new people' making videogames

If you could type a series of sentences into a computer and have it generate a videogame for you, would you?

Well, you can—kind of. Tesana is one of a few companies promising that we can all make games simply by typing prompts into a generative AI system. In tandem with third-party AI services like Claude and asset generators, its proprietary game engine "translates descriptions of environments, mechanics, characters, and rules into structured game code."

Judging by the examples on its website, Tesana's users are mostly testing it out by generating garbled versions of other games. Valdenholt, for example, is like Skyrim if you took it out of the computer, stomped on it, and then smushed some of the pieces back in. But it is technically a game-shaped piece of software.

The service attracted around 10,000 paying customers in its first couple weeks, according to co-founder Johannes Vermandois, who has previously founded a skincare company and an AI marketing platform. Vermandois showed off Tesana at the Game Developers Conference in March.

Vermandois thinks Tesana will be useful for prototyping games, and imagines that the speed at which users can iterate on ideas will lead to the invention of entirely new kinds of games. He also foresees full, finished games being produced with prompts, and wants future versions of the software to be capable of creating games as good as, for example, Valheim.

"If we can have 100 million new people being able to make games, you know, that will be an exciting thing," Vermandois told PC Gamer.

For Vermandois, it's a given that more people making more games is good. He agrees that many of those games will be bad (lots of bad games are made without the help of AI, after all), but thinks the quality bar will rise as more people are able to turn their ideas into software.

"I feel like indie games are really the future of gaming," he said. "And I think that these kinds of tools and platforms that we're building at Tesana could probably help more people make games, and then the sheer volume of games will increase … And I think the quality bar of games will increase as well, because there's so many more games."

I'm not so sure about that myself. Game development is more accessible than ever—you don't need a PhD in computer science to make something in open-source engine Godot—but it has never been easy. It seems to me that generative AI offers ease in exchange for control over the details, which is no small thing if you believe that "art is all in the details."

If 999,999,999 other people can generate the same game as you, why would anyone care about yours?

"I think people will still pay for other people's creative vision," says Vermandois. "Like for the same reason people go to a movie: You want to experience someone else's story or vision. I think that's still true. For example, I've been playing around with [AI music generator] Suno. I can make some really cool music myself that I like to listen to, but then I also listen to, you know, Spotify. I like artists on Spotify."

It seems that where we differ is in how much we consider prompting generative AI systems to itself be an expression of creative vision. Would anyone bother with Spotify if it were exclusively full of other people's Suno-generated tracks?

Vermandois is obviously more optimistic about the technology than I am, but he doesn't claim to know what the future holds, either. Regarding worries that AI processing costs will go up once companies like Anthropic can no longer subsidize their users, he's hoping that model costs instead go down. And on how Tesana will compete with Roblox and other big companies slapping generative AI onto their products, he's holding onto a startup founder's necessary optimism that the small guy can beat the big guy.

Success for Vermandois will be creating a new market for new kinds of games, and "a new hobby for a lot of people." For now, the first test games made by Tesana users have nothing to offer the rest of us, but as we often hear when talking about generative AI, "it's very early days."