How artificial intelligence changes video game industry – NBC Bay Area

In a congress hall in San Francisco, which used to be full with large stands of brands such as Sony Playstation and Epic Games, the 2025 Game Developers conference had a very different mood: lines and rows of tiny stands at which Indie developers showed their latest games, controllers and technological innovations and human quantities that gathered.

But the indie takeover of the South Hall of the Moscone Center was not the only change at this year's GDC, which has been held annually since 1988 (except while covid). Although the last few years have concentrated virtual reality and other technologies that give players a portal to a fully intangible meta -verses, the “IT” technology was clear, and it became artificial intelligence everywhere.

Nowhere was it so clearer than in one of these tiny stands in which a startup called Ovomind offered a demo of a creepy horror game. There was no VR headset, and no players swung through the air wildly -just a laptop, an Xbox controller and a small black bracelet that looked like a fitbit.

“We develop a technology that can measure emotions,” said Yann Frachi, CEO of Ovomind, and points to the small sensors on the bracelet. “We measure the skin temperature, micro -sweat end, heart rate behavior.”

All of these individual measurements are fed into a generative AI model in the cloud, explained Frachi, which uses this data to determine how the player feels, and reports it with adjectives such as “bored”, “excited” or “alarmed”. It also sends other data back, such as the player's heart rate and breathing destruction. These factors can in turn influence elements of the game such as endurance, the accuracy of shooting and the peripheral view of a character – or even when and where enemies appear.

“It will be a different experience every time,” said Frachi. “You can't get bored because it can adapt and change to inspire you … … this is the Game Changer.”

According to Frachi, the constantly growing amount of data for the training of Ovominds Ki began with a decade of academic research. The company sells a developer kit that includes the bracelet, but finally hopes to use the sensors in an ordinary smartwatch such as those of Apple or Samsung to expand the range of technology.

AI who can read their feelings may be a player for the players, but the game changes in other ways for the tens of thousands that participate in the Game Developers conference – including many with the hope of landing their next job in a game studio.

“Mass decisions, so to speak,” said Ryan Dunagan, a professor who taught game programs at the University of Denver. “There are much more talent than there are positions.”

In fact, the first participants arrived with high hopes of GDC, but the expectations. You can see that you play this quest in hard mode.

“My main interest is to get a job,” said Robert Breglio, whose all-access conference badge listened to his company as N/A. “Everyone listens, I would love a job. I am a really hard worker.”

It is a problem that we heard for the first time about a year ago when we went through the conference floor with the legendary game designer Louis Castle.

“Our industry has a certain crisis because the cost of developing content has become so high,” he said at that time.

But for a short moment, the industry found a magical healing potion.

“The game industry saw quite a elevator during Covid,” said Chris Hewish, Chief Strategy Officer at Game Commerce Company Xsolla.

Nothing to do, he said, people turned to games – and many of them still play today. But now the world is open outside again.

“Hyper growth slowed down a little,” he said. “We are still growing, but we don't see this high, double -digit growth.”

And that's the other place where AI comes into play. When we spoke last year, Castle expressed hope that Ai could help the studios to accelerate the process of building games and bring them onto the market. For example, when he designed the classic 90s games -Blade runner, the Optical Motion Capture only developed as a way to encourage characters. His team set up a full Motion Capture studio in their offices with specialized Vicon devices.

“Every single person moves a little differently,” said David Edwards David Edwards Vicon VFX. “And you can also communicate a lot of how you move.”

Motion Capture was created for the de -facto -art, expressive animations for games, since the use of the performance of an actor was dramatically faster than any animation from scratch. Now Vicon wants to accelerate things even more, which it calls “Markerlos” motion capture. Instead of commiting the actors to wear special black suits that are littered with dozens of gray plastic balls or “markers” on every joint and fingertips, the new system uses a series of video cameras and a AI model to extract animation data from the movements of the people in their regular clothes.

“You spend less time to worry that everything that is perfectly furnished,” said Edwards, “and more time to worry, 'Is that the performance I need? Is that the animation that my game needs?” “

During his first public demo of the technology at the GDC, Vicon said that his new AI-based movement recording should not replace the “old path” with suits and markings what is more precise when collecting detailed information up to the most subtle finger movements. Instead, the new technology should accelerate the process of creating prototypes and coarse designs – a goal that is supported by the International Game Developers Association.

“AI has great uses such as fast iteration, fast prototyping when you see if it is fun to play things,” said Jakin Vela, Executive by IGDA. “It is when Studio heads or managers believe that AI should be substitute Workers, The Is the concern. “

Vela said that there are also concerns about AI training data, the massive sentences of real examples that are fed into AI models so that they can learn to generate images, sounds and text. Vicon and Ovomind both say that they collect all of these examples in their own house, so there is no question about who belongs. But if the source of the data is not so black and white, as for some AI models for online image generation-this can be a problem for artists, including the actors.

“Language actors also have some concerns,” added Vela. “To ensure that the AI ​​does not tear off its voices or runs out its similarity.”

And then there is a challenge to ensure that generative AI models are suitable for children-and ensure that your training data only includes clean, healthy examples, so you can only learn how to generate G-rating content. This has a priority for Roblox, a game in which a great increase in popularity recorded during the Covid pandemic. In Roblox, players can create their own interactive 3D experiences – essentially – and become a game designer themselves. Roblox now has a new open source AI model called Cube that is ready to facilitate this process.

“The creation in 3D is very, very difficult,” said Kiran Bhat, senior engineering director of Roblox. “So … it is a way for our users, especially the younger users, to be able to experience the magic of creation.”

The new AI model can generate 3D objects from a text description, and Roblox is actively working to design it at once. Bhat sees a future in which creator can try dozens of ideas for a new Roblox experience and choose those who like it best – everything by tapping words and phrases and magically appearing in 3D – and soon what Roblox 4D calls.

“The fourth dimension of functionality is in the Roblox context,” said Bhat. “So if you think of a door, you have created the geometry of the door, it is 3D – but the ability to open and close the door is 4d.”

When a generation of players grows up to build worlds with AI, Bhat hopes that some of them will be the next generation of game developers. And Hewish believes it will happen.

“Every new technology comes with the fear that people will be sold and replaced to the printing press,” said Hewish. “I think we will have opportunities for creators who are committed to AI – who can actually work with AI as a partner.”

Because as Vela emphasizes, there is no return of the clock to AI. The fourth industrial revolution was even called.

“It's here,” said Vela. “It doesn't go anywhere. We have to learn how to work together.”