How autism may stem from problems with prediction | Hacker News Reader

How autism may stem from problems with prediction

spectrumnews.org

142 pointsakvadrako46 comments

In Ayaya’s telling, her autism involves a host of perceptual disconnects. For example, she feels in exquisite detail all the sensations that typical people readily identify as hunger, but she can’t piece them together. “It’s very hard for me to conclude I’m hungry,” she says. “I feel irritated, or I feel sad, or I feel something [is] wrong. This information is separated, not connected.” It takes her so long to realize she is hungry that she often feels faint and gets something to eat only after someone suggests it to her.

This sounds more like a failure to integrate sensory information than a predictive error. That also fits with existing theories of autism relating to sensory processing.

Isn't that sort of the same thing? To what end does one integrate sensory information other than to make predictions or classify things?

"Making predictions" sounds too broad. Everything humans do is basically pattern matching and prediction making.

The idea that there is a conflict of prediction vs. reality, and a simple failure to connect disparate sensations into a whole are nothing alike. The former is a characteristic of something like motion or VR sickness; your body expects one thing, but receives conflicting information. Integration failure doesn’t require prediction, just different sensations not creating a whole impression.

The author describes experiencing hunger as a bunch of disconnected sensations, that don’t add up to one thing. She feels pain in her stomach, irritability, tiredness, but doesn’t integrate that into “I’m hungry” and just experiences them as different incidental feelings indicative of nothing in particular.

It's the difference between knowing you're hungry and knowing you will be hungry. I can tell you that if I drive a car far enough it will eventually run out of gas, so I should probably go to the gas station before a big trip. But that's a very different problem from having a gas gauge that's broken entirely.

Speaking from personal experience with the exact conundrum described in the quote, I could not agree with your comment more.

There's a medical term for this — https://en.wikipedia.org/wiki/Alexithymia.

Alexithymia can overlap with ASD, but it can certainly occur in individuals who do not otherwise meet diagnostic criteria for ASD.

It's quite a different thing. Sensory information like the limits of your body, vision and sound input.

http://carlysvoice.com/home/aboutcarly/

I believe the woman above described that she does not feel the limits of her body and needs constant jerking and slamming of the body parts on the walls to find comfort.

When I'm using ADD meds, I have to set a clock or I forget to eat, because I don't feel hunger. I feel every other symptom surrounding hunger, like becoming light-headed due to low blood sugar, or getting more sluggish when working out problems. However, it all happens without feeling a desire to eat (which is what hunger is, no?)

I have to make a rational decision to act on the signals and put nutrients in my stomach. And I chose that description for a reason: once I do eat, I taste the food just like I normally would, but it doesn't trigger any emotional response. The satisfaction of eating something tasty, or filling up an empty stomach is absent. To be clear: the medication does not make me a robot, I still feel emotions in general at this stage and I do expect them to be there. Normally, not enjoying food that you like is basically the body's signal that you've eaten enough, just before feeling nauseous due to overeating. Except that I'm starving. So eating a meal is like forcing food down my throat. It also means that I have to guestimate the proper meal size, because the normal cues for having had enough food are absent.

My emotional workaround is focusing on the the fact that after eating, I will have energy again. So I try to find emotional joy in anticipating the consequence of eating a healthy meal rather than the taste of the meal itself.

Is that the same phenomenon, or something different?

Out of curiosity, can I ask which ADD meds you use? I find Dexedrine incredibly useful, and I expected it to lower hunger as a side effect, but didn't really find it did anything in that regard.

I'm not the person you asked, but I figured I'd add my two cents. I have tried two different medications, with very different results in the side-effects department.

The first one I tried for about a year was Adderall (the regular release one, not the XR). With Adderall I noticed significant loss of appetite, to the point that I had to actively remind myself to eat, like other people here have mentioned. It was interesting, because I would feel my stomach sending the "I'm hungry" signal, but it was trivial to ignore it. It wasn't until I would go 12 hours without eating and start to feel light headed that I would realize, "oh... I need to eat something." However, I still enjoyed food very much. Unlike someone else in these comments said, I didn't feel like I was merely ingesting sustenance vs. enjoying my meal. Whether a few hours or 12 since my last meal, as soon as I started eating I would think "oh wow, I was really hungry. and this is tasty."

