Adobe has integrated AI into a new tool in the Photoshop beta called generative fill. It has several things it can do that are pretty cool. For one thing, if there is something in the picture that you want to get rid of, you can do a loose selection of it and press the generative fill button, and poof! It’s gone; And not only is it gone, you can’t tell it was ever there and it creates new information for the area it removed. It is pretty much undetectable. Like most AI that’s coming out now, it can also use a text input to describe what you want added to the picture. For instance, you can select an area on the composition, press the button and in a text box type something like “blue VW beetle” and it will insert one. In addition, it will match the lighting and color shading of the existing composite. It will start you with 3 choices and if those don’t suit you, push it again and it will give you 3 more.
I sat down and started to play with this tonight and decided I would take a little landscape picture I have and make it bigger by having the AI build the new areas. This worked better than I could have imagined. I made it 3 times its original size with all new content added and you couldn’t tell. So I decided to take this a step further. I created a blank canvas and told it I wanted a “beach at sunset facing out at the ocean”. It made my whole background. I then told it “Girl walking on the beach looking at the ocean deep in thought”. It did it and you couldn’t tell any of the three it brought back weren’t live models on a beach. I then told it “Small Sail Boats” and it added them in the water! I then fiddled around with textures, filters and other editing and came up with what you see above. It looks kind of like a painting now, but I assure you it looked like a photograph before I started editing. In fact, here is the original composition that the AI made from the prompts above.
See?! It looks like I shot this on the beach at sunset. I was quite excited then the more I thought about it, I began to wonder if this art practice I have been working so hard at is about to be replaced and anyone who can write a good description of the composition can create beautiful art. But are they? Or did they just provide a prompt to a computer and the computer created the art? There is no doubt this is a beautiful art that is being created, but is the artist the talent or the programmer? I guess since I am both, I should be happy, but I don’t think I am. I wonder if this is what the painters of the world felt like when Photoshop came out?