online learning is stuck in 2012

The best way to learn a piece of software used to be to leave it.

You'd be stuck in Photoshop, or Excel, or the accounting app your job just switched to, and the help was never in the app. It was on YouTube, or in a blog post, or in a course you'd bought and left open in another tab. So you'd stop what you were doing, go find the help, watch someone do the thing on their screen, then come back and try to do it on yours.

Nobody planned it this way. It was just what was possible. In 2012 a program had no way to see what was on your screen. It couldn't tell whether you were in Photoshop or a spreadsheet, never mind which button you were reaching for. So help couldn't come to you. You had to go to the help.

Once you notice that constraint, you start seeing it everywhere. It's the shape of nearly everything we still call learning.

A video is someone else's screen. You watch them click, in their version, with their files, and you do the translation in your head - their menu is where on mine? An article is worse, because now the screen has become a list of numbered steps and you're rebuilding the motion from a description of it. Even the newest thing, asking an AI in a chat window, has the same shape. You describe your screen to it in words, it answers in words, and you carry those words back across the gap to the real app and hope they still fit.

Every one of these makes you leave. You go and get the help, and then you do the hardest part - putting it back where it belongs - by yourself.

This is expensive in a way the receipt doesn't show. A course costs money, a hundred dollars and up, twenty hours and up, but the real cost is the trip. You break your own concentration, go off somewhere to learn the task in the abstract, and come back to a screen that has moved on without you. Most people never finish. Under one in ten finish an online course. I don't think that's because people are lazy; it's because the format asks you to learn the task everywhere except where the task is.

The constraint is gone now. Software can see your screen. It can tell which app you're in - and the average company runs about a hundred of them - and it can point at the actual button, the one in front of you, instead of a picture of a different one.

That small change turns the whole thing inside out. For twenty years the help lived somewhere else and you fetched it. Now it can come to where you already are and stay there while you work. You don't leave Photoshop to learn Photoshop. You do the real task, on your real file, and something stands beside you and points.

for twenty years the help lived somewhere else, and you were the one who had to carry it back.

It's worth seeing how narrow the old choices really were. A lesson was either fixed - a video or an article, made once and identical for everyone - or it was alive but somewhere else, a person or a chat you had to travel to. The one thing no one could give you was the combination that turns out to matter most: alive, and right here, on the screen you're actually looking at. That option was empty because it was impossible.

It isn't impossible anymore. That's the whole thing.

It's roughly what we're building HeyBraza to do: you hold a key, ask out loud in whatever app you're in, and it points at the next step on your own screen. It takes a beat to answer, and it's early. But the shape is right, and the shape was always the point. It competes with the course, not with the app.

The tell, when you look at how people learn software today, is that all of it still assumes you have to leave. Video, articles, chat - every one of them was built for a screen no program could see. That screen is gone. Most of what we call online learning is still talking to it.