Recently I discovered an AI bot named Midjourney, a bot hosted on a server in an app called Discord, which is a chatroom and community app. It is an AI that produces pictures for members of the server, who can get twenty five free images or can pay for membership which is a monthly subscription that has multiple tiers. After doing a bit of a play around with the trial version of the AI, I decided to subscribe to basic tier to allow me to play around a bit more and discover what different kinds of things Midjourney could do.
The AI works by taking prompt words from the user and producing a two by two grid of images based on the details it has received from the user in 60 seconds. There seems to be very few limits on what the AI can do and how it quickly each prompt is processed.
Users then have the option to either upscale any or all of the images produced by the AI. Some of the results that I personally have received from the AI after giving it different prompts have been seriously impressive, I’ll include examples below with the word prompts that I gave to produce the images
I started off thinking it would be interesting to experiment and see what sorts of interesting images it could produce, before moving on to images that were photorealistic or close to looking as real as possible. I thought it would be interesting to see an artificial intelligences interpretation of different real world artifacts and how to render them to look a certain way. I felt like if I could get something to look like a photo it might make for some interesting subjects.
With a new month coming soon with a credit reset for paying members I look forward to more experimenting and seeing what the AI can do and what new features the hardworking staff are working on in the background and I look forward to seeing what they are.