Writing Help: Emails, Letters & Messages
Writing is the most immediately useful thing AI can help you with. Not because it replaces your voice — it doesn't, and shouldn't — but because getting the right words onto a page is genuinely hard, and AI is remarkably good at helping you do it faster, more clearly, and with less stress.
In this lesson we'll work through the writing situations most people face regularly: professional emails, formal letters, tricky messages to people you know, and those situations where you know what you want to say but can't find the right way to say it.
Why AI is so good at writing help
AI has been trained on billions of examples of human writing — formal and informal, technical and conversational, in hundreds of styles and contexts. When you describe a writing situation, it draws on all of that to produce something appropriate to the context. It doesn't just generate generic text: it calibrates tone, formality, length, and structure to what you've told it.
The key insight is that you're not outsourcing your voice — you're borrowing a first draft. The best approach is to give AI the raw ingredients (the situation, the tone, the key points you need to make) and let it produce something you then read, adjust, and make your own.
Professional emails
This is where most people start, and for good reason. Professional emails are one of the most time-consuming, anxiety-inducing parts of many people's working day. Getting the tone exactly right — assertive without being aggressive, friendly without being unprofessional, direct without being rude — takes real skill and time.
Here's the approach that works best:
Write a professional email to [who it's going to] about [the topic]. The context is [brief background]. I need to [what you want to achieve — request, inform, complain, follow up, etc.]. The tone should be [polite and firm / warm and friendly / formal / etc.]. Keep it to around [length].
Let's see this in practice. Here's a real example:
Write a professional email to my line manager requesting a meeting to discuss my workload. I've been consistently working late for the past six weeks and I want to raise this formally but constructively — I'm not looking to complain, just to find a solution together. Tone should be professional and collaborative. Around 150 words.
A well-structured email with a clear subject line, a professional opening, a factual description of the situation without emotional language, a collaborative framing of the desired outcome, and a polite request for a meeting. Ready to personalise and send.
The before-and-after difference
One of the most useful things AI can do is improve something you've already written. Paste in your draft and ask it to make it clearer, shorter, more professional, or less confrontational.
"Hi, I've emailed you three times about this and still haven't heard back. Can you please respond to my query about the invoice? I need this sorted urgently."
"Hi [Name], I wanted to follow up on my previous messages regarding invoice #1042. Could you let me know when might be convenient to discuss? I'd be grateful for an update at your earliest convenience."
The second version says the same thing — you've chased three times, you're frustrated, you need a response — but it says it in a way that's more likely to get a reply and less likely to damage the relationship. That's the value AI brings to difficult correspondence.
Formal letters
Formal letters — to a council, a landlord, a bank, an insurer, a solicitor — follow conventions that most people only encounter occasionally and therefore find stressful. AI handles these with ease.
Write a formal complaint letter to my energy supplier. I was overcharged by £340 on my last bill and have been trying to resolve this for six weeks without success. I'm now formally requesting a refund and a written explanation within 14 days. Include the appropriate formal letter structure, today's date, and a reference to my right to escalate to the Energy Ombudsman if not resolved.
AI may include specific regulatory references, deadlines, or legal terms. These are usually accurate but should be verified for anything important. For genuine legal disputes, treat AI output as a starting point and confirm the specifics with Citizens Advice or a solicitor.
Difficult personal messages
Sometimes the hardest writing isn't professional — it's personal. Telling a friend something difficult. Responding to a family member you've fallen out with. Writing a message of condolence when you don't know what to say.
AI is surprisingly good at this. Not because it feels empathy, but because it has processed enormous amounts of human emotional communication and can help you find words that are kind, honest, and appropriately calibrated to the relationship.
I need to write a text message to a close friend declining their wedding invitation. I can't attend for financial reasons but don't want to make it about money. I want to express that I'm genuinely sorry and that I care about them deeply, while keeping it warm and personal rather than formal. About 3-4 sentences.
Three quick writing tasks AI handles brilliantly
- Proofreading and editing. Paste any text and ask AI to "proofread this for grammar, clarity, and flow without changing my voice."
- Making things shorter. "Reduce this to 100 words without losing the key points." Remarkably useful for job applications, reports, and social media.
- Translating tone. "Rewrite this to sound less aggressive" or "Make this more formal" — instant improvements on anything you've drafted.
Think of a real email you've been putting off writing — or one you've already written that you're not quite happy with. Use AI to either draft it from scratch or improve your existing draft. Notice specifically how it handles the tone.
AI is excellent at writing help because it calibrates tone and structure to context. The best approach is brief-and-refine: give it the situation, objective, tone, and length, then make the output your own. It works for professional emails, formal letters, and difficult personal messages alike. Verify anything with legal or financial implications.
Research & Finding Information
For most of history, getting a well-informed answer to a complex question meant going to a library, finding a knowledgeable person, or spending hours searching the internet and piecing things together yourself. AI has fundamentally changed this. You can now ask nuanced, specific questions and get well-structured, synthesised answers in seconds.
But — and this connects directly to Lesson 1.4 — using AI for research requires knowing exactly where to trust it and where to verify. This lesson covers both.
What AI research is brilliant for
AI excels at research tasks that involve synthesis, explanation, and exploration rather than precise factual verification.
- Understanding a topic you know nothing about. Getting a clear, structured overview of something unfamiliar — a medical condition, a legal concept, a financial product, a historical event.
- Comparing options. "What are the pros and cons of X versus Y?" — asked conversationally, with your specific situation as context.
- Unpacking jargon. Pasting in a complex document and asking "explain this in plain English."
- Generating questions to ask. Before seeing a doctor, solicitor, or financial adviser: "What questions should I be asking about [situation]?"
- Getting a starting point. When you're about to research something properly, AI can give you the lay of the land so your reading is more efficient.
Unlike a search engine, AI lets you ask questions in your specific context. Not "what is a fixed-rate mortgage" but "I'm 52, planning to retire in 8 years, and considering switching from a variable to a fixed-rate mortgage — what should I be thinking about?" The personalised answer is infinitely more useful than a generic article.
Explaining complicated things simply
One of AI's most underrated capabilities is explaining complex topics at exactly the right level for you. You can calibrate this directly:
Explain how interest rates affect house prices. I have a basic understanding of economics but no formal training. Use an analogy if it helps, and keep it to a few paragraphs — I just need enough to follow the news intelligently.
Or use the classic instruction that unlocks surprisingly clear explanations:
Explain what a pension is and how it works — as if explaining to a 12-year-old who has never heard the word before.
This works for genuinely complex topics too. The instruction to explain simply forces AI to drop jargon and find the clearest possible framing — which is often more useful than a technically accurate but dense explanation.
Summarising long documents
One of the most practical time-savers AI offers is reading things for you. Paste in a long document, a terms and conditions page, a research paper, or a news article, and ask AI to summarise it.
[Paste the full text here] Summarise this in bullet points. I want to know: the main argument or purpose, any commitments or obligations I'm agreeing to, any important deadlines or conditions, and anything unusual I should be aware of.
This is particularly useful for contracts, insurance documents, tenancy agreements, and anywhere else you're expected to read something long and dense but rarely actually do.
Preparing for appointments and conversations
Before any important meeting where you need to be well-informed, AI can help you prepare:
I've just been diagnosed with type 2 diabetes and have my first appointment with a specialist next week. I want to make the most of the appointment. What are the most important questions I should ask? I'm 58, currently managing with diet alone, and I'm particularly worried about long-term effects on my eyes and kidneys.
AI gives you context and helps you ask better questions — it doesn't replace professional advice. For anything with real consequences: use AI to understand the landscape and prepare, then speak to the relevant professional for the actual decision.
The verification habit
Remember Lesson 1.4. When you use AI for research, two habits protect you:
- Ask AI to flag its own uncertainty. Add "and tell me which parts of this you're less certain about" to any research prompt. Good AI tools will honestly identify where they're less reliable.
- Verify anything that matters before acting on it. Use AI to understand — then confirm with an authoritative source before doing anything consequential.
