[porknubbins]

Having worked with some brilliant people with autism, I argue that deficits in cognitive empathy/ inability understand intuitively what other people are thiking in real time is the hallmark of autism.

Moreso than anything about emotions, body language, social skills etc this is the most common trait. It pops up in odd places no matter how much you mask or learn the visible skills.

[exmadscientist]

I think the actual hallmark is one layer deeper: the brain's "input filter" is different and doesn't assign "important" versus "not important" distinctions in the same way as a "neurotypical" brain might. (In fact, the brain just might not be very good at input filtering, period.)

This shows up in a lot of ways: sensory (can't just filter out the tag on the back of your shirt), interests (can't recognize that you're spending just a liiiitle too much time on that model train), food (don't categorize trying new foods as interesting, so you don't), etc. But human socialization is incredibly complex, and while many of us can do decently well at learning it, imagine trying to get a grip on social cues without being able to tell what's relevant and what's not. Sounds pretty bad, doesn't it?

It also is a working theory that well matches my experience with the autistic people in my life. "Deficits in cognitive empathy/ inability understand intuitively what other people are thinking", aka "theory of mind", is a really good description of how things are often difficult, and I think "different stimulus input filter" is a pretty good hypothesis for how it might arise.

[shazbotter]

As an autistic person, what you say is correct to my experience. I have all sorts of sensory processing issues that all basically are variations on "I can't ignore signals".

[accidentallfact]

It's the other way round - let's say you have a thousand inputs, and those contain 1 bit/s in a way that you can't pick one or a few inputs that allow you to correctly decode that one bit of information.

Or let's say you have a million inputs, or pixels, and you need to determine if there is a cat in the picture. Selective attention won't work for that either. You can't pick six pixels that allow you to reliably answer this question.

You need dimensionality reduction, that will reduce the data into a manageable amount of abstract features, from which you can pick what features matter to you.

Neurotypical people lack (or only have remnants of) this second filter.

[accidentallfact]

No, the core difference is the level of abstraction, and the inability to communicate between people with different levels of abstraction.

The opposite of autism is schizophrenia, where abstract thinking fails completely, and the person is unable to find correct answers to everyday problems.

Neurotypicality is merely a socially acceptable level of schizophrenia.

Autism results when the level of abstraction is significantly higher than the surrounding society:

You can't automatically understand the concrete speech, and you especially can't understand the "implications" that rely on the concrete magical thinking.

People can't understand you, because their level of abstraction isn't sufficient to understand the actual meaning, so they assume you talk about something random.

People overread the gaze of more abstract thinkers, and underread the gaze of less abstract thinkers, due to the difference in the field of view.

Compare Taylor Swift (ultra concrete, easily understood by neurotypical people) vs Rihanna (very abstract)

[orthoxerox]

> The opposite of autism is schizophrenia, where abstract thinking fails completely, and the person is unable to find correct answers to everyday problems.

I've heard this one before, but expressed in causal fallacy terms.

Correlation doesn't imply causation, but it's often right to assume a causal relationship even though you are unaware of the underlying mechanism.

If your criteria for assuming causality are too lax, you are schizophrenic (seeing relationships that aren't there), if they are too strict, you are autistic (missing the obvious).

[sillywabbit]

Having worked with some brilliant people who don't have autism, I'd argue that their hallmark is deficits in cognitive empathy / inability to understand intuitively what autistic people are thinking in real time.

[sillywabbit]

Not even my theory: https://en.wikipedia.org/wiki/Double_empathy_problem

[lcnPylGDnU4H9OF]

I'm not sure how to describe this phenomenon where one person says $CLASS has $DEFICIENCY and they are respected and understood. Then someone else uses the same phrasing to reframe the statement to mean the same thing from a different perspective and that idea is rejected without explanation as to why.

Allistic people do fail to empathize with autistic people everyday. To suggest otherwise only ignorantly refutes my lived experience. Only gaslighting will convince autistic folk that they should disregard that. (And that happens everyday, sadly.)

[FollowingTheDao]

But as someone with Asperger’s I can tell you we have a lot of emotional empathy. In fact, we have too much.

[shazbotter]

Yep. Too much empathy can be a diagnostic signal for folks with autism. (These days Asperger's is now diagnosed as autism and not a separate diagnosis.)