How to learn AI with no technical background: where to start

Using AI tools today is mostly a matter of writing, or speaking, in plain English. If you can write a clear email or explain a task to a new colleague, you already have the most useful skill. You don't need a coding course, and you certainly don't need a master's degree.

Plenty of people have already started. Ofcom's 2026 survey found that 54% of UK adults use AI tools such as ChatGPT, Copilot or Gemini, up from 31% a year earlier. Confidence lags behind use: in a 2024 survey for the UK government, only 21% of adults felt confident using AI at work. That gap is about practice, not technical ability.

Experience counts for more than technical knowledge. Someone who knows their work well can tell a good answer from a bad one straight away, and that judgement is exactly what AI lacks. I teach digital tools one-to-one every week, and confidence nearly always comes from using a tool on your own real work, not from theory.

There's no fixed starting line. Find the description that fits you and begin there:

• You've never tried it. Open a free tool and ask it about something you know well, such as your job, your town or a hobby. Notice where it's right and where it's wrong. That one conversation teaches you more than an hour of videos.
• You tried it once and weren't impressed. The usual reason is a short, vague request. Give it the background a colleague would need, or ask it to ask you questions before it answers.
• You use it now and then. Pick one task you do every week and use AI for it every time for a fortnight. Save the prompts that work so you can reuse them.
• You use it most days. Try what you haven't yet: uploading a document to question it, using voice on the move, or giving the tool a standing brief about you and your work.

1. Choose one free tool. ChatGPT, Microsoft Copilot, Google Gemini or Claude. Sign up with your email address. If your employer provides one, use that for work tasks.
2. Change one setting. In a personal account, switch off the use of your chats for training. In ChatGPT it's Settings, then Data controls, then "Improve the model for everyone". The others have similar switches.
3. Pick a real task from today. An email you need to send, a long document you need to understand, a plan you need to make.
4. Explain it as you would to a new colleague. Who you are, what you need, who it's for and what a good result looks like.
5. Reply to improve it. "Shorter." "Warmer." "Put that in a table." "What have you assumed?" The first answer is a draft, not the final word.
6. Check it. Read it as the person receiving it would. Check any fact, figure or name you'll rely on.
7. Save what worked. Copy the prompt into a document so you can use it again next time.

Try: "I'm a team manager and I've never used AI before. Ask me five questions about my week, one at a time, then suggest three tasks you could help me with and show me how to ask for each one."

1. Asking. Good requests usually have four parts: a role ("You're an experienced bookkeeper"), the context, the task and the format you want back. You can watch this structure work in the prompt playground.
2. Conversing. Treat it as a conversation, not a search box. Push back, ask for options and ask it to explain its reasoning.
3. Checking. AI sometimes makes things up and sounds sure of itself while doing it; this is called hallucination. Check facts, numbers, quotes and links before you use them.
4. Protecting. Keep personal and confidential information out of free and personal accounts, and follow your employer's policy.

None of these is technical. When the UK public were asked which AI skills matter most, the top answers were keeping information safe and private (62%) and judging whether AI's answers are accurate (56%).

Weekends off. By the end you'll have a handful of prompts that work for you and a clear sense of where AI helps in your week. If you run your own business, the next step is putting AI into your daily workflow.

• AI (artificial intelligence): software that does tasks we used to think needed a person, such as writing, recognising speech or spotting patterns.
• Generative AI: AI that creates new text, images, audio or video in response to a request. ChatGPT, Copilot, Gemini and Claude are all generative AI.
• Chatbot or AI assistant: a generative AI tool you talk to in a chat window or by voice.
• Large language model (LLM): the engine inside a chatbot, trained on huge amounts of text to predict what words should come next.
• Prompt: whatever you type or say to the AI. A good prompt is simply a clear request.
• Hallucination: when AI states something false as if it were true.
• AI agent: AI that can carry out several steps on its own, such as searching the web, filling in a form or booking something, usually with you approving along the way.
• Automation: connecting tools so that one thing triggers another, for example a new enquiry creating a draft reply and a calendar entry.