Those "spatial reasoning" questions look a lot like the IQ test I took as a kid (my parent brought me to the psychology department to be tested for some reason). I really struggled with those and it wasn't until years later, talking with some other folks, that I realized I couldn't visualize mental rotations as well as most people. because of that, I invested heavily in 3d computer graphics, which helped tremendously when I worked in structural biology.
> Those "spatial reasoning" questions look a lot like the IQ test I took as a kid
When we took my boy for a WISC (children's IQ test) when he was ten, as part of a larger neuropsych test, he had enormous trouble with the spatial reasoning parts. The funny thing was that the test wouldn't score, as he was getting less than 5% on some of sections and more than 95% on others. This is a pattern that is apparently unlikely enough that the test assumes you aren't taking it seriously and gives up. So his IQ result was effectively "NaN". Based on that and other parts of the neuropsych, he was informally diagnosed with Non-Verbal Learning Disability (informally because NVLD isn't currently in the DSM).
While the IQ test was useless for him, thankfully the diagnosis was enough to get him into a school who understood his learning style and he went from completely failing out of school in sixth grade to now just getting accepted to a highly competitive college in NYC.
> I realized I couldn't visualize mental rotations as well as most people
I always struggled with those questions as well. Only decades later, when I realized I have aphantasia [0], did that, just like many other things, suddenly make sense.
For what it's worth, I have perhaps the opposite[1], unless imagining a conversation or reading/writing I think almost entirely in mental pictures and the manipulation thereof. It's part of the reason my memory is so effective, and verbal reasoning confounds me to some degree. Cognitive behavioral therapy is definitely not set up for someone with no internal monologue, only an an internal "visiologue."
PS "The vividness of mental imagery has a key role in the development and continuation of intrusive memories," ouch, does it ever. Every embarrassing moment I've ever had is right there at hand.
I do indeed find that interesting. Checking the paper, it does make some sense to me. It’s similar to what I said in another comment [0], the people with aphantasia do a slower, more methodical approach. It then fails for me when there are too many steps required, which is not the case for simple rotations as in the paper.
I was decent at the simple ones, where I could essentially map them logically. As long as it was simple enough to keep every step in my brain kinda "as text", I could solve them. Those in school were simple like that.
But IQ tests have more complicated ones, and there I’d not manage to keep everything in mind.
I think I have kind of aphantasia as well, but I usually score well in those, but what I think is going on is that I solve everything kind of like an LLM would, with my internal monologue I will try to come up with the first simple idea, and then count the objects to see if it fits that idea, and then if not (likely not if it's a tricky one), I will start to bruteforce more ideas and then count the matching with my internal monologue rather than imagining anything. How are you at math, e.g. calculations in your head?
All of it is much easier of course when you can write things down, or make notes.
I'm not truly sure if I have aphantasia, but I do think I don't get as vivid imagery in my head considering what other people claim to see. Although I always thought they were embellishing it. But for example when reading books I never found any use for scene etc descriptions, and they seemed just words, while I think other people got images in their head, and for this reason those descriptions were good.
Talking to Claude it thinks I don't have aphantasia, since I think I can see vague flashes of things. E.g. I try to recall my family member's face, I get what I think are vague flashes of the face. But it seems to require a lot of effort and definitely not voluntary based on text.
But I may be better at imagining some abstract content, like numbers. E.g. if I have to remember a 2FA code, sometimes I can keep visual imagery for 3 of the numbers and the other 3 I keep repeating in my head. Maybe text and numbers I can keep visualisation better, but not true 2d or 3d detailed imagery.
Out of curiousity can you imagine any numbers or abstract shapes at all?
It sounds like you are close to full aphantasia, but not quite. I usually describe it as having a memory of seeing something.
> Out of curiousity can you imagine any numbers or abstract shapes at all?
No, nothing at all, I have to (e.g. a 2FA code) repeat it to myself repeatedly, and then I’ll remember it by rhythm. Which is also how I remember e.g. months in a year, or the order of cardinal directions.
