Unity is joining other companies in providing users with generative AI tools, but ensuring that those tools (unlike some) are built on a foundation that is not based on theft. I have been careful to check. Muse, a new suite of AI-powered tools, starts with texture and sprite generation and gradually moves into animation and coding as it matures.

The company announced these features at the Unite conference in San Francisco, along with Unity 6, the next big version of its cloud-based platform and its engine. After a turbulent few months that saw major product plans completely scrapped and a CEO ousted, you’re probably looking to get back to business as usual if possible.

Unity has traditionally positioned itself as a champion for small developers who lack the resources to adopt broader development platforms like rival Unreal. Therefore, the use of AI tools can be considered a useful addition for a developer who cannot afford to spend days creating, for example, 32 slightly different wooden wall textures in high resolution. can.

There are many tools out there to help you create and modify assets like this, but it’s often desirable to be able to say “make something more like it” without leaving your main development environment. The simpler your workflow, the more you can do without worrying about details like formatting or siled resources.

AI assets are also often used in prototyping, where things like artifacts and slightly wonky quality (which these days are common regardless of model) don’t really matter. However, illustrating your gameplay concept with original, well-made art rather than stock sprites or free sample 3D models can make the difference in communicating your vision to publishers and investors.

Another new AI feature, Sentis, is a little harder to understand. “It enables developers to bring complex AI data models into the Unity runtime to create new gameplay experiences and features,” Unity’s press release states. So it’s kind of his BYO model, with some features built in, and it’s currently in open beta.

AI for animation and movement is in development and will be added next year. These highly specialized scripting and design processes can greatly benefit from generative first drafts or multiplicative helpers.

Image credits: unity

The Unity team emphasized that a big part of this release is to ensure that these tools are not overshadowed by future IP infringement lawsuits. Image generators like Stable Diffusion are fun to play with, but they’re built using assets from artists who never agreed to have their work taken and regurgitated.

“To provide usable output that is safe, responsible, and respectful of the copyrights of other creators, we challenged ourselves to innovate the training techniques for the AI ​​models that power Muse’s sprite and texture generation.” says a blog post on responsible AI. Techniques associated with presentations.

The company said it used a completely custom model trained with images owned or licensed by Unity. However, they essentially used stable diffusion to generate a larger synthetic dataset from the small, carefully selected datasets they had assembled.

Image credits: unity

For example, this wood wall texture may be rendered with several variations and color types using a stable diffusion model, but no new content will be added. At least that’s how it’s described to work. But as a result, new datasets are not only based on responsibly sourced data, but also one step removed from it, making it less likely that a particular artist or style will be duplicated.

Although this approach is more secure, Unity admitted that the quality of the initial models it was providing was reduced. However, as mentioned above, the actual quality of the generated assets is not necessarily important.

Unity Muse costs $30 per month as a standalone product. We’re sure you’ll soon hear from the community about whether this product is worth its price.