Prompt Engineering Is Just Common Sense With Structure

Most developers who get bad results from LLMs blame the model. The model is usually not the problem. The problem is that the instructions were vague. And vague instructions get plausible-sounding answ