When using image generating AIs like Midjourney or Dall-E the protocol is to think of something you want to render and then instruct the AI to make it using a prompt. For example: "a rabbit eating carrots on Mount Everest". And it's an art to engineer these prompts in a way the AI will understand and give you good results.
In this process, you're generally trying to reduce unexpected results. But what if you wanted to increase unexpected results? What if you wanted to create noise and let the AI show you what it sees?
This tool is a helper for creating prompts meant to generate unexpected results from image generating AIs.
Save these and upload them somewhere, direct linking is unreliable.