Learn to write a clear prompt first, because every AI tool you will use at work runs on the same skill: telling it who it is, what it needs to know, what to do and what the answer should look like. Then learn what you must never paste into AI, and only after that try one small automation, starting with Prompt Writing 101, a free 8-minute lesson.
That order matters. Most people start with a tool, get a generic answer and decide AI is not for them. The problem is almost never the tool. It is the one-line request.
What are the three things to learn, in order?
- Prompting. A good prompt has four parts: Role, Context, Task and Format. With those four, ChatGPT, Claude, Gemini and Copilot all give you something you can use instead of something you have to rewrite.
- Safe use. Before you paste a client email or a spreadsheet into a chatbot, know which settings to check and which details to take out. This takes ten minutes to learn and protects you from the mistake that costs the most.
- One small automation. Once you can get a good answer by hand, learn how a tool like n8n, Make or Zapier can run that same request for you every time an email or form arrives.
You do not need to learn coding, machine learning or the math behind AI models for any of this. Those are useful for people who build AI products. They are not what makes office work faster.
How do I start this week?
- Day 1: Take one task you already do, such as a customer email or a meeting recap, and write it as a four-part prompt. Compare the answer with what you get from a one-line request.
- Day 2: Save the prompt that worked as a template with blanks, so you never start from an empty box again.
- Day 3: Check your chatbot's data settings and learn your company's AI policy, if it has one.
- Day 4 and 5: Learn the difference between a chatbot, an agent and an automation, then pick the one task you repeat most and see whether an automation could do the first draft.
What does this look like in a real lesson?
In Prompt Writing 101 you ask ChatGPT for a customer email twice. The first time you type one vague sentence and get back a long, generic email full of blanks like "[Your Company Name]". The second time you add a role (the owner of a small lawn care company), the context (existing customers, a new fall service), the task (a short announcement with one call to action) and the format (under 120 words, plain, subject line first). The second answer is ready to send after a quick read. You finish by saving the prompt as a reusable template in a Google Doc.
The next lesson, Use AI at Work Without Leaking Client Data, has you turn off model training in ChatGPT's data controls, strip a tenant's bank details and address out of a request, then catch two promises ChatGPT made up before the reply goes out.
What mistakes do beginners make?
- Starting with the most advanced tool. Agents and automations are easier once you can write a prompt that works by hand.
- Trusting the first answer. AI fills gaps with things that sound right. Check every fact, number and promise before anything leaves your desk.
- Pasting everything. Names, account numbers and addresses usually are not needed for a good draft. Use placeholders.
- Learning in the abstract. Use your own real tasks. A prompt for a made-up email teaches you less than one for the email you have to send today.
Key takeaways
- Learn prompting first: Role, Context, Task and Format.
- Learn safe use second: settings, what to take out, what to check.
- Add automation third, once one prompt works reliably by hand.
- No coding or math is needed for AI at work.
Frequently asked questions
- Do I need to learn coding to use AI at work?
No. Writing prompts, checking answers and building simple automations in tools like n8n, Make or Zapier all work without code. Coding helps later if you want to build custom tools, but most office uses never need it.
- Which AI tool should I learn first?
Start with the chat assistant your workplace already allows, usually ChatGPT, Claude, Gemini or Copilot. The prompting skill carries over between them, so the choice matters less than practicing on real tasks.
- How long does it take to learn the basics?
The three basics fit in about a week of short sessions. The first two free lessons take about 15 minutes together, and the habits come from using them on your own work every day.