Extracting training data from ChatGPT

This is pretty interesting.

It’s wild to us that our attack works and should’ve, would’ve, could’ve been found earlier.

Most attacks in a nutshell lol