What if you look at a some sort of writing, e.g. 4 large letters and then close your eyes can you, can you at any capacity visualize them or.also absolutely nothing?
Not the guy you were asking, but it's like old tube tv when you switch it off. There is an after image for an fraction of a second at 2% opacity and then blank, nothing.
Ideas live like a non-visual, non-textual concepts more akin to feelings but not quite so.
I am good at maths, IT, programming and reasoning problems.
Bad at any sort of visual, mehanical problems. For example in a car engine I can not connect naturally how one part effects the other without knowing operational concept beforehand.
I also don't have any narrator voice in my head but my dreams can be extremely vivid.
I remember in meditation guides when they tell you to imagine you are on the beach and whatnot, I always assumed this is figure of speech.
I'm not sure what the opacity is for me, I think higher than 2%, but I wouldn't necessarily describe it as low opacity, it's more like very low frame rate of the actual image vs everything black. Or sometimes it's off color like grayish blueish instead of colorful.
I also have to focus hard. When I look at an object and then I close my eyes, if I put effort in I can see it blinking fast between black and the object and then black smoothly becoming more and more dominant, but it's longer than a second, and I guess how long depends on how much effort I put in.
I'm also terrible at real life 3d things and remembering how to navigate without GPS, or remembering places.
> I remember in meditation guides when they tell you to imagine you are on the beach and whatnot, I always assumed this is figure of speech.
I kind of also thought it was figure of speech, I do feel like I can get some sort of low imagery to black framerate going on if I imagine penguin in a snow cave, but it takes so much effort it doesn't feel pleasurable or worth doing. I never understood counting sheep either, thought it was some sort of joke.
I wonder how much miscommunication or misunderstanding there is between people, because everyone assumes they process things in the same way.
Also I wonder when saying disgusting, shocky or dark humor related things, it seems nothing to me since it's just words, but I wonder if it offends or shocks some people so much because they get the actual imagery.
I asked someone with hyperphantasia, and they said that they do indeed see images, but they also don’t bother them or are a positive even. Someone on the more normal side said they need to actively picture them, it’s not unbidden. And finally, someone else said that they can’t read about child sexual abuse without seeing it, which makes them avoid any such accounts.
> Also I wonder when saying disgusting, shocky or dark humor related things, it seems nothing to me since it's just words, but I wonder if it offends or shocks some people so much because they get the actual imagery.
Ohh, that is yet another thing I never thought of. Would be interesting to know.
Yeah, this is spot on. I just tried it again with 4 large letters I had conveniently in front of me, they were instantly gone, and I could only remember what they said, but have no visual.
I'm more of a wordcel than a shape rotator myself. It's not that I can't, it's more so that I get terribly bored when doing it. The last time I can remember being excited about it was in elementary school, where I was tested for spatial reasoning and put in some gifted mathematics class. Incidentally, my verbal skills were very poor at the time, I tended towards being very quiet and keeping to myself.
As I grew older and my verbal skills improved, I switched to the abstraction that comes with language. I also became more outgoing; I'm still an introvert but nowhere near what I was when I was a child. I suspect if you took a random Masaai hunter, they would do very well at these kinds of spatial reasoning tests.
Even in my career as a software engineer, I still tend to rubber duck a lot and really excelled at pair-programming. That's why I really enjoy using LLMs when programming now. When I transitioned to engineering management, I learned over the years to get out of the way of my directs who performed better alone; many times I wouldn't even meet with them for a month, and that made them extremely productive.
The issue with a lot of organizations is that they don't account for the variety in problem-solving techniques and abstractions that engineers use. Engineers are not some monolith that always need to be managed from the top-down; you should always play to their strengths and support a variety of problem-solving styles.
> many times I wouldn't even meet with them for a month, and that made them extremely productive.
