I haven’t seen this done, yet, but a brilliant introduction to the concept of generative AI is the well-known improv game One Word At A Time.

The idea of the game is simple, a team tells a story one word at a time. That’s it.

Which is roughly what Large Language Models are doing; just on a grander scale and with some secret sanity simulations built in.

In a group setting, try it. Give a team a prompt and see how easy it is to get close to a coherent answer.