Slashdot Mirror
Computer Game Reveals 'Space-Time' Neurons In the Eye
sciencehabit (1205606) writes news that the EyeWire project from MIT has yielded some exciting results. "You open the overstuffed kitchen cabinet and a drinking glass tumbles out. With a ninjalike reflex, you snatch it before it shatters on the floor, as if the movement of the object were being tracked before the information even reached your brain. According to one idea of how the circuitry of the eye processes visual data, that is literally what happens. Now, a deep anatomical study of a mouse retina — carried out by 120,000 members of the public — is bringing scientists a step closer to confirming the hypothesis."
The paper (paywalled), and a gallery of screenshots of the game.
gotta have 'em.
After looking at the screenshots. It blows my mind that that can compute. Who's crazy idea was it to build a computer out of jello and solutions?
The retina and optic nerve are very complex and dense. The almost certainly perform some level of preprocessing, and as such are really just an extension of the brain. However to say that you react before -any- info reaches the brain smacks of a physical impossibility as the brain has to receive some sort of data to trigger motor action. Unless the motor nerves are connected directly to the eyeballs.
Silence is a state of mime.
The eyes are actually part of the brain and have computational and comparative circuits built into them. This has been known for many decades.
Take a nasally voice, add lots of extraneous noise, and you have the walkthrough from Hell
The "space-time" intro at first made me think they discovered quantum sensors in the eyes that detect action slightly before it happens using parallel universes or the like. But they are just talking about motion-sensing pre-processing by the retina itself.
Disappointment. I wanted the ability to walk into my boss's office and say, "Before you get up to fire me, I quit!"
Table-ized A.I.
I had a friend and he was one of those friends who would always get me into trouble when I was a kid. He was four years older than me... and this was in grade 9 for me. He was the kind of dude that would just throw something at you and yell your name last second. It got so that I had developed Jedi reflexes around this kid. Something told me exactly what to expect. One day the bastard throws a big knotted wooden log towards my head, and calls it out last second as it's about to hit my face.
Without any hesitation I caught it!! About 45-55lbs, which isn't that much -- but it's a hell of a lot to catch without warning.
My point is that there is probably some kind of zone of effect to this type of thing where in a kitchen for example you could expect that a plate or glass might get knocked off the counter so you would be queued up to catch something whenever you enter the kitchen.
The dangers of knowledge trigger emotional distress in human beings.
are real.
Happiness in intelligent people is the rarest thing I know.
Ernest Hemingway
Eye, for one, welcome our new ninja mouse retina overlords
While this finding is interesting it's not the first time neuroscientists have found complicated functions being performed in the retina of non-primates; the extrapolation in the summary to implications on human vision is a bit of stretch. Mice have poor high-frequency vision, but they can sort of make up for it with vision that's sensitive to motion. Many others mammal have this feature as well (rabbits, cats, etc).
Our brains don't solve complex simultaneous differential equations. we use a more half-assed but faster approach. Faster means we have some chance in hell of coming up with a reaction in time.
Now, the basics of the approach is merely successive approximations. We start with a guess. If it's something we do a lot, like a ninja catching an arrow out of mid air, then their "first guess" is going to be pretty good, and their approximations will be even better, meaning they might just do it. If you've never played baseball before and you have to catch a high fly ball before it bonks you off the forehead, well, since you have no practice, your first guess at the motion of the ball will likely be wrong, and it may take you too long to figure it out, so you won't catch it. instead you will go home with a lump in your head. oops.
Getting the correct answer, too late, is no good. Not when some of the problems that might happen, is a Tiger pouncing on you. Getting a "good enough for now" answer fast enough to start making descisions, is often better.
We don't need time travel to catch the falling glass. We need enough practice with falling objects in general, to start moving our hand in the right general direction, and then adjust its course mid-way as we get a better idea of where it's going.
This story brings a couple of quotes from the movie Enter the Dragon: 1) To a student he's instructing - "Don't think, feel" 2) Holding up his fist - "It hits all by itself"
A summary / cut&paste of the information in the articles:
"In 1964, scientists showed that some neurons in the retina fire up only in response to motion. These detectors have so-called direction selectivity, each one sensitive to objects moving in different directions. No one knew the fine-grained anatomical detail about how the neurons in the retina are wired up to each other. We have no computer model that can figure it out. Only humans have good enough spatial reasoning to trace out the borders of all the neurons to map their circuitry."
They knew that our eyes contain motion tracking receptors that have their own path to the brain. They had lots of maps but couldn't get a computer to find the way so they asked a bunch of people to do it instead. Now they have a better understanding of how, where and why impulses are sent from those receptors.
"space-time", come on. Mixing neuroscience with terms most frequently used in physics to get people excited. Tsk, tsk.
The first time I experienced this affect was a few year ago while I was walking in some local trails. All of a sudden my body ducked and only after did I realise that I was about to hit a low hanging tree branch. Our mind is living in the past.
ayottesoftware.com
Now we know why even if it's not preventable, time slows down when "something big" is about to happen (eg, you're in a car crash)...
