> As an example: I'm autistic and I learn inside-out, building larger new concepts out of smaller existing ones; those with Asperger's on the other paw, learn outside-in instead, breaking down larger existing concepts into smaller new ones; both are part of the "autism spectrum", but differ very fundamentally.

This is a really interesting observation - can you expand on this a bit more, please? How did you first notice this distinction?

When, for example, learning a new concept in math or physics, what would outside-in look like vs inside out? Would you characterise neurotypical learning in one way or the other?

> This is a really interesting observation - can you expand on this a bit more, please? How did you first notice this distinction?

A year or two ago I interacted with someone with Asperger's, and since that very rarely happens I hit a sort of uncanny valley with the way they wrote to me, because their writing gave me a lot of neurotypical vibes, but at the same time seemed a lot more logical and structured than actual neurotypical writing. They seemed to be building their writing and ideas out of logical tokens and constructs in an evidently autistic way, but so much of their writing looked neurotypical.

It was sort of like Scratch blocks, where they learned general templates for sentences and paragraphs from neurotypicals, and then substituted entire phrases at a time within them to achieve the desired communication. So while their templates and phrases were learned mostly from neurotypicals, the structure of their communication and usage of it still had a sort of logical system to it.

They remembered concepts by the phrases that had been used to explain it to them in the past, and reused those phrases verbatim, slotting them into the templates wholesale. They didn't fully parse everything like I do, they only broke things down as needed to create new logical tokens for new concepts. It's very interesting and fascinating. I mostly only know the one experience I've had, since I didn't get to speak to them again, but I've started to see it absolutely everywhere since.

Plenty of engineering blogs, for example, share that sort of tokenized writing style, like for instance this one I recently noticed: https://www.righto.com/2024/12/this-die-photo-of-pentium-sho...

> When, for example, learning a new concept in math or physics, what would outside-in look like vs inside out?

I can only guess, because I haven't interviewed someone with Asperger's about this, but I know what it looks like for me and I can guess what it could look like for them (at least for the purposes of the outside-in analogy).

For me, when I learn a new concept in math or physics, I want to build an intuition about that concept so I can come up with strategies about it and involving it. This actually mostly does not consist of algorithms related to the concept, but rather more of a fundamental intuition about the nature of the concept itself; what shape it is, what holes it fills, what types and classes of things it can do or model, etc. With this I can come up with my own algorithms on the fly, and also know when the concept might be relevant.

An outside-in learner, however, would be different. I have very little experience with the learning process or even with the execution process, but it's likely that they would not bother to break the concept all the way down and try to fully fundamentally understand it. Rather they might look to learn examples of it, algorithms and procedures relating to it, and ways to prove that their approach is correct. It would be less of "understand all the intricacies of this thing and every property it could possibly have and integrate that with everything else" and more of "fill a mental knowledge base with examples, applications, procedures relating to this thing".

> Would you characterise neurotypical learning in one way or the other?

I would not. I don't mean to imply that neurotypical logic is inherently more simplistic or inferior to autistic logic, but to me it seems far more like rote learning. They don't necessarily seem to break things down to understand the smaller details, but they don't really commit full logical tokens to memory either. They seem somewhere in the squishy middle, where they can be good at doing something that they don't understand at all, just because they practiced a lot, and they can be terrible at doing something that they understand far better, just because they're not practiced.

I've found that given I understand something well enough, I can perform better on the first try than some others who practice. That's not really a reliable indicator of much, but for me a lot comes from similar sources of truth, whereas it seems for a neurotypical they don't always necessarily organize or canonicalize their learnings, so their knowledge can be completely disjoint in areas that could totally be combined.

This is all just my personal experience/opinion though, fwiw.