Currently I am taking Vyvanse. All other things aside, one of the very noticeable differences between Vyvanse and Adderall (for me) is the loss of appetite. With the Vyvanse, although I am able to go without eating for much longer than without, I still get hungry. With no meds I might say I feel like I'm starving after 4 hours. With Vyvanse I might say that after 5-6, though I will usually eat before that, because it is definitely noticeable in a normal time frame. With Adderall, I had to actively remind myself to eat, otherwise I might get dizzy and light headed first.

I started out with methylphenidate, but that gave me extremely high blood pressure and anxiety. It also completely removed my appetite Since about half a year I'm on lisdexamphetamine and it works a lot better, except for the hunger bit (although that also is a bit better).

Also, I can't drink coffee, not even a single cup in the morning, or I will end up skipping an entire night of sleep. That part is supposedly fairly unusual for AD(H)D so I'm wondering if I might be slightly misdiagnosed or something

EDIT: adderall = ritalin, vivanse = lisdexamphetamine, so my story has similarities to awsedrftg except my appetite is affected differently.

Re: your edit, you meant methylphenidate = ritalin, of course.

Ooops... yes, thank you.

As someone who also has ADHD, I too have to remind myself to eat because stimulants supress my appetite.

I have read that people who have damage to their ventromedial prefrontal cortex can have greatly reduced ability to experience emotions but otherwise have normal cognitive functions. This can be incredibly debilitating because these people end up having to rely on “rational” thinking only and can take incredible amounts of time to make what seems to be a simple decision.

I wonder how much of hunger is an emotion, but regardless, having surprised hunger can be incredibly draining mentally because you are manually micromanaging something that should be handled “automatically.”

https://en.m.wikipedia.org/wiki/Ventromedial_prefrontal_cort...

> I wonder how much of hunger is an emotion

I don't know what the scientific definition of emotion is (wikipedia seems to suggest there is no real consensus), but I am inclined to believe that at least part of the hunger process is an emotion, as it calls one to action, and it involves a desire for food. Rational thinking, typically seen as a contrast to emotions, does not do involve either.

And to phrase myself differently: I've noticed that I do sense hunger, as in all the sensory aspects of hunger (an empty stomach, consequences of low-blood sugar), but I don't feel hunger (the discomfort of having an empty stomach or low blood sugar, the craving for food). The latter feels more like an emotional aspect to me.

Interesting link to the ventromedial prefrontal cortex - it's not like my experience, but it reminds me of a few people I know.

> This can be incredibly debilitating because these people end up having to rely on “rational” thinking only and can take incredible amounts of time to make what seems to be a simple decision.

This makes a lot of sense to me for a number of reasons. There is nothing in rational thought that immediately calls to action. Yes, you can use it to reason and figure out truths and what you really want to do, but underlying all that is an emotional foundation of what you want. Without emotion, without wanting, rationality can give truths without leading to conclusions of what to do with those truths. There also is probably less of a sense of urgency without emotions, so one can take their time to weigh all the options (even though there might not be any time).

This sounds like she needs to set up an eating schedule.

IMO,autism tends to effect high-level cognitive domains because those domains depends on a lot of neural machinery. The particular neurochemical pathways that disrupt those high-level cognitive domains will differ between individuals though, and the gene-environment interactions involved seems to be a total mishmash of mutations in genes and environmental factors regulating the brain...If more genes are affected relating to attention, we might give a diagnosis of ADHD. It IS a spectrum. So when a title says Autism stems from problems with prediction, I would rather say that prediction and balance between prediction and stimuli sensation are frequently disrupted because they are high level forms of cognition that can get disrupted in a number of ways.