Yet I can't see at that rate all the time. I wonder if this has persistence / jitter implications for the Oculus Rift?
One of the more interesting explanations that I've heard for the so-called slow motion effect that seems to happen during periods of high stress or anxiety is actually just the conscious mind's interpretation of the relationship between the amount of new information being recalled by the act of remembering something and the passage of time.
The brain has a way of focusing a larger amount of attention to anything that is novel or different, and in certain types of "big" situations, the circumstances can be extremely novel, warranting a high level of attention to immediate detail. This effectively increases the quantity of new data that will be processed by the brain over that span of time. Your mind is accustomed to generally receiving such data at certain speeds through daily experience, however, and amounts that are in excess of that will cause you to recall the experience as taking proportionally more time than it actually did. In fact, it's really your memory fooling you about how much time actually passed, not how fast time actually seemed to take for you while it was happening. It's only when you try to recall the event that it will seem to have took so long. Although it will genuinely seem like you are remembering time seeming to slow down for you as the events were happening, this illusion is only caused by a skewed sense of time impressed into your memory by the amount of new information you were absorbing.
File under 'M' for 'Manic ranting'
I suspect this is a semi reflex action with small amounts of preemptive or situational processing. Possibly certain types of stimulus or environments cause a bypass of usual processing systems. Consider the same situation but it is a knife. The same reaction is rarely triggered, otherwise we would have a lot of Chefs missing fingers. You know which directions the knife and cups are positioned in a room and your brain knows beforehand that anything falling to the left of you from cupboards is likely safe , to the right... maybe let it drop.
Most of you have probably reacted to catch something unusual or out of place, something too heavy or dangerous and had to retract your hand quickly in response, this would explain those instances.
Oh dear God, I'm part bugbrain. I thought the bugs were not like us. Please, please kill me now.
This explains why my dreams always have 2-5sec "visions" of the future... I can't believe I'm a mouse... I am hairy enough though...
It easily could be that your mind is simulating everything a few moments into the future. Trying to anticipate the actions of others, your own actions, and the possible events that could occur. Occasionally it could find something's going to fall, but can't insert an action far enough ahead to prevent it from starting like tipping a cup. The result is you responding to a falling cup prior to you knowing it was going to fall.
I hope I am missing something, but it seems like:
Thanks for your help J.Q. Public!
If you are interested in the results, buy access here:
http://www.nature.com/nature/journal/vaop/ncurrent/full/nature13240.html#figures
AC CAPTCHA = "despise" ...'nuff said...
Visual information controlling physical action without conscious thought. Think of it as a higher level of autonomous nervous system.
Peter Watts wrote a very depressing novel involving the idea which explores the possibility that consciousness is not necessary for intelligent life, and, indeed, may ultimately turn out to be an evolutionary dead end...
Lawrence Person (lawrencepersonh@gmailh.com (remove all "h"s to mail)
http://www.lawrenceperson.com/
I was at PetSmart and a girl was restocking the cat food can shelf and bent over to get the next box. The previous one fell (a 24-can package! not exactly light) and she snagged the thing backhanded as it fell off the shelf. Apparently, ninja's work at PetSmart.
I was always of the opinion that this is just perception - it feels longer because more is happening.
That was until I was driving through unknown roads, at night, in horrendous driving rain on my own. I passed through a small town, and was way under the speed limit as I could see there was a pub (hence drunken people getting home in the rain was my thought) and yet the rain on my windscreen hindered my vision slightly even with the wipers on full. I was crawling along.
So I poodled through the town, out the other side, and up a tarmac incline. It was at this point, driving along, that my brain decided to supply me with information. I'd just passed a road sign on a pole, stuck at the bottom of the incline. It wasn't a usual one. I can remember in my head clearly debating whether I'd seen it correctly, and what it could mean. The debate went on for a while, foot still gently resting on the accelerator.
The sign was a red circle, inside which was a picture of a car, tipping over an edge into some wiggly lines. It's not one that I have cause to see very often. And I swear, in my head, for several tens of seconds, I was pondering the sign while the car was still moving.
It is, of course, a "harbour ahead" warning telling that you are about to plunge into the water. The ramp I was going up was not an incline, but a ramp into where (presumably) a ferry or similar would normally dock. Except there was no ferry.
I slammed my brakes on and was left with my headlights beaming out and catching only the top few feet of blue waves. And the waves must have been 15 feet high, and went on forever - they disappeared into absolute darkness as, obviously, there was no light ahead of me but the projected by my headlights. But I was that close on the incline that I could actually project light and see the waves ahead of me, sitting on the sheer vertical drop that would go into the water.
