This idea is inspired by painting. My experience is that when you paint. You make alterations based on feelings. If the drawing doesn’t feel quite right you make an alteration until everything feels right. Thinking about feelings. Everything we do generates a feeling within us. I don’t think that it’s just for show. When a person designs a product he looks at it and feels. If he then makes a prediction on what the changes should be to amplify that good feeling. Is he not optimizing that product with a feeling as a base? My idea is then that we might be able to capture or computer recognize that feeling and others with machine learning. Then we could have the computer make optimizations based on feelings. That is. The computer makes alterations to the design to maximize the specific “feeling” we want.