Exploring the potential of generative AI in games

Beamable’s Jon Radoff, Inworld AI’s Kylan Gibbs, Berkeley Synthetic’s Matt White and Hidden Door’s Hilary Mason sat down at the GamesBeat Summit to talk about generative AI’s potential in games, now and in the future.

There’s one thing we can all probably agree on when it comes to AI in games. Anything a human might oversee in a game experience has the potential to overwhelm a human from complexity. The perception of generative AI in games is mostly aimed at work usage. A lot of the conversation there focuses on creation, and how it might impact working folks.

But there’s another angle. AI as content administrators and facilitators. Think about a game like Destiny 2, with its bite-sized open world events. They’re fun, the first few times, but the whole process gets a little stale. Imagine what an AI could accomplish if it was there to mix and match content. Or if it could scale whole events based on the amount of people in the area.

This take on AI is kind of central to Hidden Door’s whole business model.

“The thing we’re building could not be built without the underlying technology,” said Mason. “You could create a similar kind of experience, but you’d have to put a human inside of it. They’d have to improvise, and they’d have to be smart. They’d have to have access to a library of information in their head that is really out of the scope for the scale of what we can cognitively process.”

For Mason, it isn’t cutting humans out of the process entirely. It’s just shifting them from a production role to a directing role. More than that, the role of AI isn’t being looked at as content creation itself.

It’s a framework for content creators to create with.

The Wild West

There’s a problem, though. Like many newly popular technologies, there’s a whole lot of everybody doing their own thing. That level of fragmentation is a problem that needs to be addressed.

“The situation right now … it’s a little bit of a Wild West in the generative AI space,” commented White. “AI researchers are not terribly good at creating products. They’re particularly good at creating something and then moving on to something else, and just abandoning it. That helps us to create new innovations and that sort of thing, but the market’s fragmented.”

It’s the same problem VR suffers from. Or the Metaverse as a viable thing. The people really into it know it needs some kind of unification. The people really into making money from it all want their own thing to be the one that blows up.

The potential for AI in games is huge. It’s hard to overstate how huge it could be. If game developers can get the tools out of AI that AI enthusiasts are sure can exist it could be a whole new world for game creators.