One was created and hand drawn onto some medium by a talented human, the other was never seen by a human eye until it was belched completed out of a computer program.

One was created line by line with purpose and intent, using materials, textures, colors, tones and an understanding of composition. The other was cobbled together by scraping the Internet for work by artists and then analyzing millions of images to mimic them.

Machine learning cannot create anything new, only approximate an execution of what has already been done.

Having an idea for an image and not being able to execute it but only tell someone or something else to do it, is the other side of that coin. If you can type a great description, you may want to call yourself an author, and then hire an illustrator.