Stable Attribution is a tool that allows people to identify the human creators behind AI-generated images. It works by uploading an image generated by AI, and then using noisy vectors and the original numerical representation of the photo and its accompanying caption to identify the artist who created the source image. The tool was created by Anton Troynikov, the co-founder of Chroma, who wanted to find a technical solution to the problem of ensuring that technologists and creatives were not antagonists to one another. The goal of Stable Attribution is to demonstrate that this problem is not technically infeasible to tackle. By being able to do attribution, users can then proportionately compensate the contributors to their AI models. The tool has been positively received by some, with one user on Hacker News noting that the first images suggested by Stable Attribution looked a lot more like the AI image in terms of pose and everything.
