Here's a little demo that I whipped up for last week's U of Digital AI accelerator training on how to create a repeatable skill as a Google Gemini Gem - in this case, to create a one-panel cartoon in the style of the New Yorker or The Far Side.
It’s kind of mind-boggling how well Gemini was able to do a lot of the prompting and ideation itself. It still took about 2 hours and a lot of iteration to nail the prompt and context down to make sure the text and model were coming back with what I wanted.