Since I know that most people don’t go through the technical documentation when new models come out (or ever), I want to point out sections that I think will help with people’s day-to-day AI tool usage.
This is from OpenAI’s GPT-5.2 Prompting Guide. If you’ve already gone through it then this will look familiar - unless you’ve already stopped reading.
According to the guide, GPT-5.2 generally has a lower verbosity than previous models. They still recommend prompting it specifically to “steer” the output.
Here is their example to control verbosity and output shape -
<output_verbosity_spec>
- Default: 3–6 sentences or ≤5 bullets for typical answers.
- For simple “yes/no + short explanation” questions: ≤2 sentences.
- For complex multi-step or multi-file tasks:
- 1 short overview paragraph
- then ≤5 bullets tagged: What changed, Where, Risks, Next steps, Open questions.
- Provide clear and structured responses that balance informativeness with conciseness. Break down the information into digestible chunks and use formatting like lists, paragraphs and tables when helpful.
- Avoid long narrative paragraphs; prefer compact bullets and short sections.
- Do not rephrase the user’s request unless it changes semantics.
</output_verbosity_spec>
Take a look at AI Quick Tips #261 if you don’t know what those notations mean.
.png?width=600&height=496&name=Next%20Level%20AI%20Workshop%20Blog%20CTA%20-%201%20(1).png)
You’ve Read Enough About AI. Time to Do Something With It.
Hands-on workshops for people who want results, not theory.
Comments