Google's Project Genie experiment allows creating interactive worlds with generative AI

At the end of January, Google revealed Project Genie - their new experiment to have generative AI create explorable interactive worlds. So instead of you making models, textures and writing code the idea is that you write a prompt and it's all generated for you.

YouTube videos require cookies, you must accept their cookies to view. View cookie preferences.
Accept Cookies & Show   Direct Link

Tools like this as they improve and expand will only open up a bigger can of AI worms, as showcased by the likes of Jay Peters from The Verge (via Bluesky), who used it to create some legally dubious Nintendo-inspired worlds. There's many such examples across social media of people using it to make their own versions of worlds created to be very close to IP from various video game companies.

Currently Project Genie is available to Google AI Ultra subscribers in the US. So it's quite limited, and as an experiment it's not finished. Google said in their announcement that some of the features they previously announced like "promptable events that change the world as you explore it, are not yet included in this prototype".

This has seemingly caused a little investor panic as stocks from various companies dropped including Take-Two Interactive, Roblox and Unity as reported by Reuters.

In a recent earnings call Q&A session, Take-Two were asked about generative AI and what they thought about it for their business in regards to threats and opportunities. Executive Chairman & CEO Strauss Zelnick replied:

[…] Well, I have to admit, I'm a little confused. The video game business since its inception was built on the back of machine learning and artificial intelligence. We create our games in computers with technology. And ever since questions began about generative AI about 18 months ago, I've been incredibly enthusiastic about what the future can bring.

As it happens now, we're actively embracing generative AI. We have hundreds of pilots and implementations across our company, including with our studios. And we are seeing opportunities to drive efficiencies, reduce costs and create the opportunity to do what digital technology has always allowed, which is the mundane tasks become easier and less relevant, which frees up our creators to do the more interesting tasks of making superb entertainment.

The history of the interactive entertainment business has been one of great creators using technology to do amazing things to please audiences. And that's our job around here, and that remains unchanged, except perhaps accelerated. Just a reminder, our strategy has three parts: be the most creative, be the most innovative and be the most efficient company in the entertainment business. And generative AI squarely falls within the category of innovation and is already moving into the category of efficiency. I'm hopeful that it will also move into the category of creativity, as it allows our creators to use digital tools to expand what we do to make it even more beautiful and even more engaging and even more exciting.

And circling back directly there was a question about Project Genie with Take-Two's President Karl Slatoff saying:

So, in terms of commenting on the specific technology, I think I don't think we're going to go into the great details about the tech differences because, frankly, Genie is early in its iteration at this point and trying to make a comparison to a game engine is just really - they're not even in the same ballpark. Genie is not a game engine. And I would - it's very exciting technology. And I think it's - the question is how can it benefit our creators. And I think there will be a moment in time where that will become more defined. It certainly doesn't replace the creative process. And I would say, look, I mean, it looks to me more like a procedurally generated interactive video at this point. There are limitations and Google has said as much. So to compare the technologies, I think there's really no way to do that because they're so far apart.

And there are so many more elements to game development that go beyond world creation. And the question is what is the world creation. So even beyond world creation, there's everything else that's involved. There's the storyline, there's emotional connection, there's vibe, there's mission structure. All of those things you cannot capture through AI and certainly not through a world builder. So that's just a very, very small component of what we do. And if this tool bears out, it will make a component of what we do all that much better and more efficient.

As experiments like this expand, we're going to see a lot more created with it - for better or worse.

Even if the generative AI bubble does eventually pop, I do think it will end up changing the way everything software-related is made. It does have a number of genuine uses, it's just AI slop makers have a tendency to over-hype it and use it in places where it's not needed as various companies try to justify it existing in their products (while also raising prices - wonderful eh!).