How to Ask AI the Right Question: Guiding an Over-Eager Co-Worker

Imagine a colleague who is brilliant but exhausting. You ask for a short summary, and five minutes later they return with ten pages of beautifully formatted analysis, three charts, and a conclusion you never asked for. They mean well, they want to help, but they do not always know when to stop. That is what prompting an AI assistant can feel like. Artificial intelligence thrives on enthusiasm. It will never say “I don’t know,” even when it probably should. It fills the silence with plausible guesses, adds detail to look confident, and may go far beyond your original request. The key to working well with such a co-worker is not to rein them in with code or complex rules, but to learn how to guide them. The simplest way to do that is to speak clearly. A good prompt is not magic, it is just a clear instruction. Instead of asking “Tell me about climate change,” try “Summarise in two short paragraphs how climate change affects coastal cities, using plain language for a non-expert audience.” The difference is the same as asking a colleague to “find everything about this topic” versus “find me three recent examples that illustrate this point.” Adding guardrails sounds technical, but it is really just about boundaries. You can tell the AI what not to do: “Don’t give me a list,” “Avoid technical jargon,” or “Keep it under 200 words.” These limits are like telling an over-enthusiastic assistant to stick to the brief. You are not restricting creativity; you are protecting focus. If the answer you get feels off, think of it as a conversation, not a verdict. You can refine your question, narrow the scope, or ask for a different tone. Most of the time, the problem is not the AI’s ability but the vagueness of the instruction. A vague question leads to a vague answer, just as it would with a human colleague. For people who feel unsure about “how to use AI,” this is the hidden truth: you already know most of it. If you can ask a colleague for something clearly and politely, you can prompt an AI just as well. The rest comes naturally with practice. Treat AI like a bright, energetic team member who sometimes needs direction. When you give it clarity, it performs well. When you leave it guessing, it improvises – and that is where trouble begins. Learning to prompt is not a new skill, it is just good communication, repurposed for the digital age. Some common examples: 1. Research summary (scope and tone guardrail) “Summarise the main findings of recent research on renewable energy in Africa in two short paragraphs, written for a business audience. Avoid academic jargon and focus on economic opportunities.” Why it works: It limits the length, defines the audience, and prevents overly technical language, guiding the AI to stay relevant and readable. 2. Drafting communication (tone and content guardrail) “Write an email to customers announcing a small price increase, keeping the tone friendly and transparent. Do not use marketing clichés or exaggerated language. End with a note of appreciation.” Why it works: The instruction defines tone, emotional range, and what not to include, reducing the chance of a robotic or salesy message. 3. Creative support (format and style guardrail) “Give me three creative taglines for a public library campaign that promotes reading. Keep each tagline under ten words and avoid rhymes or puns.” Why it works: The format is fixed (three short lines), and stylistic boundaries are clear (no rhymes or puns). This keeps the AI concise and on-brand. 4. Decision aid (clarity and neutrality guardrail) “List the main pros and cons of remote work policies for small companies, staying neutral in tone. Do not include personal opinions or generic motivational advice.” Why it works: The model is instructed to stay balanced and factual; an effective way to prevent biased or opinion-heavy results. 5. Technical explanation (audience and complexity guardrail) “Explain how blockchain works to a 14-year-old using simple analogies, without mentioning cryptocurrency or finance.” Why it works: It defines both audience and boundaries; the AI knows to simplify the explanation and avoid its usual examples, forcing creativity within limits. This article was created by people. We have used artificial intelligence (AI) to help articulate our message and refine the text. AI was employed as a tool to assist with structuring, identifying grammatical and spelling errors, and improving readability. The final document has been carefully reviewed and approved by our team.