I cannot swim in a pool, I certainly cannot swim in a sea like that, from a car, in that amount of surprise. It was late at night, pitch black and I was actually outside the village. Any splash I made would be inaudible against the rain (in fact, nobody heard my screech of tyres at all). I was out on a drive after my wife and I had split but were still living together and she'd brought her parents around to visit, in order to disconnect for a while. To say that I would probably not have survived, and that people weren't likely to hunt too hard, and certainly not immediately, is an understatement.
In my rear view mirror, the view hadn't moved far enough to see the sign that had made me stop - that was still alongside my rear passenger window. Hence my braking distance in those "several tens of seconds" was approximately 6 feet, if that. In the wet. So I was poodling along, but still, the ramp was barely 10-15 feet long itself.
After what was, literally, ten minutes or so of very heavy breathing and watching the waves in front of me, I reversed the car slowly back (I remember having difficulty seeing as I hadn't put my rear wiper on), parked up, and got out into the torrential rain to look around. There was only one sign. The one I'd seen. There was no warning, no hint that there was even water nearby, no nautical theme at all in the whole village. Driving home, I kept my eye out for signs I might have missed in the rain, there were none, except for the one that made me stop.
But I swear, I can remember my entire thought process of - at one point - thinking how the sign must be wrong and someone was playing silly beggars, my highway code, all sorts, until the realisation hit me of what it was. It was then an awful long time before my brain supplied the notion "Well, stop then, you stupid bugger". In reality, my car would have been on the bottom of the ocean by the time I'd had all those thoughts if they'd happened as I remember them - and also how I remember thinking of them immediately afterwards.
My life didn't flash before my eyes, but there was time too. My brain mu
This study has shown, using a huge amount of money and effort, something that we've known since the 60s: neural circuits can detect motion if afferent projections along the axis of motion have increasing delays. It's a very simple idea. Say that neuron A is activated when neurons B-D are all co-active. Now say that B-D lie along a trajectory across the retina. If a bar of light falls across all three, then A is active. Now add a conduction delay to the projection from C and a larger delay from D. If a dot of light falls on B then C then D with the appropriate timing, then the signals will all reach A at the same time. A will be activated when a dot of light moves across the trajectory from B to D at the correct speed. Hence, A is a detector for this specific type of motion. Like I said, hypothesized in the 60s, called a Reichardt Detector, well-accepted in the field for decades, etc.. Google it. Also, complex calculations happen in the retina. This is known. There is a good review from 2010 by Gollisch and Meister published in Neuron called "Eye smarter than scientists believe: neural computations in circuits in the retina".
This is a Nature paper because the 1. study is the only known "success" for the trendy but controversial field of "connectomics" (trendy because "mapping the brain", controversial because it costs a lot and, even when it works, doesn't actually tell us much) 2. it uses the trendy methodology of "crowd-sourcing" to conduct the research. It doesn't teach us anything we don't already know. The approach doesn't scale (if it took years of work and for 120,000 "citizen scientists" to analyze enough imaging data to decode this simple circuit in the retina, how long would it take to do anything in the cortex? The answer is forever) and relies on a huge number of people being playing a game that is, by comparison to all of the other options, pretty boring. The approach also requires that function emerges directly from anatomy to have any chance of success (good luck using this tool to figure out the role of inhibition, neural modulators, etc.).
Sebastian Seung (the PI on this paper) is a smart and successful guy, but this is a triumph of web 2.0 hype, not an actual scientific advance. Evidence? Amy Robinson, the eyewire "creative director" whose job is to drive hype (by being cute and enthusiastic, social media, TEDx talks, etc) but not to develop technology or conduct science or analyze results, is an author.
So space-time neurons are the new midichlorians?
The study was "carried out by 120,000 members of the public" and the paper is paywalled?
The whole scientific publishing industry is a classic example of privatizing benefits and socializing costs. This is thoroughly corrupt, it is impeding science and human progress, and needs to be overthrown.
Perhaps that also explains why time seems to pass slower when tripping balls. Look at the clock after an hour, and you see it's only ten minutes later :D
Does this explain why, when my daughter dropped her plush animal toy last night I automatically grabbed for it, even though it meant I would spill my beer all over the place in order to catch something that couldn't possibly be damaged by a short fall?
The Quirkz Handbook of Self-Improvement for People Who Are Already Pretty Okay
Doesn't EVERYTHING (as we know it) exist in the space-time continuum? Now excuse me as my space-time motor neurons move my space-time legs so I can go get some space-time coffee.
The part of your brain in charge of making sense of the world and deriving "deeper meaning" is slower than the part of your brain that has pattern recognition, muscle memory and motion sensing.
So.... it doesn't matter that the actual signals to the brain are so fast... the reaosn you perceive it as being so fast is that the REST of your brain is so slow.
Instead of reacting before the info reaches your brain. I think your brain creates an assume trajectory and velocity of the falling object that you body then uses to catch the object before your visual cortex process all of the information. As someone who juggles with their feet and hands and plays ball based sports, that is my take.