AI can’t create good video game worlds yet, and may never be able to

This is it Step backa weekly newsletter covering one essential story from the world of technology. For more news about the video game industry’s opposition to generative AI, Follow Jay Peters. Step back It arrives in our subscribers’ inboxes at 8 a.m. ET. Subscribe to Step back here.

Long before the explosion of generative AI, video game developers made games that could create their own worlds. Think of titles like Minecraft Or even the 1980 original Rogue This is the basis of the term “roguelike”; These games and many others create fast worlds with specific rules and standards. Human developers work hard to make sure the worlds their games can create are engaging to explore and full of things to do, and at their best, these types of games can be replayable for years because of how new the environments and experiences can feel every time you play them.

But just as other creative industries are resisting The future of AI is rampantGenerative AI is coming to video games, too. Although they may never be able to catch up with the best that humans can do now.

Generative AI has become a thing of the past in video games lightning rodWith players getting… Angry at slipping into the game And half the developers think so Generative AI is bad for industry.

However, major video game companies are jumping into the murky waters of AI. PUBG Crafton maker turns into An “AI First” gaming company.EA is cooperating with Stable AI for “transformational” game making tools.and Ubisoft, as part of a major reorganization, is Promising to do so It will make “accelerated investments in player-facing generative AI.” CEO of Nexon, which owns the company that was a hit last year Sagittarius RaidersPut it down Perhaps the most dangerous: “I think it’s important to assume that every gaming company is now using AI.” (Some independent developers disagree.)

Larger game companies often offer their commitments as a way to simplify and help with the game development process, which means… Get increased cost. But the adoption of generative AI tools poses a potential threat to jobs in an already notorious industry Waves to Layoffs.

Last month, Google launched Project Genie, an “early research prototype” that lets users create sandbox worlds with text or image prompts that they can explore for 60 seconds. For now, the tool is only available in the US to people who subscribe to the $249.99 per month Google AI Ultra plan.

Project Genie is powered by Google’s global Genie 3 AI model, developed by the company Stadiums as well “A major stepping stone on the path to artificial general intelligence” that could enable “AI agents capable of thinking, solving problems, and taking action in the real world,” and Google says the model’s potential uses go “beyond games.” But it received a lot of attention in the industry: It was the first real indication of how generative AI tools could be used to develop video games, just as tools like DALL-E and OpenAI’s Sora showed what might be possible with AI-generated images and videos.

In my testing, Project Genie was barely able to generate even remotely interesting experiences. The Worlds don’t allow users to do much except move around using the arrow keys. When the 60 seconds are up, you can’t do anything with what you’ve created except download a recording of what you’ve made, which means you also can’t plug what you’ve created into a traditional video game engine.

Sure, the Genie Project has allowed me to generate Awesome unauthorized Nintendo knockoff (apparently based on online videos on which Genie 3 was trained), which generated a lot of interest Familiar fears About copyright and artificial intelligence tools. But they weren’t even in the same quality as the worlds in Nintendo’s handcrafted game. The worlds were silent, the physics were sloppy, and the environments were primitive.

The day after Project Genie’s announcement, the stock prices of some of the biggest video game companies rose, including Take-Two, Roblox, and Unity. dive. This resulted in little damage control. For example, Take-Two CEO Carl Slatoff strongly opposed Genie on an earnings call a few days later, arguing that Genie did not yet pose a threat to traditional gaming. “The genie is not a game engine.” He saidnoting that technology like it “certainly does not replace the creative process,” and that the tool to him looks more like “a procedurally generated interactive video at this point.” (Stock prices rose again in the following days.)

Google will almost certainly continue to improve its global Genie models and tools for generating interactive experiences. It’s unclear whether it wants to improve the experiences as games or whether it will instead focus on finding ways to help Genie on its ambitious path toward AGI.

However, other leaders in AI companies are already pushing for interactive AI experiences. Elon Musk of xAI Recently claimed “Real-time”, “high-quality” video games that are “personalised” will be available “next year”, and in December, he said that He said that Building an “AI gaming studio” is a “major project” for XAI. (Like with Many of Musk’s claimstake his predictions and timelines with a grain of salt.) Mark Zuckerberg of Meta, who is now pushing AI as New social media After the company Cut jobs into their own metaverseenvisions a future where people create a game from the prompt and share it with people in their feeds. Even Roblox, a gaming company, touts how creators can use AI world models and calls for in-game worlds to be created and changed in real time, which it calls “Dream in real time.”

But even on the more ambitious view that AI technology is practically capable of generating worlds as responsive and interesting to explore as a video game that runs locally on a home console, PC, or smartphone, there’s a lot more to making a video game than just creating a world. The best games have engaging gameplay, include interesting things to do, and feature original art, sound, writing, and characters. And it sometimes takes human developers years to make sure all the elements work together correctly.

Artificial intelligence technology is not yet ready to produce games, and anyone who thinks it is is fooling themselves. But AI-generated video was still bad, and it was still being used to produce it A bunch of bad Super Bowl adsso tech companies may still be putting a lot of effort into games made with generative AI. in An already unstable industryEven the idea that AI tools could compete with what humans can create could have huge ramifications in the future.

But the complexity of games is different from video AI, which has improved dramatically in a short period of time but has fewer variables to take into account. AI game-making tools will almost certainly improve, but the results may never close the gap to what humans can make.