I am still new to Autism research, but in my own work and others, I am seeing that ASD is related to a reduction of neuro-typical sex differences. However, I do not think I'd say that Autism stems from reduced sexual differentiation. I'd characterize it as a piece of the puzzle, and that disrupted neural machinery will likely disrupt the processes of sexual neuro-differentiation.

I try to be very humble about making claims about Autism, because I have begun to see how truly complicated and multivariate the disorder(s), an dmainly want to focus on my efforts to aspects of the disorder that affect quality of life (and certainly not extinguish all neurodiversity).

Exactly I too believe everything is being classified as Autism nowadays. Instead of focussing on improving the life for those who truly struggle.

Your first point is not quite what I meant, but the spectrum part of Autism is real, and so an expansion of what is diagnosed as Autism is probably justified until such time (hopefully soon!) where we can really fulfill the dream of identifying separable phenotypes of autism.

Perhaps then I respectfully disagree. Classifying everything as Autism means huge medical spend which could be used better.

Trying to classify every deviation from the narrowly definition "normal" is just lazy.

I believe you are over estimating the degree that the diagnosis has widened. Instead, awareness of autism is leading to more diagnoses. The instruments have not changed.

"everything is being classified as Autism" because autism diagnosis opens the door to insurance coverage for therapy and special-education services in the schools.

Yes and that's what I meant. I have Personally experienced this recently. Doctors wants to classify any deviation from their checklist as Autism.

And early behavioral intervention is one of the few things we know yields tangible benefits to quality of life outcomes

Could you give some evidence of how 'everything' is classified as autism nowadays? And could you also give some examples of how this prevents those who 'truly struggle' from improvement? In fact, could you explain what you mean with 'truly struggle'?

I've found that a pretty significant part of the struggle of people with mental health issues (even severe ones) is people who claim their struggles aren't 'real' (for some reason that always strikes me as either self-serving or somewhat arbitrary/based on pop-skepticism).

Autism is a serious condition and should be treated as such. But now as it is defined spectrum doctors muddle minor deviations and real serious issues together. In my view this does not help either as treatments are very different depending upon where exactly on the spectrum issue is. This gives rise to lazy labeling.

I am all for early intervention. Its rather that because now Autism is classified on a spectrum. Most doctors have got a lazy attitude of classifying any deviation from the normal as Autism. Instead of trying to go deeper they just start with default assumption of putting the patient on ASD spectrum.

- First born with minimum eye contact is Autism. Why it cannot be shyness?

- Doesn't pay attention unless Adult makes it interesting for the Child. Why it cannot be you are boring and I don't like you?

- Speech difficulty because of child is multilingual and mixes them two languages together. Please look at family background of the child first without declaring the judgement.

Doctors also have this issue of not crossing each other. Once one of them does a wrong diagnostic, others tends to agree with it rather than taking a contrarian view.

So, I can agree that broadening the concept of 'autism' can be detrimental, but mostly because it's seen as such a 'big' thing. I personally think it was maybe better to separate 'classical autism' from things like 'asperger's' as a label. But don't forget that these distinctions do exist in the actual field.

But everything else you're saying just isn't true in both my personal experience, that of couples I know with high-functioning children, and what I've found in my research of the matter.

I asked you for evidence (and personal experience does count, to me), but instead you're just listing/reiterating all the same things that I also keep hearing over and over from people who generally don't know what they're talking about.

Again, I do see some of your points, but for me these have become basically a 'meme' that is spread by people who are stuck with pop-psychology knowledge from sometimes decades ago. Stuff like 'ADD isn't a thing, it's just that boys need to be free' etc.

Not all untrue, but way too one-sided to be helpful or informative. It does make one feel like a clever contrarian though...

EDIT: To be clear, I'm not saying you're uninformed, but rather that I hear these things a lot from people who base their opinion on the pop-contrarian articles du jour.

See what I have listed is coming from my personal experience. I am not reiterating anything seen or heard anywhere just to be contrarian. I have struggled for more than a year with various doctors across two countries.

