[rdrd]

I’ve enjoyed using it to create random test data in csv format, something I would usually do mashing together a few APIs and a Python script. Just today I asked for a large number of rows with several different column types, including first name, last name and random date of births in ISO 8601 format. After generating very US-centric names I asked for UK based names, which it again modified and generated a very sane dataset for, but I wanted it more culturally diverse, so a final re-prompt of “can you make the names more culturally diverse” and voila, it was done. What might have taken me an hour to produce in the past was done in 2 mins.

[blakeburch]

I've been using it for the same thing. Super helpful for building sample datasets for a class I'm putting together.