A simple follow-up question can expose assumptions and gaps in AI-generated responses, according to a writer who tested the technique. The method was described as a way to improve ChatGPT’s usefulness without needing to craft long initial prompts.
log in to read full article