Question

Select all the statements about prompting AI that are true.
Research shows that using Star Trek language to speak to AI will return the best results for
math calculations.
Research shows that AI will return the best results for scientific questions if the request is
blunt and stops just short of being rude.
Research suggests that you will get better results if you follow politeness norms when interacting with AI.
Research suggests that AI performs better when given certain prompts, so you should experiment to achieve the best results

Answer 1

True: only the fourth statement.

- False: "using Star Trek language..." — no evidence that using fictional/novel dialects improves math performance.
- False: "best results if the request is blunt..." — bluntness itself isn’t shown to increase scientific accuracy; clarity, specificity, and correct context do.
- False (generally): "you will get better results if you follow politeness norms..." — politeness can affect human perception of responses and sometimes how models imitate social behavior, but it’s not a reliable way to improve factual or technical performance; clarity and good prompt structure matter more.
- True: "AI performs better when given certain prompts, so you should experiment..." — extensive research and practice in prompt engineering (instruction tuning, few-shot, chain-of-thought, etc.) shows that wording, examples, constraints, and format strongly affect model outputs.

If you want, I can give brief examples of good prompt patterns for math or scientific questions.