become tedious and challenging. We are also often challenged by our use of a limited vocabulary. We can say to a friend, “Take a portrait photo of me,” and they know what we mean. We don’t have to describe the camera, lens, setting, lighting, etc.
With AI-based text-to-art generators, like VQGAN+CLIP or Disco Diffusion, the AI would have no clue what “Take a portrait photo of me” even means. It would try to interpret the statement as best it can, but without additional modifiers that help describe further what you mean, you are likely NOT going to achieve an acceptable result.