It has been amazing to watch the feedback loop with tools like Stable Diffusion where you see two types of strategy for solving problems of missing functionality. The first strategy is coming up with some long winded manual process involving disparate tools and several steps. The other is simply train a model to do what it is you want automatically. It has fast taught me that there is no competing with this new paradigm, and those capable of being able to solve their problems by building models will totally leapfrog those that have to do things the old way.

In all seriousness I've found that asking them to generate their own prompts can be very useful. It's a bit like a perl script that outputs a perl script.

You just need an LLM to generate random prompts and a sweatshop in Manila that filters out the bad ones.