@KatyElphinstone Also, to address the article's substance, that is, the thesis that autistic differences are significantly driven by #hypopriors used in unconscious Bayesian reasoning beneath the hood of conscious thought: that makes perfect sense to me. It even fits in with my own experience.
I don't have frequent problems with autistic literalness, but it does occasionally happen. I remember particularly one mortifying instance when I was in high school, talking with two of the teachers. One of them asked me whether I could see. We were watching some high school event, and no doubt a neurotypical student would have taken it for granted that in the teacher's mind the physical functioning of a student's visual system would never be in doubt. The neurotypical student's robust prior would have ruled out any interpretation of the teacher's question based on such a doubt, and that student would therefore have realized that the teacher was asking about external hindrances to vision, such as excessive distance, or intervening obstacles obstructing the line of sight.
But my autistic mind was working from a diffuse #hypoprior that gave more than negligible weight to the possibility that the teacher might be wondering about whether my own visual system was impaired — whether I had suddenly gone blind. I was puzzled as to why the teacher would be in any such doubt as to the functioning of my eyes, but I nevertheless took the possibility seriously enough to make it the basis of my answer, without considering other interpretations. Without thinking it out further, I answered that I could see, and expressed surprise that the teacher would be in any doubt of that.
The two teachers didn't know I was autistic, but they had had enough experience with me to know that I wasn't "normal". One of them was baffled by my answer. But the other understood, laughed, and explained (correctly) the mistake I had made. He thought it hilarious that I would have understood the question as I did.
I suspect that many other instances of autistic #literalness have similar causes. Of course the #mindblindness crowd jumps to the demeaning conclusion that we come up with literal interpretations because we can't even understand nonliteralness. But I understood the teacher's explanation of my mistake perfectly well — well enough to be embarrassed to the point of mortification. We don't arrive at literal interpretations because we have any problem whatsoever in understanding nonliteral ones. It's just that our diffuse #hypopriors don't rule out literal interpretations, and thus force us to consider nonliteral ones, nearly as often as the robust, narrowly focused priors of neurotypicals would.