If the brilliant brain researchers are right, with a powerful brain scanner and enough calculations, science is able to "take" pictures - and maybe one day even read thoughts - straight from your mind.
It's already dark, on a warm Monday night in April, I'm lying face up inside a 13-ton cylinder at the Henry H. Brain Imaging Center. Wheeler at the University of California at Berkeley. The room is dimly lit, and I am alone. A white plastic shell surrounds my face, and a blue computer screen shines in my eyes. I'm here because a neuroscientist named Jack Gallant claims he can read my mind. He gave me strict instructions not to move; Even the slightest movement can affect the accuracy of what it's going to do. As I stare straight up, I feel an itch on my thigh. Don't scratch, I tell myself. I try to silence my thoughts as the beeps become faster and the fMRI machine – the scanner that will detect the changes in blood flow in my brain – starts working.
Gallant assures me that the random thoughts going through my head will not affect his results. Today he is only interested in what I see and how it is absorbed by the visual cortex, an area at the back of the brain that processes what our eyes see. It doesn't matter if I'm thinking about what I'm going to have for dinner, or I'm worried about getting a parking ticket. The only important thing, he says, is that I keep as still as I can, and in a little while he will have enough information to recreate the pictures I was looking at without him seeing them himself.
For the past ten years, Gallant has directed the Neuroscience and Psychology Laboratory at Berkeley, which is dedicated to brain imaging and vision research. He is one of the few neuroscientists in the world who are on the verge of discovering the key to mind reading by analyzing brain patterns with the help of magnetic resonance scans and algorithms. By exposing a series of random images and evaluating and analyzing fMRI scans from my primary visual cortex, Gallant says his technique can reconstruct images stored in my brain. His current method requires hours of analysis, but his goal is to refine the technology to the point where it can infer what people are seeing in real time.
If it succeeds, the method could affect the way we do almost everything. Mind-reading machines could help doctors understand the inner worlds of people suffering from hallucinations, cognitive disabilities, post-traumatic stress disorders and other impairments. Judges will be able to use them to peer into the minds of suspects by requiring them to recreate the experience and read what they see in their minds. Such machines would also be able to decide whether someone using a defensive line of insanity is deceiving, or whether someone claiming self-defense is truly in fear of their life. On the other hand, the technology raises significant ethical concerns, with critics fearing that it could one day make our personal thoughts vulnerable to snoops and hackers.
I think about all these things as I lie motionless inside the brain scanner, staring straight ahead while Gallant and two of the researchers in his lab project several dozen images before my eyes, for a few seconds each. I see sheep grazing in a meadow, a rock formation, a lake and the profile of a guy who looks like Einstein. I'm not supposed to actually look at these pictures - my job is to stare at the white dot in the middle of the screen. "Seeing" does not occur entirely in the cognitive dimension, explains Gallant. The visual cortex works like a camera, and automatically receives information through the retina and registers the image in the brain.
The ten minutes feel like an eternity, but eventually the fMRI announces the end of its program with another loud beep. The investigators take me out of the cell and escort me to the control room, where a huge screen shows 30 scanned images of my brain from different angles. I see groups of squiggly white lines and V-shapes in light gray within rows of gray circles. "that's it? Is this my brain?” I ask, my head foggy from the great effort not to move. It surprises me that all the processes that take place in my brain can be reduced to a cluster of geometric shapes. Glanter tells me that brain activity is basically just a group of neurons that fire an estimated 300 million times in the primary visual cortex alone, according to the latest research.
To help decipher the shapes, the brain scanner divides them into a grid of cube-like XNUMXD structures each called a volume pixel, or voxel. To me, each voxel looks like a random mix of whites, grays, and blacks. But to Gallant's computer model, which can see more precise information within those hues, the voxels are a meaningful matrix of zeros and ones. By calculating this matrix, he is able to transform the shapes back into an incredibly accurate rendering of the guy who looks like Einstein or the sheep in the meadow. Gallant and his team didn't have enough time to generate enough scans of my brain to make their algorithm work, but they showed me compelling results from other volunteers. "It's not perfect," says Shinji Nishimoto, one of Gallant's postdoctoral students, "but we're getting pretty close."
When I leave the lab with my thoughts stored in my head, I feel a little uncomfortable with the knowledge that they may not stay stored for much longer. Gallant's "neural decoding" - a term he prefers over "mind reading" - is becoming faster and more sophisticated every day. In fact, last October, his lab was able to recreate entire video clips just by analyzing the brain patterns of people who watched them. In one example, a replayed video of an elephant walking in the desert showed a blur similar to Dumbo walking across the screen. The exact details are lost, but even so the translation is impressive for the fact that it was drawn from someone's mind. And it's not just Gallant that is moving forward. Using similar technology, other researchers are deciphering memories and dreams.
