I agree. Also it's worth to consider that these machine learning technologies can be applied to almost ANY digital data. The images and text were first because that was just easy and quick to grab from the internet and process. That's why digital painters and writers deal with this now, it will spread to different digital art spheres - video, music, 3D (sight) etc. Programmers are already dealing with people using GPT for generating code, it will be everywhere. In HIVE it's probably best if people could post some work in progress documentation together with their work as a proof, but ultimately that also can be faked, so before someone develops some kind of working AI detector it will be hard to tell what is what.