Though AI art is in it’s infancy, it can still pump out some pretty darn awe-inspiring images. I discovered AI art a couple of weeks ago and I must admit that I’ve become quite addicted to “fishing”–typing in a descriptive prompt, to see what the digital Degas whips up. It usually takes quite a few tries to get a coherent image and there are some things that AI is currently quite challenged at. It’s very good for natural landscapes but it has a hard time doing coherent human figures and very specific requests (ie – “lego LeBron James slam dunking the moon”). Once it is coupled with more precise editing tools it will be an incredible resource for those who need to create images on the fly.
My favorite AI art creator so far is Midjourney. Some of the visions it’s able to create are really uncanny. Take a look at the gallery of images below. I made all of these and a lot more in a couple of hours. Though many of these pictures took quite a bit of fishing, they indicate what the technology is currently capable of. Maybe it can eventually be a tool to help us design new buildings and even indicate possibilities we haven’t thought of before–a new source of inspiration. Either way, I think it’s wise of us to embrace new non-destructive technologies because they’re going to radically revolutionize the way we do things, whether we like it or not. Midjourney gives you four ideas per prompt and then allows you to make variations of any images you like. The two images below are variations from the same base image.
On the Midjourney discord server, it can get quite busy and hard to keep all of your images organized so I suggest starting a private server and to connect the Midjourney bot to it and direct message prompts to it there–that way all of your art will be in one place. Thanks for reading!
Some More Examples:
