Think about the last time something stopped working. The Wi-Fi, a spreadsheet formula, the dishwasher that suddenly started blinking a code you’d never seen before.
What did you do first?
Some people look for the manual, a tutorial, or someone who has fixed it before. Others start pressing things. They unplug it, change a setting, see what happens, and try something else. They aren’t being careless. They’re gathering information the only way they ever really learned to.
That second group is full of people who taught themselves most of what they know. And most of them have no idea their approach is unusual.
A lot has been said about the difference between people who read instructions and people who just figure things out, but for self-taught people, it goes deeper than a preference. It’s how their thinking was built.
Self-taught people learned to explore because nobody showed them the one right way
When you learn something in a classroom, you usually get the method first. Here’s the formula. Here’s how the function works. Now practice it.
When you teach yourself, nobody hands you a method. You start with the problem and work backward. You try things that don’t work, and you figure out why they don’t work, which teaches you something the correct answer alone never would.
That describes a lot of people now, especially in certain fields. In software, it’s practically the norm.
In Stack Overflow’s 2024 survey of developers, 82% said they learn to code using online resources, compared with 49% who learned in school.
Plenty of those developers have degrees. But a big share of what they use every day, they picked up on their own by searching, testing, and breaking things until they worked.
Being taught one method can make people stop looking for others
A study out of MIT helps explain what self-taught people are doing differently, even if they can’t name it themselves.
Researchers built a toy out of colored tubes that could do four things: squeak, light up, play musical notes, and show a mirror image. Then they handed it to preschoolers at Boston’s Museum of Science under a few different conditions.
In one version, an adult said, essentially, this is how my toy works, and showed the kids only the squeak. In the others, the adult stumbled onto the squeak by accident, got interrupted partway through, or didn’t demonstrate anything at all.
The children who were deliberately taught the squeak played with the toy for less time and discovered fewer of its other functions than the kids in any other group.
Cognitive scientist Laura Schulz, one of the study’s authors, summed up the logic in MIT’s announcement of the findings: “If I teach you this one thing and then I stop, then you may say, ‘Well that’s probably all there is.'”
That isn’t a flaw in the kids. It’s a sensible shortcut.
If someone who knows more than you shows you how something works, it’s reasonable to assume they showed you the important part.
Self-taught people never got that signal. Nobody told them where the edges were, so they kept pushing on everything.
This was a study of preschoolers, not adults, so it’s a stretch to map it directly onto someone who taught herself bookkeeping at 35. And Schulz has been clear that the takeaway isn’t to stop teaching. Some things, like reading or driving a car, make no sense to learn by trial and error. The point is that instruction comes with a hidden cost, and self-taught people mostly skipped it.
More Bolde Stories
A familiar solution can keep people from seeing a better one
Psychologists have a name for getting stuck on the solution you already know: the Einstellung effect.
In experiments by Merim Bilalić, Peter McLeod, and their colleagues, master chess players were shown a board that could be won two ways. One was a well-known five-move checkmate. The other took only three moves.
Once the players recognized the familiar five-move pattern, they seemed unable to notice the faster route. When the researchers shifted a single bishop so the familiar pattern no longer worked, the same players found the quicker solution.
These were experts. Their training was exactly what got in the way.
This is part of why self-taught people can come across as unconventional problem solvers. They often never memorized the standard playbook, so they don’t automatically reach for it. They look at the problem in front of them instead of the one they were trained to expect.
To be fair, self-taught people aren’t immune to this. They build their own go-to moves, and those can get just as rigid as anything learned in a lecture hall. The difference is that their habits usually come from what worked for them, not from what someone told them works.
Self-taught people treat getting stuck as a normal part of learning
Ask anyone who taught themselves guitar, or coding, or how to fix a car, how much of the process involved being confused. The honest answer is usually most of it.
That turns out to matter.
Manu Kapur, a professor of learning sciences at ETH Zurich, has built his career around what he calls productive failure: letting people wrestle with a problem before they’re shown how to solve it. The struggle shows learners what they don’t understand yet, so the explanation actually sticks when it arrives.
The idea still has critics, especially among researchers who favor teaching first, but a 2021 meta-analysis of 53 studies that Kapur co-authored found that students who tackled problems before instruction generally came out ahead.
“Being frustrated and struggling are normal things,” Kapur told Edutopia. “In fact, if you’re not feeling those things, that means you’re probably not learning.”
Self-taught people absorbed that lesson without ever hearing the term. Being stuck doesn’t register as proof they’re failing. It registers as the middle of the process.
Part of that tolerance comes from why they were learning in the first place.
Writing in Psychology Today, Neil Farber notes that self-learning tends to be driven by genuine interest rather than external rewards. When nobody is grading you, getting stuck isn’t a mark against you. It’s just Tuesday.
That’s also why they can look oddly calm when a problem comes up at work. Someone who learned through clear instructions may feel lost once the instructions run out. Someone who never had them is already on familiar ground.
Self-taught people assume everyone else solves problems the same way
So why don’t they notice any of this?
Mostly because they’ve never watched themselves from the outside. When you’ve always solved problems by poking at them, that just feels like what solving a problem is.
Psychologists call the broader tendency the false consensus effect. In a classic Stanford study from the 1970s, Lee Ross and his colleagues asked students to walk around campus wearing a sandwich board sign.
The students who agreed believed most of their peers would agree too, while the students who refused believed most of their peers would refuse.
We use ourselves as the default setting. So the self-taught person who reverse-engineers every problem assumes that’s what everyone does. Then a coworker asks, wait, how did you know to try that? and they genuinely don’t have a better answer than I don’t know, I just figured it out.
Part of the reason they can’t explain it is that nobody gave them the vocabulary.
People who learn a field formally pick up its names for things. People who teach themselves often know how to do something long before they know what it’s called.
More Bolde Stories
Teaching yourself also leaves gaps you may not know about
None of this makes self-taught people better problem solvers across the board. It makes them different, and different comes with blind spots.
Kapur himself is careful about this.
In the same interview, he points out that failure on its own isn’t enough. In his model, the struggle has to be followed by someone with expertise who helps the learner make sense of what they found. Without that step, people can land on answers that work without ever reaching the best one.
Self-taught people usually skip that second step, because there’s nobody around to provide it.
In one Hacker News thread about self-taught programmers, a commenter described the result as knowledge gaps covered by “reinforced work-arounds” people built because they didn’t know the gap was there.
It’s a fair criticism, and plenty of self-taught people would admit it. You can spend years solving a problem the long way because you never came across the short one. Being unaware of how you think also means being unaware of what your thinking is missing.
The difference comes from how they learned, not how smart they are
It’s tempting to turn all of this into a story about self-taught people being naturally more creative. That’s not really what the research says.
What it says is more ordinary.
The way you first learn to handle a problem shapes how you approach the next one. Show someone the answer, and they tend to stop looking. Let them find their own way, and they tend to keep looking, for better and for worse.
Self-taught people just got the second version, often by accident. Maybe nobody was around to teach them. Maybe they couldn’t afford the class. Maybe they were too impatient to wait for one.
Whatever the reason, the habit stuck. They still open the problem before they open the manual. And most of them will go on assuming that’s what everyone does, right up until someone watches them work and asks how they knew to try that.
