Topic hub
Prompt Engineering
Prompt engineering is the skill of asking an AI model for what you actually want — clearly enough that the first answer is usable and the second is right. Everything on this site about it lives here.
The name makes it sound technical. It is not. A language model continues text in the most plausible way it can, so a vague request is consistent with thousands of possible answers and you receive the most generic one. A specific request — who the reader is, what you need, the facts the model could not know, and the shape you want back — narrows that range to the answers you would actually accept.
That is the whole mechanism, and it explains why every genuine technique works: giving an example narrows the range, naming the audience narrows it, saying what to avoid removes a region of likely answers. Once you see it that way, you stop collecting magic phrases and start reasoning about what information the model is missing.
The other habit worth building is treating the first response as a draft. Say precisely what is wrong with it and ask again. Almost every good result comes from two or three rounds of that, and people who find these tools disappointing are usually judging them on the first attempt.
None of this depends on which assistant you use. The course, the guides, and the templates below apply equally to any AI chat tool, and they will keep applying to the tools that replace the current ones.
Start with the course
Structured techniques for getting reliable, high-quality output from AI tools on real professional tasks — not just clever one-liners.
4 lessons · about 8 min of reading · free, no sign-up
Read the guides
What Is Prompt Engineering? A Beginner's Guide
The difference between a vague AI prompt and a great one usually isn't cleverness — it's structure. Here's the structure.
A Beginner's Checklist for Better AI Prompts
A short, practical checklist to run through before you hit send — the fastest way to get noticeably better AI output.
10 ChatGPT Tips That Improve Your Results
Not tricks or magic words — specific, repeatable habits that make a real difference to the quality of what you get back.
Writing With AI Without Sounding Like AI
AI-generated text has recognizable tells. Here's what they are and how to work with AI so your writing still sounds like you.
The prompt library, by task
Every template explains why its wording works, what good output looks like, and the mistakes that produce poor results — so the reasoning transfers to prompts you write yourself.
Marketing
Coding
Learning
- Break Down a Complex Topic Into a Study PlanTurn an intimidating subject into a sequence of learnable steps.
- Explain a Concept at Three LevelsUnderstand a difficult topic by hearing it explained simply, then with more depth.
- Make Flashcards From Your NotesTurn class or reading notes into question-answer flashcards for active recall practice.
Productivity
- Turn a Brain Dump Into a Task ListConvert an unstructured list of to-dos into something actionable.
- Turn a Done-List Into a Status UpdateConvert your scattered list of completed work into a clear weekly update for your manager or team.
- Make the AI Plan Before It ActsGet the full plan, and the risks in it, before any action is taken on your behalf.
- Delegate a Task to an AI Agent SafelyA briefing that sets the goal, the boundaries, and the stopping conditions before work starts.
Business
- Draft a Polite Follow-Up EmailGet a follow-up message that's firm but not pushy.
- Draft a Reply to a Frustrated CustomerRespond to a complaint with genuine acknowledgment and a concrete next step — not corporate filler.
- Review an AI Agent's Work Like an AuditorCheck what an agent actually did rather than trusting its own summary of it.
Career
- Tailor a Resume Bullet to a Job DescriptionRewrite an existing accomplishment to speak directly to a specific job posting.
- Practice Interview Answers With FeedbackRehearse answers to likely interview questions and get specific, critical feedback.
- Turn a Job Description Into Prep NotesExtract what an employer is really looking for and prepare targeted talking points.
Content Creation
Key terms
The vocabulary you will meet in prompting advice, defined in plain English with an honest note on limits.
Assistants these techniques apply to
Independent write-ups — including where each one falls short — of the general assistants most people practise on.
Questions about prompt engineering
- Is prompt engineering a real skill, or a passing trick?
- The underlying skill is writing a clear request: naming the reader, the task, the facts the model cannot know, and the shape you want back. That is durable because it is really about communication, and it transfers between tools and model versions. The tricks — magic phrases, secret templates — are the part that does not last.
- How long does it take to get noticeably better at prompting?
- Most people notice a difference within a few hours of deliberate practice on their own real tasks, because the biggest gains come from a handful of habits: adding context, being specific about format, and treating the first answer as a draft. Getting consistently good takes longer, and mostly comes from noticing why a particular answer disappointed you.
- Do I need to learn a special syntax or format?
- No. Plain sentences work. Structure helps — a short list of what you want and what to avoid is easier for a model to follow than a paragraph — but there is no required format, no special punctuation, and nothing you have to memorize. The templates on this site are written in ordinary English for that reason.
- Why does the same prompt give different answers on different days?
- Models choose among likely continuations with a deliberate element of randomness, so two runs of the same prompt can differ even minutes apart. Providers also update models over time. If you need consistency, be more specific about the format and include an example; if the difference matters for a decision, ask twice and compare rather than trusting the first run.
Where to go next
If the mechanics of models are unfamiliar, AI Fundamentals explains tokens, context windows, and why models make things up — all of which shape how prompts behave. Once you can write a specific request, Using AI in Everyday Work applies it to emails, documents, and meetings, and the Agentic AI hub covers what changes when a model can act rather than only answer.