That's an interesting idea. Since I feel like it's same for me. When I meet with people they have their ideas, suggestions and requirements that they put out and I feel like I have some sort of social obligation to meet them and eventually it might not fit very well into what I had planned and then I start to overthink related to how I can fit their ideas into what I was planning to do and it derails a lot. If I ignore the ideas I might need to justify it later or at some point, or argue immediately, but then so much energy will go towards justifying and debating, it will derail what I was to do anyway. It just I have this very strict big picture idea of what I'm planning to do in my head, and it feels impossible to share it with people without actually going through with it. Even if I wrote a massive document it wouldn't help I feel.
I've actually got feedback in my past that I should consider other's opinions more when doing something, but then it just breaks down everything I was to do, and I feel like I could execute really quickly.
I have trouble with group meetings as well, e.g. where people are going through tasks by splitting them down into subtasks, then discussing subtasks one by one, while I feel like I have a big picture thing, where I do not want to divide things into subtasks, since it's everything supposed to work together, and they can't be discussed individually. Although it's considered common best practice to divide and conquer, I feel like it's not how I best think. And I have idea for each subtask, but it's all coming from that "big picture" idea, and if I propose it I will have to start explaining the whole big picture I have in mind, and it would derail the whole thing, so I usually just let people discuss things out, because I can't explain the whole thing anyway. I can never be in sync with people in group meetings, so I'll just try to come up with things to pretend I'm being an active participant.
I'm obsessed with this big picture idea, and excited about executing it, but everything around it demotivates me and kills the excitement.
One thing to consider is that your code will be hard to maintain by others if you can’t at least communicate your big picture. And ideally, everything should be subdivided into modules with clearly defined interfaces, so that each module can be independently reasoned about based on (only) its interface definition. This increases maintainability because things can be reasoned about locally. People also differ in how much detailed context they can hold in their head at any given time, so the smaller the required context can be made for reasoning about any given piece of code or part of the design, the better.
Yeah, but I think my brain just doesn't work like that, I kind of have to work around that. I've never managed to change this with 10+ years. I do feel guilt around it all the time when working with others.
People with aphantasia are actually _better_ at mental rotation tests! Lack of visualization is no obstacle to the brain, which of course has all sorts of interesting implications.
Fascinating. Thank you for providing the paper. As someone with moderate aphantasia, it didn’t even occur to me that people performed “mental rotation” versus “analytic” strategies.
Looking up one of the phrases from Langdon's statistical report ("Four Sigma Members") led me briefly to the rabbit hole of high-IQ societies, of which there seem to be many.
Its not clear to me what the purpose of the high IQ societies Hoeflin founded was other than to give him a prominent role that he did not have at Mensa.
> I was also very shy and unable to put myself forward socially in Mensa groups. At the higher-IQ levels, however, I had the prominent role of editor and even founder
Its not my experience of Mensa which mostly operates in small groups. I just do not find people having a high IQ necessarily gives them a great deal in common.
I was just thinking recently about how much I hate the imposition of time limits on things like this! Although I'm not super interested in figuring out my own or others' IQ, these are a fun set of questions to mull over.
I have always disliked any kind of speed tests, because that parameter becomes the trivial and over-ruling influence on the outcome ... knee-jerk versus mulling. It discriminates against multiple sub-populations, e.g. the very learned, the aged, the ill, and those who hadn't enjoyed a good night's sleep.
I sympathize because I tend to be a slow thinker as well in some ways, but really, how fast one can solve a problem is a feature of how intelligent one is, and bad sleep or illness does decrease one’s ability to think intelligently (and not just in terms of speed).
Who's more intelligent, person A that would score 80% on a test given 30 minutes, and 85% given 60 minutes; or person B that would score 50% given 30 minutes, but 99% given 60 minutes?
Person A might be perceived as quick, have an accurate intuition, and able to come up with a lot of ideas. Person B on the other hand is of the slow and steady kind that will methodically and tirelessly work on a problem until it's solved.
