Learning · 7 min read
How to learn a new skill when you're not sure you'll stick with it
Most advice assumes you've already committed. This is the version for people who are still deciding — pick a skill, get one hour of real evidence, and let the result decide whether you continue.
18 August 2026
The short answer
To learn a new skill, start with one hour of the real activity rather than a course about it. Choose a skill you can produce something small with, do the smallest complete version of it, notice what the hour actually felt like, then decide whether to book the next hour. Commitment should be the result of evidence, not the price of entry.
There is a specific kind of stuck that nobody writes guides for. You're not lazy. You're not unmotivated. You just don't know whether this thing is yours yet, and every piece of advice on the internet starts at step two: 'commit, build a habit, show up daily.' Committing is exactly the thing you can't do honestly, because you don't have the information yet.
So here's the version that starts at step one. The goal of your first session is not progress. It's evidence.
Pick a skill you can make something with
Some skills produce a visible object in an hour — a drawing, a working script, a loaf, a 30-second edit, a repaired bike brake. Others only pay out after months: a language, an instrument, statistics. Start with a producing skill even if the long one is what you secretly want, because a producing skill gives you feedback while you still have the energy to use it.
If the long skill is the real target, find the producing version of it. Not 'learn Spanish' but 'order food in Spanish at the place down the road'. Not 'learn to code' but 'make a page that shows my football team's next fixture'. The narrow version is not a lesser goal. It's the same goal, sized so you find out something today.
Do the smallest complete version
Complete matters more than small. Half a thing teaches you almost nothing, because the interesting difficulty of most skills lives in the finishing — the bit where it has to actually work, actually be readable, actually be eaten. People who abandon skills usually abandon them in the messy middle, having never once experienced the finish.
- 01Write down what 'done' looks like in one sentence, before you start.
- 02Set a timer for 60 minutes and make the worst acceptable version.
- 03Finish it even if it's embarrassing. Especially if it's embarrassing.
- 04Show it to one person who will react honestly.
The first version is not a performance. It's a probe you send into the skill to see what comes back.
Read the hour honestly
This is the part that actually decides whether you continue, and most people skip it. Right after you finish, before the feeling fades, answer three questions:
- Where did time disappear? The stretch you didn't notice passing is the part of this skill that fits you.
- What was hard — interesting-hard or heavy-hard? Interesting-hard is a puzzle you keep turning over. Heavy-hard is dread. They feel completely different in the body and people routinely mislabel one as the other.
- Did you want to show someone? Not 'were you proud'. Wanting to show someone is a stronger predictor of continuing than satisfaction is, because it means the thing has social meaning for you as well as private meaning.
One 'no' is not a verdict. Three 'no's after two honest attempts is real information, and dropping the skill at that point is a good decision, not a failure. You spent two hours and learned something true about yourself. That's a cheap price for accuracy.
How long it actually takes
The 10,000-hour number that floats around describes people at the top of competitive fields. It has almost nothing to do with you learning to do something usefully. The more relevant research finding is far kinder: most of the improvement in a new skill happens early and steeply, and roughly 20 hours of focused, feedback-rich practice takes most people from hopeless to visibly capable.
Twenty hours is forty-five minutes, three times a week, for ten weeks. That's the real scale of 'learning a new skill'. Not a life sentence. A season.
Stack the second hour before motivation leaves
Motivation after a first session is a resource with a short shelf life. The single highest-return thing you can do with it is not to practise more — it's to decide, in that moment, when the next hour happens and what its 'done' will be. Write it somewhere you'll see it. A specific next session survives a flat Tuesday. A general intention does not.
Why this works better than choosing carefully
People assume the way to avoid wasting time is to think harder before starting. In practice, thinking produces predictions about yourself, and human predictions about what we'll enjoy are famously poor. Doing produces data. An hour of the real thing beats a month of imagining the real thing, every single time — and it costs less.
That's the whole loop, and it doesn't only apply to skills: explore what's out there, try a small version, learn from what actually happened, update what you believe about yourself. Run it enough times and the question stops being 'what should I learn' and becomes 'what am I testing next'.
Questions people ask
- How long does it take to learn a new skill?
- Long enough to be useful is usually 15 to 25 hours spread over a few weeks — not 10,000. The 10,000-hour figure describes elite performance, not competence. For deciding whether a skill is for you, one honest hour is normally enough.
- What's the best way to learn a new skill fast?
- Work on a real, small output from the first session. Producing something badly teaches you more per hour than consuming lessons about producing it well, because it exposes exactly which sub-skill you're missing.
- How do I know if a skill is worth learning?
- Check two things after your first hour: did the difficulty feel interesting or just heavy, and did you want to show someone the result. Wanting to show someone is the most reliable early signal that you'll keep going.