You can definitely input the wrong words in your prompt, swap the positive and negative prompts, forget to select the correct controlnet pre-processor. These errors do turn out interesting results, I wish I had some to show you. How much experience do you have working with this stuff?
Trust me there are a lot of accidents with AI generators
But they’re never interesting. It’s never due to some human factor, but always a “huh, I guess the stochastic model doesn’t work properly.”
You can definitely input the wrong words in your prompt, swap the positive and negative prompts, forget to select the correct controlnet pre-processor. These errors do turn out interesting results, I wish I had some to show you. How much experience do you have working with this stuff?