So, my question is fairly simple, how do I teach the AI to be better? It seems like it can do things that are totally subjective fairly well, but if it is a real actual place or thing, it completely breaks down and generates images that laughable, and could never actually be used. I see amazing AI created artworks that are being touted all over the internet, things that look nearly photo realistic. And yet Adobe Firefly is THAT far behind in quality? You can ask for some very specific characteristics, and it will do nearly the opposite.
Even if you do manage to get an image that has potential, you don't have the ability to choose one of the images and make modifications, it will automatically generate a completely separate image. Wouldn't it be so much more intuitive to be able to generate an image, and then for example, ask it to "Make the cowboy hat smaller." And have it do exactly that?
But all of that comes after being able to teach it what a cowboy is, because it clearly needs some more education. So, How do I teach the AI to be better?
1
Reply
AdChoices