Rockstar Co-Founder Calls AI Overhyped — Still Puts It to the Test for His Next Game

Dan Houser is back in the spotlight, but not for another GTA. The Rockstar Games co-founder popped up on Channel 4's Sunday Brunch to promote his new novel and ended up giving a refreshingly blunt take on AI in games: he is using it, sure, but he does not think it is the magic wand a lot of people are selling.

Houser's read on AI: useful, not miraculous

Houser told Sunday Brunch (via Eurogamer) that his team is experimenting with AI tools on an upcoming game, but he is wary of the hype cycle around the tech. In his words:

"We are dabbling in using AI, but the truth is a lot of it's not as useful as some of the companies would have you believe yet. It's not going to solve all of the problems."

He went further, saying a chunk of today's AI marketing exists to "sell AI stock" and persuade investors and the public that the revolution has already arrived. He did give credit where it is due: some AI use cases are genuinely impressive. But he also pointed out that plenty of tasks people now credit to AI were already doable with existing tools.

This lines up with things he has said elsewhere about writing: he is not into handing narrative over to large language models, arguing they do not generate real ideas or the kind of big, connective concepts humans bring to a story.

So, what is he building?

Houser's new novel is titled "A Better Paradise - Volume One: An Aftermath," and it centers on an AI that is, as he describes it, self-loathing. The game he is making is set in the same universe but does not retell the book verbatim. As he put it: they are making a game in the same world, just not the same story.

The project sits under Absurd Ventures, Houser's post-Rockstar company, and it is designed as part of a bigger, connected slate that spans novels, podcasts, live-action projects, and games. The game itself has been in development for about 18 months. And yes, in addition to being a theme in the fiction, AI is something the team is testing in the actual development process too.

The broader picture: everyone is experimenting, not everyone is aligned

Houser expects games to split into two tracks: projects greenlit for creative reasons and projects greenlit to maximize profit, with both coexisting. That tracks with where the industry is right now. Publishers like Activision, EA, and Krafton are openly weaving AI into their pipelines, while the rest of us are trying to sort out what that means for actual games. Meanwhile, debates over AI-generated art and voice work keep flaring up as studios push and players push back.

• Where he said it: Channel 4's Sunday Brunch, while promoting "A Better Paradise - Volume One: An Aftermath" (via Eurogamer)
• The stance: AI is interesting but oversold; not a fix-all and sometimes hyped to sell stock
• On writing: he does not want LLMs doing the storytelling; says they lack real ideas and large-scale concepts
• The game: set in the novel's universe, different story; in development at Absurd Ventures for about 18 months
• The plan: a connected IP spanning books, podcasts, live-action, and games
• The industry: big publishers are integrating AI; the field is likely splitting between creative-first and profit-first projects

Short version: Houser is curious about AI but not drinking the Kool-Aid. It is a tool, not a revolution, at least not yet. That cautious, pragmatic approach makes his new world about a self-loathing AI sound even more intriguing.