Think of something you've been meaning to understand better — a topic in the news, a financial concept, a health question you've been wondering about. Ask AI to explain it at your level. Then ask a follow-up question that takes it deeper. Notice how it feels compared to searching the internet for the same thing.
AI excels at synthesising, explaining, and exploring information — especially when you can give it personal context. It's brilliant for understanding topics, comparing options, summarising documents, and preparing questions. Always verify anything with real consequences. The habit of asking AI to flag its own uncertainty is your best defence against confident errors.
Planning & Organisation
Planning is one of the most satisfying things to use AI for — because the output is immediately actionable. Whether it's a holiday itinerary, a project plan, a weekly schedule, or a home renovation checklist, AI can turn a vague goal into a structured, practical plan in minutes.
The reason it works so well is that planning tasks follow predictable structures. A good plan has goals, steps, a sequence, timelines, and dependencies. AI understands all of these and applies them automatically when you describe what you're trying to achieve.
Holiday and travel planning
This is where many people first discover just how useful AI can be for planning. Instead of spending hours on TripAdvisor, travel blogs, and Google Maps, you can describe exactly what you're looking for and get a thoughtful, personalised itinerary.
Plan a 7-day trip to Portugal for two adults in their late 50s. We love food, history, and walking — but nothing too strenuous. We don't enjoy touristy crowds and prefer to eat where locals eat. We'll be visiting in May. We'll be based in Lisbon for the first 3 days and want to do a day trip or two from there, then move somewhere else for the final 4 days. Include suggested restaurants for 2-3 evenings and any practical tips specific to our age group.
What you get is a structured, day-by-day plan that reflects your specific preferences. Not a generic "top ten things to do in Lisbon" — a personalised response to your actual requirements. Follow up by asking it to add a packing list, suggest which travel insurance features matter most for your trip, or recommend what to research further before booking.
AI may suggest restaurants, hotels, or attractions that have closed, changed, or are simply inaccurate. Use its output as a strong starting point, then verify specific bookings with current sources before committing.
Project planning
Any project — at work or at home — benefits from a structured plan. AI can help you build one quickly, whether that's a kitchen renovation, a website redesign, a house move, or a work deliverable.
I'm planning to redecorate my living room on a budget of around £1,500. It's currently quite dated — magnolia walls, old carpet, heavy curtains. I want it to feel lighter, more modern, and comfortable. I'm doing most of the work myself but may hire someone for painting. Create a step-by-step project plan including what to decide first, what to buy, a rough sequence of tasks, and things I'm likely to forget.
The "things I'm likely to forget" instruction is particularly useful — it prompts AI to add the practical details that experienced people know but beginners miss. For the renovation example, this might include: allowing paint to dry fully between coats, ordering 10% extra flooring for cuts and waste, measuring door frames before buying furniture, and testing paint colours in natural light.
Weekly and daily planning
AI can help you structure your time, not just individual projects. If you struggle with organising your week or tend to leave important things until the last minute, try this:
Help me plan my week. Here's what I need to accomplish: [list your tasks and commitments]. I work best in the mornings, lose focus after 3pm, and have these fixed commitments: [your meetings, school runs, etc.]. I also want to protect time for exercise and one evening for myself. Create a realistic daily schedule that doesn't overload any single day.
Making decisions
Planning often involves decisions, and AI can help you think through them structurally. Rather than just asking "should I do X?", frame it as a structured decision:
I'm trying to decide whether to take on a freelance project on top of my full-time job. The project would pay around £800 over 6 weeks and require about 5 hours per week. Help me think through this decision. What factors should I weigh up? What questions should I be asking myself? What are the risks I might not have considered?
The most powerful planning technique is to ask AI to identify what you haven't thought of. "What am I likely to forget?" "What could go wrong?" "What would an experienced person add to this plan?" These prompts consistently surface useful things that don't occur to people in the moment.
Think of something you're planning — even something small: a dinner party, a weekend away, a home task you keep putting off. Ask AI to create a practical checklist or plan for it, and include the instruction "and add anything I'm likely to overlook."