My thinking not against diagnostic but clumping a broad range of symptoms which affects life very differently is just lazy.

Definitely some interesting bits and pieces. I was recently diagnosed and the focus on predictive ability reminded me of a particular weakness I have: inability to follow plots. I can't figure out what's going to happen in most movies and usually I load up the Wikipedia entry to follow the plot if things aren't explicitly mentioned in the dialogue :-D

I work in film and visual art, and have always, odd as it is, associated my difficulty following film plots with my intrigue or indiosyncratic relationship to the art forms. I am also diagnosed. Film editing offers an interesting boundary of sorts in thinking about scope of perception or hierarchies of attention. I’ve always had a hard time just watching a film as a story experience but I find and appreciate plenty of elements about films to be plenty intrigued by almost any of them. I got very interested in media affect as a result, that is the experience of media intake.

I have come to believe my (seemingly) unique sensitivity to aesthetics is somewhat undermined by a lack of sensitivity to poetics. This conclusion was painful to reach, but so far the realization has paid off in the way of understanding the limits of my abilities. On a more related note, I tend to experience poetics from the outside in, which causes me to miss out on much of the human experience, including the way many people enjoy films.

This article reminded me a lot of my struggle with ADD. Noises that are repeating and usually filtered out by most people are the thing that keeps me from being able to focus on a single thing. I was never diagnosed with ASD, but I can commiserate.

Diagnosed with ADHD and ASD over here. My Dr. says there is some thinking that ADHD is potentially on the lighter side of the same scale of "thing" and the symptoms often exhibit together. I was diagnosed with ADHD first and the ASD thing came much later. It really helped put some of the pieces into place for me. I definitely see their relation.

I found this interesting. I’m diagonsed with ADD and carry all of the tell tale signs. My mind is geared to constantly fill in—like I need to over predict and guess what someone is going to say. When I watch a movie I am constantly hunting for the plot twist even if there is not one. I also have a higher than normal error rate because my mind tends to come to a resolution on scant evidence and I need to slow myself down and often double or triple check. I’ve often felt that autism was on the other end of the spectrum. Everything needed to be explicit, complete, and literal.

I relate strongly to your comment, specifically about the need to over-predict and be wrong. With my ADHD, I think my lack of a verbal 'filter' also contributes to that error rate, where my brain has a tendency to express the first thought that pops in it, rather than the best one :/

Came out of the other end of this article totally convinced im autistic. How much of this like hearing your own voice echo is part of everyday experience to lesser degree?

Or possibly, how problems with prediction may stem from autism.

That seams like a very good question.

Why would they not mention that in the article? seams odd

edit:

But on second thought...

If something is caused by autism it would be part of the definition for the behaviors that we call autism.

I should thank Michael Crichton for getting me in the habit of looking out for potential "Wet Streets Cause Rain" articles.

The article gets into that near the end, that maybe social issues make it harder to learn how to make predictions rather than the other way around.

At first I expected this to be something about the process of synaptic pruning, which doesn’t seem to work for people with autism and could be a big part of how autism works. I guess in retrospect this is only somewhat related to prediction in that the brain has to predict which pathways will be used and which won’t in the future, and eliminate the less used (and less valuable) ones. That’s really a key process to learning and associating things, which sort of ties into what’s described in this article.

Prediction is the one thing a AI excels at at this point. I wonder if someone could devise an apparatus to assist with such a disability.

Shit HN says.

http://n-gate.com/

Ha, I'd say we already have such an apparatus: a computer.

Throughout my life I've used computers in various ways to 'assist' me, but only in the past year or to have I been setting up 'systems' and tools specifically to help me with the things that are unusually challenging to me. So far it's already helped a bunch.

However, I think one problem of AI as a solution is that while the AI might do a decent job 'predicting' the needs of someone on the spectrum, the resulting unpredictability inherent in such an AI might make someone with ASD not want to use it.

https://en.wikipedia.org/wiki/Sally%E2%80%93Anne_test

Reply on news.ycombinator.com