Beyond the nebulous dimension of the supernatural, mind reading may simply be a matter of having the right tools. "As long as we have good measurements of brain activity and good computer models of the brain," Gallant wrote in an appendix to an article he published in the journal Nature in 2008, "it should in principle be possible to decipher the visual content of mental processes such as dreams, memories, and images." .
The magic of the MRI
Gallant is a lean, muscular man with a horseshoe mustache and a Willy Wonka energy about him. He tends to use friendly and clear comparisons when he talks. "The mind is like a Thanksgiving turkey," he told me last summer during a visit to his modest office in Berkeley. He excitedly drew on a blackboard, trying to explain in simple terms the inner workings of the visual cortex. "The outer part of the turkey is the skin, or cortex. All the chicken folds inside are subcortical nuclei. This" - he tapped with his chalk on the huge balloon-like space at the back of his "Turkey" painting - "is the primary visual cortex, the center of our visual system.
The brain operates a complex assembly line to build the world around us. The primary visual cortex, or V1, connects to a maze of other brain areas called V2, V3, and so on. ("Nobody knows exactly how many areas are up there," says Gallant, pointing a finger above his head). Each area performs specific vision-related operations, such as differentiating between colors, distinguishing between shapes, determining depth of field, or sensing movement. When I look at a dog for example, I don't just see the shape of a four-legged animal; I recognize him as the brown and white dog I had when I was a child, hopping around familiarly in the backyard where I grew up. The picture may even trigger the memory of playing with him. Each of these aspects of "seeing" was represented by different patterns in the visual cortex.
The main function of V1 relevant to Gallant's research—the reception of visual stimuli—was discovered in the early 20th century, when soldiers with bullet wounds to the back of their heads, presumably in their visual cortex, experienced partial blindness even though their eyes were perfectly healthy. Experiments on rodents have confirmed that the location and shape of things we see are reproduced in V1. If I were to look at a tree, for example, the back of the eye would pick up an inverted tree representation on V1. But in the late 90s, when neuroscientists used a process called multivoxel pattern recognition, they were able to locate exactly these representations noninvasively in humans. The technique uses fMRI to map the visual cortex in tiny structures - voxels - that parallel the blood flow patterns. One pattern in the region responsible for shape, for example, could imply that the person is looking at a dog, while another pattern in the region responsible for color could imply that the dog is brown.
Gallant's project takes the technique to a new level, using a computer model that not only recognizes the figures but is also able to reproduce them. On the night of my fMRI scan, I met five team members from Gallant's lab who, for the past three years, have been wrestling with probability theory to arrive at the best algorithm for running the model. When I asked them exactly how they programmed the code, Thomas Nasleris, a tall, curly-haired postdoctoral fellow, wrote a long equation on the board called Bayes' Law. "This is the fundamental principle of probability theory that calculates how probabilities change in response to new information," he explained, and is the key to their technique.
To calculate the probability that patterns in someone's brain represent a particular character, the researchers have to first fit their special equation with a significant sample of data, and input 1,750 fMRI scans of the subject. "Bays' law tells you, for each possible image that the subject looks at, the probability that what we are constructing is the correct image," Naselris says. It's kind of like predicting the model of a car hidden under a tarp: to make an accurate guess, you first have to analyze all the available clues - the shape of the tarp, its size, maybe the type of person who owns the car, maybe the sound of the engine. The more information you have, the more informed your guess will be. Similarly, the more information you put into the equation, the more accurate its predictions will be.
dancing bears
Being able to pull an image from someone's mind is an impressive feat, but a far greater challenge is finding the way to understand the very thoughts associated with that image. Gallant has no way of knowing, for example, what I was thinking while I was lying in the scanner. This is because thoughts, unlike images, are not neatly recorded in the back of the brain.
So where are they recorded? Tom Mitchell, a computer scientist at Carnegie Mellon University, along with his colleague Marcel Just, are using fMRI and multivoxel pattern recognition to answer this question. Mitchell believes that by mapping the brain's response to images, words and emotions, his lab will be able to decipher thoughts, not just images, within ten years.
