Rather than only telling an AI what you want, you also explicitly tell it what you don't want—exclude certain tones, avoid specific topics, skip particular formatting—which often produces better results than positive instructions alone. This mirrors how humans clarify requests: not just 'write something funny' but 'write something funny, but not mean-spirited.'
Negative prompting is the practice of explicitly stating what you do not want in an AI response, such as excluding certain formats, topics, tones, assumptions, or types of content from the output.
Adding clear negative instructions to your prompts eliminates the most common sources of unhelpful AI output and dramatically reduces the number of follow-up correction messages you need to send.
Peri can explain this concept, give practical examples, help you decide whether it applies to your situation, or recommend a journey if appropriate.
Explore related journeys or tell Peri what you're working through.