Generative AI Video Game Art Startup Scenario Raises $6M
Generative AI art for video games startup Scenario has raised $6 million in a seed funding round led by Play Ventures and opened early access to its GenAI Engine. Scenario enables artists and developers to set up and train their own image generators, producing thousands of assets in their own style.
Scenario’s GenAI helps developers design custom art generators trained on their own content. Developers only need a few example images of the kinds of characters, props, or environments they want for the AI to learn the style. It can then create new assets from text prompts. The images are usable for games as well as related material like advertising and marketing material. Those interested can access GenAI online, through a mobile app, or embed Scenario’s API into their own system for $20 a month.
“Digital artists must remain in control of how their art is used in AI products, working alongside AI and not in competition with it,” Scenario CEO Emmanuel de Maistre said. “Generative AI will be as transformational for game development as Photoshop has been for digital photography, but it cannot get there without the same commitment to stylistic consistency, ease of use, and the rights of the individual artist.”
Applying synthetic image generation to video games is on the rise. Game developer Latitude created AI Dungeon to use advanced language models to produce the text of the game and interact with players through its characters. The game currently uses the Jurassic-1 model created by AI21 Labs. Meanwhile, Google has experimented with using its Imagen text-to-image AI tool to augment classic 1970s text adventure Zork with graphics. One notable aspect of Scenario’s approach is that it insists clients train the AI using their own data. That way, they might avoid the issue of ownership rights for training images that led artists to file a lawsuit against Stability AI, Midjourney, and DeviantArt over their Stable Diffusion-based generative AI.