I guess you're technically correct, but the task you're describing isn't generating an image from a prompt. It would be to maintain context across distinct-but-

I guess you're technically correct, but the task you're describing isn't generating an image from a prompt. It would be to maintain context across distinct-but-related statements based on an internalized model of reality. That's like discounting the advent of the calculator because you still need an accountant.