I think this is getting close to the idea of liquid intelligence and crystallized intelligence. The scary thing about valuing the liquid kind is that it’s the one that shrinks with age. Crystallized intelligence grows with age.
Working memory size, processing speed, fluid intelligence (novel problem solving / creativity), and crystallized intelligence (knowledge), are are all different components of intelligence that are pairwise not perfectly correlated.
Some person A can solve problem X in 1 minute and problem Y in 1 hour, while some person B can solve both in 10 minutes, even with the same crystalized intelligence.
Not necessarily. If the problem was solved more easily if one possessed some specific amount of crystallized intelligence, one could say the "tardy" person turns out to have superior fluid intelligence.
Practice will clearly improve your performance on puzzle questions like these. Is that really testing intelligence, or just interest and motivation for solving puzzles?
It's also testing for social class, educational quality, and attention span. While mostly validating existing hierarchies. And giving loads of people a reason to feel good about themselves.
Don't forget the very real and horrifying links with eugenics and the use of IQ testing to justify sterilisation and worse! It's the gift that keeps on giving.
Not just that. Some IQ test puzzles that were difficult to understand when I encountered them "in the wild" went from bizarre to somewhere between trivial to doable when I read an explanation online of what they were even asking. My measured IQ would have taken quite a jump on just that one bit of information, were I taking those tests.
Without sarcasm or rancor, you can profitably debate whether or not that means that those questions were just too hard and I really was indicating that I wasn't smart enough for them. I get that, and my point isn't about my intelligence or if it is measured properly; whatever it is, it is, and you are welcome to conclude that I'm just a dufus, especially if you get them without that hint. My point is more that if the tests are supposed to be good measures of whatever intelligence is, it is rather debatable whether they should have that characteristic.
I do know I have never liked the "find the next number" tests; even before I learned about polynomial interpolation in school my protest has always been that any series of any numbers can be a pattern; there are an infinite number of patterns, and even constraining them down to the "human interesting" patterns isn't enough to nail down the patterns the IQ test writer is actually using. As a passing part of this post [1] I show a number of patterns in the OEIS that start with 1, 2, 4, 8, 16 and do not proceed to 32, each of which does not proceed to 32 with a different number, and I just stopped looking after I had enough for a blog post and already was filtering the results to "things you might actually recognize", and I probably could have kept going for a while. (Yes, I'm aware that a normal IQ test won't be looking for "divisors of 496" at the beginning parts... but towards the end of a high-IQ test...?) Just realizing that they tend to confine themselves to varying levels of Newton's calculus [2] is enough to unstick me, and, again, bump my score just because I know something rather than because of "intelligence".
(This also implies that the SAT proxy measure, even though it is not a direct IQ measurement, is likely to be more satisfactory, in the cases where the test takers have the relevant mathematical knowledge in advance. The SAT did not involve any problems that I had not encountered before, and to the extent that I passed and failed the various questions it can not be said it was due to lack of knowledge of what the questions even properly were on my part.)
http://www.lumifont.co.uk/omnitest/omniQ55.htm rewired knowledge of how loading a revolver works. If you don't spin between bullets, where does the 2nd bullet go? What if you spin, but it stops in the spot where the first bullet went?
I think if it stops at where the first bullet is you just put the second one next to it.
Btw, this question can be gambled a bit by using the fact that it has exactly one correct answer and going after easiest cases. This also makes the details of where you put the second bullet after a spin irrelevant provided that you don't always put it next to the first one.
Often I'm surpised and delighted when problems are accidentally constructed so that "this has exactly one correct answer" provides a lot of infromation, sometimes more than half of the actual question provides.
I don't think you need to spin to load multiple bullets into a revolver. You can put all the bullets in the cylinder once you've opened it with any need readjusting/spinning.
Why does it include Chinese "bons mots" [see below] in so many of its answers? That would be an annoying quirk even if it were right, but it's especially obnoxious when it's being smugly stupid.
