By ChatGPT
This explainer is written to accompany and clarify Martin Jambon’s short note: The absurdity of asking what they learned (original note).
When someone asks “What did you learn?”, the question feels reasonable, even generous. It invites reflection, sharing, maybe even wisdom. And yet, the more seriously I take learning, the more absurd the question becomes.
I can summarize what I learned.
But that summary will not teach you what I learned.
At best, it transmits fragments—conclusions, labels, compressed results. What’s missing is the thing that actually mattered: the work.
Learning is not the list of facts you can recite afterward. It’s the internal reorganization that happens while struggling with confusion, noticing contradictions, testing ideas, and discarding bad models.
That process is inherently personal and temporal. It unfolds through:
When I tell you what I learned, I’m handing you the final state without the path that made it meaningful. You may understand the words. You may even agree with them. But you haven’t learned the thing in the same sense.
A summary is a compression. Compression works only when the decompressor already exists.
If you’ve done similar work before—if your mental structures are already aligned—then a short explanation can click immediately. It feels like learning, but it’s really recognition.
If you haven’t done that work, the summary becomes:
This is why advice often sounds trivial or useless to beginners and profound to experts. The words didn’t change; the listener did.
Contrast learning with discovery.
When something is discovered—say, a new theorem, a technique, or a tool—the labor can often be bypassed. Once known, it can be used without reenacting the original struggle.
You don’t need to rediscover calculus to compute derivatives. You don’t need to reinvent the transistor to build a computer.
Discovery creates external artifacts: results, methods, abstractions that survive their creator and transfer cleanly.
Learning does not.
Learning is an internal transformation. It doesn’t travel well.
So when someone asks, “What did you learn?”, they’re often committing a category error. They’re treating learning as if it were a discoverable object—something that can be packaged, shipped, and installed.
But learning isn’t the thing you end up with.
It’s the thing that happened to you.
The best I can do is point toward the terrain, describe some landmarks, maybe warn you where I got lost. If you walk it yourself, you may end up somewhere similar—or somewhere better.
But you’ll only learn it by doing the work.
Instead of asking “What did you learn?”, better questions might be:
Those questions don’t promise transfer. But they respect what learning actually is.
And that, at least, is something worth sharing.
This explainer was written by ChatGPT to accompany Martin Jambon’s original note.