To pinpoint exactly where thoughts live in the brain, he put volunteers, during a recent study, into an fMRI machine, showed them two objects - a hammer and a house, for example - and used software to analyze voxel patterns that activated different parts of the brain, ultimately determining which object the research subject was thinking about. . Like Gallant, Mitchell can do it with 90 percent accuracy. "When you think of a hammer you think of all its aspects. You can think about swinging it, a thought that would fire neurons in the motor cortex," he says. "You can think about how it looks, which activates the visual cortex." His team also collected fMRI data from the amygdala region (a sensory center in the center of the brain) and the frontal cortex - areas associated with emotions such as anger and love - to map the brain patterns that are created when people hear words like "love", "justice" and "anxiety".
Yukyasu Kamitani, a computational neuroscientist at the International Institute for Advanced Telecommunications in Japan, believes he can take the technology further and decipher dreams. This summer, he plans to put people under fMRI to read their brain signals and, like Gallant, reproduce them.
Meanwhile, Gallant and Nishimoto try to restore movies stored in the brain. After I finished my fMRI scans, Gallant showed me a video clip on the computer that showed psychedelic bears flying in front of mountains. Every few seconds, a new bear came to the front and then floated like a ball thrown into the air. From time to time a colored cube passed by the bears. Just looking at the movie made me dizzy. "This is a motion-enhanced film," says Gallant excitedly. "It makes your visual system go completely crazy, so you get a lot of blood flow and signals."
Nishimoto, the lab's local "motion expert," can reconstruct the colors, position, and movement of these bears from brain scans, and produces reproductions of original video footage. In a similar experiment, he asked a volunteer to watch two hours of movie trailers inside an fMRI machine. So, the computer matched the brain patterns of the research subject with colors and moving shapes in the film. To build the computer model's reference library of associations - and to equip it with data - the researchers fed thousands of hours of YouTube videos and asked the model to predict how a person's brain would react to watching them. Then, when the research subject watched a new series of videos, the computer was able to match the new brain patterns to images in its library to assemble a reproduction of the original video clip. The reconstructed video captured the general flow of the movement, as well as shapes and colors, although it missed small details such as facial features. The resolution will improve, the researchers say, as more data is added to the computer model. "Every time I tell someone we can do it," says Gallant, "they say there's no way."
When I think of the rat's nest of lines created from my fMRI readings - all from looking at a simple black and white image - it's a little creepy to think that our thought processes can be reduced to binary code in this way. But then, the idea of a black box of neurons controlling everything we do and think is also creepy. "It's all about numbers," says Gallant. "The trick is to do proper accounting."
Lisa Katiyama is a freelance reporter living in San Francisco. This is her first article in Popular Science.
77 תגובות
I have a feeling he's just wrong. The attempt to see the thought by scanning - cannot work.
It is possible, perhaps, to compare it to trying to measure the speed of a car on highway 1 according to the shape of the headlight.
I am willing to risk and suggest a completely different way. But before that I will describe the state of Ploni at a random point in time (let's say 1-1-2015). And our so-and-so Bn *doesn't* know how to walk on a beam.
Now, I ask him to walk on the beam and after several days / weeks he walked on the beam, say, 2000 times.
Most likely, at this point our so-and-so controls the taps of walking on the beam much (much...) better than before (let's say 30-12-2014).
In other words: the brain of so-and-so, who now has a new skill - necessarily (!!!) is different from the same brain at the end of 2014.
In my opinion - you need to know how to check these gaps - and not by visual scanning.
Need to check something else.
(The experiment described in this article tries to test (electrical) currents by photography. Even if it succeeds, it will work really hard.)
{On the other hand… what do I understand by that…………………..}
What is?
People from this excavation?!?
Well of course it hurts him, what a question.
I suggest as a start to build something simpler
Like for example the brain of a mosquito or a cockroach
Then we can tell if it hurts when you crush it against the wall.
Just like Liza says (wait, isn't Liza a girl's name??) If the artificial brain behaves and learns and solves problems, and is just as creative as a biological brain, and claims to have self-awareness and emotions, then there is no reason to doubt that, because if it is, then The same doubt can be cast against any person who stands and speaks in front of you.
(I actually heard a senior neuroscientist claim that we are all basically robots, biological robots but robots, and I think he is quite right depending on how you look at it)
Continued:
And for a more serious response - I believe that if we have a robot that by any standard behaves and reacts like a person - then for all intents and purposes this robot is a person
Amadeus:
I say that it is enough to have a robot that is able to have a conversation like a person and understand what is being said to it
To say that he is able to hold a conversation like a person and understand what is being said to him
So you're basically saying that it's enough to have a robot that knows how to play the right voice recording at the right moment, to say that it's able to have a conversation like a person and understand what's being said to it?