AI turns vague goals into structured, actionable plans quickly. It works for travel, projects, time management, and decisions. The most valuable instruction is asking it to surface what you haven't thought of. Always verify specific recommendations — restaurants, hotels, tradespeople — with current sources before committing.
Learning New Things
AI is one of the most patient, available, and adaptable teachers imaginable. It never gets frustrated when you ask the same question twice. It can explain the same concept six different ways until one clicks. It's available at any hour, on any topic, at exactly your level. For anyone who loves learning — or who needs to pick up new skills without formal courses — this is genuinely transformative.
AI as a personal tutor
The traditional barrier to having a personal tutor is cost. A good one charges £40–80 an hour, and even a mediocre one is beyond most people's budgets for ongoing learning. AI gives everyone access to something that functions like a knowledgeable, infinitely patient tutor — available any time, for any subject.
The key is learning how to use it like a tutor rather than like a search engine. A search engine gives you a link. A tutor checks your understanding, adapts to your gaps, and pushes you further when you're ready.
I want to understand how the stock market works. Don't just explain it — teach me. Start with what I should know first, then ask me a question to check my understanding before moving on. I'm a complete beginner.
This prompt sets up a genuine back-and-forth teaching session. AI will explain a concept, ask you a question, receive your answer, correct any misunderstanding, and build on what you've got right. The learning is active rather than passive — which means you retain far more.
Learning practical skills
AI is excellent for practical skill learning — anything where the knowledge can be described in text. Cooking, language learning, DIY, gardening, photography, playing an instrument — if you can ask questions about it, AI can help you learn it.
I want to learn to cook properly. Right now I can only make very basic things. I'd like to learn five fundamental techniques that, once mastered, would open up a wide range of recipes. For each technique, explain what it is, why it matters, and give me one simple recipe I can use to practice it this week.
The "explain it back to me" technique
One of the most powerful learning techniques — used in education for decades, recently formalised as the Feynman technique — is to explain what you've just learned in your own words. AI is a perfect partner for this:
I've just read that [explanation of a concept you're learning]. Let me explain it back to you in my own words: [your explanation]. Can you tell me what I've got right, what I've got wrong or missed, and what I should focus on to deepen my understanding?
This is the single most effective way to use AI for learning. Articulating what you know reveals what you don't. AI's feedback on your explanation is precise, constructive, and immediately actionable.
Building a learning plan
For anything you want to learn seriously over time, AI can help you build a curriculum:
I want to learn Spanish to conversational level over the next 12 months. I have about 30 minutes a day available. I've studied French at school, which gives me some European language foundation. Build me a realistic month-by-month learning plan, recommending the best free and low-cost resources for each stage, and the milestones I should be hitting along the way.
Each learning session builds on the last because AI has your full conversation in context. You can return to a subject days later and say "last time we discussed X — I've been practising and I have a question about Y." It remembers the thread. Most tutors don't offer this kind of continuity at any price.
Pick something you've always wanted to understand better — a subject, a skill, a concept. Ask AI to teach you the first thing you'd need to know, then ask it to check your understanding with a question. Have the conversation for at least ten minutes. Notice the difference between this and reading an article about the same topic.
AI is a patient, available, adaptable personal tutor for virtually any subject. The most effective techniques are: asking it to teach rather than just explain, using the "explain it back" method to test your understanding, and building conversation-based learning sessions that go deeper over time. For any serious learning goal, ask it to help you build a curriculum.
Part 2 Exercise
Part 2 has been practical from the start — now bring it all together with four tasks drawn directly from your own life. The goal is to produce something genuinely useful, not just to complete an exercise.
Each task should produce something you actually keep and use. Work through them in ChatGPT or Claude and note what you learned.
You've now applied AI to four of the most valuable everyday use cases: writing, research, planning, and learning. These are skills that compound — the more you practice them, the faster and more natural they become. Part 3 is where things get more sophisticated: we go deeper on prompting, learning how to get dramatically better results through structure, roles, and iteration.