There has been a lot of talk saying that AI Art isn't real art. Some people feel this way because you can use a generative AI program type a few words and get an image back very quickly. While you certainly can get quick output off of a few words its most likely not going to be anywhere near good enough to use. I am going to breakdown the process behind getting a really good image in this post, because most people just don't understand the work that goes on behind real Ai Art.
The reason I love generative Ai is you are only limited by your imagination. If you can think it up, and understand how to take your ideas put them into words, you can create some amazing things. While you can just type cat in a bed and get an output like this.
Its not going to be the image you had in your head, because it wasn't descriptive at all. However by continuing to refine and elaborate your input the output can become something truly amazing. Learning how to create a prompt that can help make the image in your mind is something that takes 100s if not 1000s of hours to become good at.
Lets take our cat in the bed idea but really elaborate on it and create some art. Here is the prompt I used.
"Create an image of a beautiful bedroom that exudes tranquility and elegance, with large windows offering a stunning view of a beach at dusk. The room is bathed in the soft, warm hues of the setting sun, casting a golden glow across the interior. The decor is tasteful and serene, with a comfortable bed at the center, adorned with plush bedding and fluffy pillows. On the bed, a graceful Siamese cat lounges contentedly, its blue eyes gazing out towards the horizon where the sky meets the sea. The room opens up to a breathtaking panorama of the beach, where the gentle waves whisper against the shore and the sky is painted in shades of pink, orange, and purple. This scene captures a moment of perfect peace, blending the beauty of nature with the comfort of a well-appointed bedroom."
For the first time in human history you are able to paint an actual picture others can see with their eyes but just using words. For someone like myself who has always been extremely creative but had no physical skill to paint, this is absolutely amazing. I am able to create the image in my head and share it with the world!
However even after a well written prompt there is almost always some part of the the image that is not final product ready. In order to make something is worthy of being printed there is a lot of work that needs to be done after getting the AI generated image.
Lets take a closer look at the Dall-E image from the start of the post.
When we look closely we can see the easel that was in the original output has a 4th leg that doesn't even go down to the floor!
Wall-E also has some really odd things going on. Take a look at the original output for his head. What are those weird antennas?
Ai images are filled with these out of place things. As an artist its my job to go in and take the original image and spend a ton of time working to remove all the odd artifacts and misplaced parts of the image in order to make a final piece that lines up with my original vision and is worthy of being printed.
This includes many hours working on an image in photoshop, removing things, playing the colors to make them pop, upscaling the images so the details are sharp and vibrant in a larger print. The end result after hours of creation is an PEICE OF ART and something worthy of being hung on the wall!