One idea per prompt
A common mistake when people get comfortable with a model is to stuff more and more into a single prompt: the task, the constraints, the edge cases, the output format, three examples, a style guide, and a reminder not to hallucinate. The prompt gets longer; the output gets worse.
The fix is almost always the same: ask for less in one turn.
Why long prompts fail
Models attend to their whole context, but not evenly. Long prompts bury the actual ask under scaffolding, and the model quietly optimizes for whichever part of the instruction was most salient — often not the part you cared about.
A tighter prompt gives the model a single target. You can add the next constraint on the next turn, after you’ve seen that the first one stuck.
The pattern that works
- State the task in one sentence.
- Let the model respond.
- Correct or extend: “good, now also handle X,” “that’s wrong because Y, try again.”
This turns prompting into a conversation you can steer, instead of a specification you have to get right on the first try. It’s also much faster in wall-clock time, because the first response comes back sooner and you catch misunderstandings earlier.
When to break the rule
Batch work — “process these 200 records the same way” — does want the full spec up front, because you are not going to iterate. But for any task where you are the reader of the output, start small.