3 HN users recommended this demo3 months2011–2025score 3.5
Years with an HN recommendation, 2007–2026
HN uses it constantly as the classic demonstration that attention is selective, in threads on driving, motorcycles, AI and perception.
The original invisible gorilla awareness test by Daniel Simons and Christopher Chabris, in which viewers count basketball passes and often miss a gorilla crossing the scene. It demonstrates inattentional blindness.
“a longer but more relevant "motorcycle invisibility training" video” (HN, 2019)
As a motorcyclist -- and lifelong learner w/ a degree in cognitive psychology -- I really appreciate demonstrations like this one^1 from Daniel Simons. It's a short video (< 90s), but for many first-time viewers, it's an eye-opening experience. ^1. [this video] Edit: just saw @elosarv's link to a longer but more relevant "motorcycle invisibility training" video ( [link]
Short demonstration that is eye-opening for first-time viewers
This is sound advice. The only time I've made contact with a moving car as a cyclist was when I was going past an off-ramp. The weird thing was that I thought the driver "saw me". It was late morning, no problem with visibility. From my POV, she was looking right at me. In fact, she was looking "through" me. She simply was expecting other cars and a person on a bike was not what she was looking for. Human attention is very much selective. Check out this shocking video: [this video] After that, I ride with a bright flashing headlight at all times, even daylight. It gives some measure of protection, I think.
Demonstrates selective attention relevant to drivers missing cyclists
It's called Selective Attention which explained in another comment: > Notably, people tend to see what they expect to see. > When they are looking for cars, they tend not to see motorcycles, even when they are in plain view. Watch/take this Selective Attention test (1 min 22 sec): [this video]
However, the argument that multitasking on computers is bad because humans can’t multitask is flawed. That was right in the article itself. I believe the author's point is that focus/attention is more important than being able to "multi-task". Refer to Daniel Simons's experiment (the invisible gorilla) on selective attention. [this video]
Selective attention experiment supporting focus over multitasking
I am not sure we even know that normal consciousness is continuous. We experience it that way, but that says nothing. We think our vision is continuous, too, but it isn't ( [this video] . More info at [link] (had to use a shortener because the URL gives information away for those who don't know this experiment))
Your video reminded me of this selective attention test: [this video] . This Wikipedia article also seems to be related to the phenomenon and talks about the video too: [link] .
> The notion of not knowing why recording sound could be useful to me is hard to grasp as someone who has always been around it. This seems like (from their point of view) the Blub paradox ( [link] ), and (from our point of view) something of its converse. From their point of view, they couldn't see the importance of sound recording because they didn't have any existing mechanisms to make use of it. From our point of view, we can't not see the importance of sound recording, because we have so many ways of making use of it. (Kind of like how, once you know the effect being demonstrated in [this video] , you not only no longer experience it, but find it hard to believe that you ever did.)
Analogy: once you know the effect you cannot unsee it
> You see a handful of lines and text across the screen and it's sufficient to make you go ":o"? Enough to conclude it must be better than a brain? I think he's referring to selective attention. Can't comment on the validity of this "study", but I recall it catching me off guard the first time I saw it. [this video]
Oh definitely. The selective attention test [0] really shows how humans can perceive certain things in a way that's different from how computers perceive them. The research on GANs also shows how computers can be fooled by things that wouldn't confuse humans. [0] [this video]
The moonwalking bear/gorilla comes to mind for this. Gorillas in our midst: sustained inattentional blindness for dynamic events - [this video] / [link] But I'm not sure what that means, is it good or bad to see the bear. You are driving not looking at pretty things. You want to reduce dimensions. It's a 1 tonne vehicle, color shouldn't matter.
Inattentional blindness study related to driver awareness
Yes. Eyetracking is rather useless as what we see depends on the task we have to perform. We filter out what we don't need and don't even register it consciously. As such, general eye tracking studies can only tell you so much... Case in point: dancing gorilla awareness test, [this video] .
This old psychology video comes to mind. [this video] Also, in San Francisco a lot of the "paint" in the crosswalks is little plastic pellets melted. I have had my rear wheel slide out going over the thick white crosswalk bars. Best to keep off the throttle and roll over them. Same for the giant metal plates they use to cover up ongoing construction projects. They can be very slippery.
Classic psychology video relevant to motorcycle invisibility
> it doesn't understand its own limitations, ie it has no self awareness. And how much self awareness do humans have, exactly? We walk around with heads filled with false memories[1], accidentally ignore surprisingly prominent stimuli[2], and struggle to understand things as simple as material color[3]. And people are mostly unaware of these limitations, or those examples would not be considered surprising! I think we plan around the limitations we observe, not our real limitations. If that's the case, then LLMs simply lack training data to notice those limitations. [1] [link] [2] [this video] (old selective attention test) [3] [link]
Yup, you can't move to them. Weirdly I didn't notice them at all until several attempts in (and was getting confused as to why certain moves weren't allowed), in a way that reminded me of the Selective Attention Test[0]. Once I'd clocked them I was able to solve it pretty quickly. [0] [this video]
A friend who lives in North Carolina sent me a video of the raging floodwaters in his state- at least that's what the superimposed text claimed it was. When I looked closer, it was clearly an Indian city filled with Indian people and Indian cars. He hadn't noticed anything except the flood water. It reminded me of that famous selective attention test video[1]. I won't ruin it for those who haven't seen it, but it's amazing what details we can miss when we aren't looking for them. I suspect this is made even worse when we're casually viewing videos in a disjointed way as on social media and we're not even giving one part of the video our full attention. 1. [this video]
Famous example of missing details when not looking
> maybe you shouldn't dress like that if you don't want the attention. Have you ever taken the selective attention test? [this video] (short 1:22 video)
Never made that claim, but dawn and dusk are the most risky. Mine happened on a lunch break, can't see through rock either. But you only see a fraction of the wildlife that is there. If your sole stratagy is to see them you may be surprised how many deer hit the sides of cars. Complacency and selective attention are very real human problems. The classic basketball game example of selective attention if you don't buy it. [this video]
Humans seem to me to be just as likely to make these kinds of errors in the general case. See the classic [this video] , which has an interesting parallel with this paper.
Classic human parallel to LLMs missing gorilla in data
It is true most humans can not track more than 5 objects concurrently. However, lidar data sets tend to produce sparse unambiguous point cloud data that is a lot more feasible to process. For example, a picture/reflection of a road on a bus is not a viable option with 3D data. Cheers =3 [this video]
Video list and Hacker News comment labels from HackerTube (prompted by @karpathy), released under ODC-By 1.0, built on the open-index/hacker-news archive. Hacker News data is provided by Y Combinator; comments remain the work of their authors. Video titles, channel names and thumbnails belong to their creators and are shown via YouTube. HackerTube Watch is an unofficial edition. Data export of 2026-10-31.
Plays via YouTube's privacy-enhanced embed
20 Hacker News comments
From the HN archive; up to 20 kept per video, favoring different commenters. Recommendations first.
Read more
Citations, criticism and jokes
Read more
Read more
Read more