Unlike some others here who philosophize a lot but generally talk nonsense.
Lisa, I really liked your comments, finally someone here who speaks to the matter.
To Amadeus:
you said:
"No, such a robot will not necessarily have experiences and mental content. I think that since the cognitive revolution and the fall of behaviorism, it's pretty much agreed that it's not enough to just have the right behaviors and reactions to be in a mental state, but something else is needed for that."
It really sounds very mystical to me
To Mr. Roschild:
"All atoms act according to exactly the same rule and are attracted in the same direction."
Now Tori Lamar said:
I find it hard to believe that you are serious about what you wrote
Do you want to fix? Or shall I?
In any case - I'm leaving the discussion because the last question addressed to me was too emotionally charged and I'm tired of emotional discussions.
Amadeus:
If so - it is not clear to me what you are asking and why specifically me (after all, between Lisa and me - I am the one who claims that the task is impossible).
Lisa:
I find it hard to believe that you are serious about what you wrote.
All atoms act according to exactly the same rule and are attracted in the same direction.
If the example you gave before was a mistake in orders of magnitude, then the current one is a mistake in the category.
With your permission Mr. Roschild, I would like to disagree with the relevance of the analogy you brought. It was equally possible to think that I would not be able to predict the movement of a ball that I drop from my hand. Think about the number of atoms in a sphere! Much, much more than a billion!!! How can we even approach the possibility of such a prediction?
Well, it turns out that it is possible - and the thing that makes this possible is the celebrated separation of levels of description - the separation of levels allows us in a computationally efficient way to describe and predict phenomena at different levels.
In addition, when trying to predict and decipher brain activity - this is done based on the study of many brains.
If we still use your analogy - this would be equivalent to researching several hundred separate Chinese countries that operate according to similar principles. It seems to me most likely to predict the results of elections with great accuracy in this situation (which, as you know, is not the situation in which pollsters stand when they conduct public opinion polls)
Lisa-
No, such a robot will not necessarily have experiences and mental content. I think since the cognitive revolution and the fall of behaviorism, it's pretty much agreed that it's not enough to just have the right behaviors and reactions to be in a mental state, but something else is needed for that. And so we will never know whether a robot has real mental states or not, until we know how these things happen in our brains. Only then can we try to copy it into a robot and claim that we have created a creature with awareness and all the other things.
Michael-
I really wasn't talking about qualia, I was talking about mental content, which is a completely different problem, which also seems to me to be related to what you are talking about with Lisa.
Amadeus:
I don't know exactly what you are asking.
If your question is about "how is qualia formed" - then I already said that I don't know, although in my opinion it is possible and even probable that in the end we will understand that we must identify it with a characteristic pattern of activity in the brain (after all, inside our brain there is no other person who gives meaning to the activity of the neurons).
If your question is practical - then the article here demonstrates how information about the subject's feelings is extracted from a measurement of the activity in his mind. There are quite a few successes in this regard, including lie detectors based on measuring brain activity.
And following the discussion I had with Lisa - I talked about analogies that are not relevant and I would like to say something about relevant analogies.
It is useful to give these examples a scale based on one of the analogies brought by Lisa - that of decoding speech.
Think about the question of what is speech.
Speech goes out into the air space as a series of combinations of sound waves at different intensities and different frequencies.
From a technical point of view - it is quite similar to the electrical activity of a single neuron.
Not only does it resemble a single neuron but it resembles a single neuron that we connected directly to.
Because it is speech that belongs to a certain language - the number of classes to which we must associate each segment of speech is not infinite - in total there is a finite number of letters, syllables, and even words in the dictionary, and our entire search is focused on determining the class to which the particular wave sequence belongs with the highest probability.
So let's still assume that it's a bit more complex than a single neuron and maybe - when we're discussing a single neuron - what's interesting is only if it's "on" or off (this isn't the case, but let's still assume that's all there is to a neuron).
In such a situation, the speech sound (on the order of magnitude of 30 letters) corresponds to the activity of 5 two-state neurons (2 to the power of 5 equals 32).
Now – compare that to the brain.
Here we are talking about a cacophony of about 10 billion neurons that talk at the same time and that we are not connected to any of them directly - and that do not speak the same language!
The so-called "Whoahed Cocktail Party" - one in which the number of participants is greater than the number of Chinese!
China is a very big country - so let's look at Israel.
We have an example of a problem that an entire country needs to make decisions about with the help of binary decisions of its individuals.
I'm talking, of course, about the election system.
This is a very simple problem overall because there is a finite number of parties and a finite number of seats in the Knesset.
