You have a guiding hand (choreographed) in the VR setup and you put your hand where it is and ‘trace’ with them. Of course you can’t see what you are drawing, and in reality your hand is of course moving but has a pen/pencil on the end and is creating a visual on a papered wall for others to see. paper roll things and mount it to a wall so then you also have a “scroll” that’s a record of what the virtual world told us. The AI strictly ‘judges’ (is part of the evaluation formula) if you do or do not do what you are told (aka trace accuracy), but judged by masses for what you create visually.

Let the audience respond and use their responses for the machine learning — so the AI is using human data [data set of responses] to generate prompts instead of humans creating prompts for AI. Repurpose my ‘swipe right/left’ rating app to log the audience responses to the crafted work.

Meanwhile, the choreographed art itself would be randomly selected through an algorithm — like a fractal formula — and one of the heuristic variable parameters used in that formula would be calculated through machine learning (based on the ‘votes’ for physical art from associated art prompt).

cave paintings, archaeology, trace history, audience, personal, performance, judgement, volition, etc.

Is ml ai better as a tool, a toy, or a reinforcement guardian? Do we want or need any of these? Will we be content as puppets on strings so long as the choreographer is a collective group-think presented in an ai algorithm as opposed to any identifiable (creditable) human or group of humans? Is this ego, socialization, insecurity, naivete?