I got the chance to be part of the beta testers of the AI tool Midjourney. I could see it’s gaining traction and as a creative and artist myself I was immediately interested in trying out this new tech. Its use is very simple. You just joint a discord group and input a command to a bot that translates your words or sentences into art. Then you have 25 requests to the server for free. What I was curious to find out is, if my imagination, expressed as a dozen of words, can be translated into picture by this tool. The results I got were quite interesting and even surprised me and gave me new some new ideas.
Impressionist futuristic architecture
The first input I tested was related to one of my visions of the future of architecture. I really want to see levitating houses sooner than later, that’s why I checked what Midjourney thinks about it. I wanted a lot of vegetation as well as flying things around and in the end I just wrote the name Monet so that it can take some style from the famous impressionist.
/imagine white futuristic flying houses over a green and yellow forest on a new habitable planet with strange plants and flying cars and people around monet
This was the result from this input which is a small resolution file:
From it I liked the second image on the first row and generated a bigger version by the upscale button. Later I generated the other two images in the big resolution:
Then I iterated on the third one with flying yellow houses on green to see what variations we can get then and decided I like the first two for upscale.
Then I was set to try another planet with a different mood. Some good contrast with the lush vegetation of my blockbuster story with the flying houses.
/imagine beautiful young woman with a futuristic spacesuite on mars, show habitat in backgorund Rembrandt van Rijn
That would have been amazing command if I didn’t make two typos on spacesuit and background. But still the software guessed kind of pretty good for my stone age English skills.
I felt I need to iterate quite a bit with this one and I guess it was because of my typos. Also I was not seeing any habitat anywhere.
I started to miss the green fresh vibes so at the end I tried to see how a future city might look like. I use the following command:
/imagine detailed future city with very tall and thin buildings connected by bridges and highways with a forest between the buildings and on top of the buildings, and a lot flying vehicles and drones, show many people in foreground in a lawn with floweres, half sunny half stormy weather
In conclusion, I really enjoyed this one hour of generating stuff. I can definitely see this as one more creative tool to explore ideas but I feel there should be some more ways to give parameters specific for each object you want in the image. I maybe missed that but can try to find out more next time.
You can also get access to the beta here: discord.gg/midjourney