How to properly
learn AI
TL;DR: Start with a free course to understand the basics, then pick a real task and try doing it with AI. If you can’t think of a task, ask AI to help you find one. Learn as you go: look for help when you get stuck, and keep noticing where AI might be useful in your work or daily routine. Start with the cheapest model that does the job.
There is no single right way to learn AI. People have different goals, different tools, and different ways of learning.
But my preferred way is learning by doing.
You can spend a long time reading about AI, watching videos, or collecting courses. That can help you understand the vocabulary, but it doesn’t necessarily teach you how AI could be useful in your work or life. The fastest way to find that out is to try it on a real task.
Start with the basics, for free
I don’t recommend starting with paid courses. At least, you don’t need to buy one just to get started. There are good free resources that can introduce you to the basic ideas and show you what these tools can do.
For example:
You don’t need to memorise everything in a course. The goal is to get familiar with the basics: what AI tools can do, what they’re good at, and where you should check their work.
Then move on to trying something yourself.
Pick a task you actually need to do
Think of something you’d like AI to help with. It doesn’t have to be a big project or an impressive use case. It could be a task you do often, something that takes too much time, or something you’ve been putting off because you don’t know where to start.
Maybe you need to organise your notes, prepare for a meeting, turn a long document into a summary, or come up with a first draft. The task should be small enough that you can tell whether the AI helped.
If nothing comes to mind, ask AI to help you find a starting point. You can use this prompt:
I want to learn how to use AI in my everyday life or work. Ask me questions one at a time about the tasks I do, the things I find time-consuming, and the problems I’d like to solve. Then suggest three simple, low-risk tasks I could try with AI.
For each suggestion, explain what I should ask AI to do and how I can tell whether it helped. Keep the suggestions realistic for a beginner. Don’t assume I have access to paid tools.
Answer the questions honestly and give the tool enough detail to make useful suggestions. If its first ideas don’t fit, tell it why and ask for different ones.
Try it, then adjust
Once you have a task, explain what you want the AI to do. You don’t need to write a perfect prompt. Start with a clear request, look at the result, and then say what’s missing.
For example, you might say:
Summarise these notes into the three decisions we made, the open questions, and the next steps. Keep the wording simple.
If the result is too vague, tell it what to change. If it misunderstands the task, explain the context. If you’re not sure what to ask for, you can ask the tool to help you improve your request.
The point is not to get everything right on the first try. You’re learning what information the tool needs, what kinds of answers it gives, and how to steer it toward something useful.
And remember to check the result. AI can make mistakes, leave out important details, or sound confident when it’s wrong. For anything important, verify the facts and use your own judgment.
Get help when you get stuck
When you hit a problem, you don’t have to figure it out alone. Search YouTube for a walkthrough, or ask the AI tool itself to explain what to try next.
You can be specific:
I’m trying to use AI to [describe your task]. I tried [describe what you did], but [describe what went wrong]. Explain what I could try next, one step at a time.
A useful skill is learning how to ask for the kind of help you need. If an explanation is too technical, say so. If you want an example, ask for one. If you’re using a particular tool, name it.
Make it a daily habit
After your first experiment, keep an eye out for other possibilities. Each day, ask yourself: Is there something from my work or daily routine I could try doing with AI?
You might not find a good use for it every day. That’s fine. The point is to build the habit of noticing tasks you could experiment with, then testing whether AI actually makes them easier or better.
Keep the experiments small. Try one task, see what happens, and decide whether it’s worth using again. Over time, you’ll learn more about what works for you than you would by collecting a long list of theoretical use cases.
Pay attention to the cost
AI tools and models can have different prices. Start with the cheapest model that works for your task. If it gives you a good enough result, there’s no reason to pay more.
Try a more expensive model when you have a specific reason: the cheaper one keeps getting the task wrong, the task needs more careful reasoning, or the result matters enough that the difference is worth the cost.
You don’t need the most powerful model for every little thing. Learn what the task requires, compare the results when it matters, and spend accordingly.
That’s how I’d start: learn the basics for free, choose a real task, and try it. Then keep experimenting.