What do you get from a single-word prompt?
A one-word prompt hands almost every decision to the model, which makes it a surprisingly good way to see its defaults and biases.
Which single word has given you the most surprising result, and on which model?