Growth: Learning Finally Became Positive Feedback

Looking back on how I learned growing up, I realise almost all the feedback I got was negative.

Red crosses in exercise books, points circled and deducted on tests, those few seconds of silence when a teacher called on me and I had no answer. We were rarely rewarded for "getting it" and far more often reminded that we "didn't get it." Over time, learning got wired to one feeling in my head: tension, fear of being wrong, fear of being found out.

No one there when you get stuck

What really hurt wasn't that the problems were hard. It was that once you got stuck, nobody caught you.

The teacher explained it once at the front of the room, and you didn't follow. The lesson wasn't going to slow down for you; raising your hand again meant worrying everyone would think you were slow. At home you opened the textbook, and its explanation was the same words the teacher had used, just laid out differently. You asked a classmate, who only half understood it too. More often than not it ended the same way: you memorised that step, passed the exam, and never quite felt sure of it.

That unsureness compounds. One concept you never really absorbed makes the next three built on top of it wobble too. Plenty of people decide they're "just not cut out for this." I increasingly suspect the problem isn't being cut out for it; it's that every time they got stuck, there was nobody around they could keep asking.

Asking stopped costing anything

After AI arrived, I felt for the first time that learning could feel different.

I have no programming background; my undergraduate and master's degrees were both in finance. When I started building my own products, a lot of it had to be learned from scratch.

While I was building duetfolio, a wall of red error text suddenly appeared in the terminal, and I couldn't understand a word of it. In the past, that moment would probably have been the beginning of giving up. That day I pasted the whole error in and asked: what does this mean? Once it explained, I asked: why did it happen? Then: how do I avoid it next time? By the end of that chain of questions the problem was fixed, and for the first time I could more or less read what an error message was telling me.

The biggest change for me wasn't that answers came faster. It was that asking questions stopped costing anything.

No worrying that the question is too basic, that I'm wasting someone's time, that they'll get impatient. Ask the same question three different ways and it answers seriously every time; tell it to "explain it more simply" or "give me an everyday example" and it does. The reluctance that sits between people, the fear of looking like you don't know, simply isn't there.

Feedback went from a term to a single exchange

Later I worked out that positive feedback isn't being praised. It's "the confusion I just had has been cleared up."

Every "why" I understand is a small confirmation: so I can figure this out. Enough of those, and the feeling of learning flips. I used to dread running into things I didn't know. Now, when I hit something I don't know, my first reaction is curiosity, and I want to ask.

Learning used to be hard to stick with because the payoff came too late and too rarely: you waited until the end of term, until the exam, to find out how you'd done, and most of the time in between was just frustration. AI has shrunk the feedback loop from "a term" to "one exchange." Ask once, understand a bit; ask again, understand a bit more. When feedback is fast and frequent enough, people naturally want to keep going.

Where the positive feedback ends

If I'm going to talk about positive feedback, I should be clear about its limits too.

AI gets things wrong, and when it's wrong it sounds exactly as sure as when it's right. So I keep two habits: for anything that matters, I check it against the official documentation or what actually happens when it runs; and once I think I understand something, I try to explain it in my own words. If I can't, I don't really understand it yet.

One more thing matters even more: asking "why" again and again only works if you're doing the thinking. If you just toss the question over and copy the answer back, what you get isn't positive feedback, just a faster way to copy homework. The real change happens in those follow-up questions; every "why" is your own thinking pushing forward.

Once the barrier is gone

As kids we were always told "if you don't understand, ask." In reality, the bar for asking was always high. Now that bar is nearly gone.

Our generation may be the first group of learners who can keep asking about anything, at any time, for as long as we need. That's not just an efficiency gain. The underlying colour of learning has changed: from fear of making mistakes to the pleasure of understanding.

If you've ever felt you "just can't learn this," try again a different way. This time, when you get stuck, you can keep asking.