This seems like a lack of experience. The more I work with LLMs, the better I get at predicting what they’ll get wrong. I then shape my prompts to avoid the mis

This seems like a lack of experience. The more I work with LLMs, the better I get at predicting what they’ll get wrong. I then shape my prompts to avoid the mistakes.