This is Part 3 of our AI-Proof Your Career series. If you're just joining, start with the foundational article (How to AI-Proof Your Career) and Part 2 (The 5 AI Skills Worth Learning First). Our series will conclude with an expert-led AI training course later this year.
In Part 2 we covered which AI skills are worth learning. Now for the question that tends to trip people up, how to go about actually learning them.
This is where a lot of us get stuck. You know you should build these skills. You may have even started watching a video or two. But somewhere between good intentions and a full life, "learn AI" becomes one of those goals that keeps rolling over to next month.
Here's the good news: learning to use AI is a lot more achievable than the noise around it makes it seem. You don't need a degree, a coding bootcamp, or hundreds of hours. You need the right approach, and you need to sidestep a few roadblocks that keep people stuck.
Learning About AI vs. Learning to Use It
There's a difference between understanding how AI works and knowing how to use it. You could read everything about how AI works and what it’s capable of and still freeze the moment you open the tool and stare at an empty prompt box.
You don't need to know what's happening under the hood (remember, we're not trying to build the technology, we're trying to use it). You need to be able to sit down, give a tool a real task, and get something useful back. That's a practical skill, and practical skills are built by doing.
The Trap of Passive Learning
Here's the most common way people stall out: they treat learning AI like a subject to study instead of a skill to practice.
They watch the tutorials. They save the posts. They collect information like they're studying for a test that never comes. Months go by, and they've consumed a bunch of content without ever using AI on a single real task. It feels productive because it looks like effort, but nothing has actually changed.
Passive learning is comfortable. There's no risk of getting it wrong, no awkward fumbling around. But that comfort is exactly the problem. You can't watch your way to a skill. At some point you have to open the tool and mess up a few times at first (and that's completely fine).
Learn by Doing—Inside Your Real Work
The fastest way to learn is also the most practical: use AI on the work that's already on your plate.
Don't set aside a separate AI learning time that you know you'll never get to. Instead, pick one real task this week (a presentation you need to create, some background info you need to gather) and do it with the assistance of an AI tool. It'll feel clumsy the first time. That's not a sign you're failing, that's you learning how to use it.
This works because it removes your two biggest excuses at once. You don't need extra time, because you were doing the task anyway. And you don't need a pretend practice scenario, because the work is actually something you need to do. Try this for a few weeks and you'll have built more genuine skill than months of watching ever could.
Where Teaching Yourself Tends to Stall
Learning by doing is a great start, but it has its limitations
The first is that you don't know what you don't know. When you're teaching yourself, you aren’t able to see the better techniques you’re unaware of. You get to "good enough" and stall out, with no idea that some small changes could lead to a huge increase in output.
The second is that you’re missing out on feedback. When you practice alone, no one is there to tell you whether you're doing it well. You could reinforce a bad habit for weeks without realizing it. Skills improve fastest when someone can look at your work and say "close - but try it this way".
The third limitation is accountability. How many self-directed goals have you actually finished? Without a little structure or a few people expecting you to show up and make progress, it’s so much easier to put off the learning for something else.
None of this means teaching yourself is a bad idea. Plenty of people build real skills on their own. But if you've tried and stalled out, it may be time to apply a bit of structure to your approach.
When a Little Structure Goes a Long Way
This is why we're building an expert-led AI training course to close out this series.
A good training program solves the exact three things that trip up solo learners. It shows you the techniques you wouldn't have found on your own. It gives you feedback in real time, so you're not playing the guessing game. And it creates structure with accountability.
Ours will run as small, live cohorts for that reason. You'll practice on real prompts, get feedback in the moment, and learn alongside other professionals working through the same things. More information on the training will be announced shortly!
The Bottom Line
Skip the passive consumption, practice on real work, and find feedback and structure. Whether you go it alone or with a little guidance, you learn this by doing it, not by reading about doing it.
Next up in the series: Once you've built the skills, how do you actually get credit for them at work, and turning "I've been using AI" into a raise, a new role, or a resume that stands out.
