Artificial intelligence does not have enough creativity to create something new when developing video games, said John Romero, co-founder of ID Software and co-author of Doom, in an interview with Ars Technica.
He expressed skepticism about the impact of generative AI on the gaming industry.
"It is important for us to be original. Generative neural networks depend on databases that already exist, and we are trying to create a completely new thing," Romero said.Game studios use neural networks to solve various tasks such as character improvement or chat moderation. The programmer believes that an AI that is still developing is not capable of designing and helping in development at the human level.
"Creating completely new things is not yet what generative AI does. I know they're working on it. But when you create an original game, you use your design flair, work with people and come up with things that didn't exist before. Generative AI limits you if you try to use it that way," Romero said.Among the advantages of generative AI, the developer named the field of information search, noting that the neural network is a "super—Google" for getting answers to questions. He attributed the model training system to the disadvantages.
"When AI is used to create images or creative concepts, it is problematic because it does not rely on your own data," Romero added, stressing the importance of ethical use of neural networks and compensation to sources that were used for training.DeepMind previously released a Genie model that creates games based on a text query.
Recall that in March, CognitionAI startup introduced an AI that is able to generate complex code for websites or games based on prompta.
In the same month, Google's AI was taught how to play video games.