will have images that they can pull from to reference when they need, for example, "a woman with green eyes", based on what it thinks a woman and a person with green eyes looks like. An artbot won't create a "new" image of a person. It will take what it is asked to generate, then