See what efforts are invested in predicting the results of the elections (and I repeat and remind - these are far fewer results - many orders of magnitude less - than the number of thoughts that can go through a person's mind) and still - the successes are not impressive.
Now - back to China's order of magnitude.
And now add to that making a decision on the order of fractions of a second without being in contact with even a million of the Chinese.
And now we have moved from the order of magnitude of the number of possibilities of the election results to that of the number of thoughts that can go through a person's mind (and every person can, obviously, think among other things about each of the possible results of the election system and this still does not begin to scratch the number of thoughts that he can think Overall).
In other words - even if it were possible to predict the election results of all of China based on information coming from a million unidentified Chinese (who we did try to select from different regions, but we have no idea if we are talking to a mayor or a farmer) - which is impossible, and even if It was possible to do this in real time (and something that is impossible at all becomes even less possible - if there is such a thing - when it is necessary to do it in real time) - we would still have a problem that is much easier than reading minds.
Amadeus:
If we assume that the day will come when scientists will succeed in building a robot that behaves in the same way as a normal human. But in exactly the same way. He will even be able to describe his feelings while watching the sunset and how much he loves chocolate.
Will the problem then be less mysterious?
Lisa, I didn't say that there is anything in the brain beyond what we already know. I meant that mental content raises a conceptual, even mysterious problem, and I don't know how to even begin to answer it, or what kind of answer there might be.
Amadeus:
What is your claim? Are you claiming that there is something in our mind other than the material it is made of that gives meaning to experiences? Or is there something (that is not the brain itself) that looks at the neuronal activity and based on the reading of the electrical signals gives them meaning?
If you claim that the experience itself is something fundamentally different from the neural activity that accompanies it, there should be some way to distinguish between the two.
If they can be distinguished - how can this be done?
If not, aren't these two names for the same thing?
Let's assume for a moment that there is indeed a way to distinguish between the experience and the neural activity that accompanies it - how would our situation be different? If until now you wondered "how nerve cells can have original content", now the question changes to how nerve cells + the new thing (which is 'responsible' for the experience) can have original content... that is, you are left with the same questions - what would be the answer to these questions for you ? What will explain the experience or the "qualia" for you?
A question for Michael-
After all, if we assume that a certain cluster of nerve cells constitutes a thought about a dog, how can one decipher from it that it is a dog and not something else? How can nerve cells even represent an idea or an object in the world? It is true that, for example, certain areas on the hard disk of a computer can encode all kinds of information - text, images, etc., but it is all about numerical coding, which without a person to give it meaning, it will remain an abstract code and not an actual representation of something. So how can nerve cells have original content, and not just a formal code in the form of zeros and ones? Ultimately, when we think about any concepts and ideas, we don't think of them as codes that represent sensory input.
And I explained why. My opinion is not based on mere optimism or pessimism, nor on analogies that are not similar, but on reasons that none of those who claim otherwise have come up with.
Well - then my opinion is different and according to it there is no chance that we will ever reach a solution that is not human-oriented.
To Mr. Roschild:
You did understand my words as I meant them.
It is necessary to separate what can be done today and what will be possible in the future.
As I mentioned, the ability to read minds does require prior knowledge of the person (this is the level of tuning I'm talking about and nothing more). For example, a person who has never seen an elephant in his life will represent a new image of an elephant differently than a person who grew up in a jungle with elephants.
Regarding a baby that requires adjustment of the activation of its limbs - as mentioned, this also does not contradict my words - two babies who will grow up in a similar environment will adjust in the same way.
Likewise, the representation of a new idea or a new object must be separated from one that is common to many people (such as elephants), while the former may be more difficult for an external device to understand, the latter will be much simpler.
In conclusion, I will point out that indeed the task of identifying a person that is not adjusted is more difficult than an adjusted one, but this does not mean that it is impossible. I'll take speech recognition technology for example - you can build a speech recognition system that depends on the speaker (i.e. recognizes only words spoken by that speaker) quite easily - this currently requires a little tuning of the recognition system by saying a small number of words or pronunciations by the speaker . Developing a parallel system that will recognize speech independently of the speaker also requires tuning by a larger sample of people with a wider range of voices - and this system is even more complicated than the first - but in the end the system 'generalizes' the speech recognition task to new speakers as well. Today there are such systems that do not depend on the speaker and their performance is not bad at all. The challenge of including learning is indeed a great challenge for learning systems and today there is no competition for this ability in our brains (for comparison, toddlers learn to recognize speech independent of a speaker at a very early stage in their lives - even after hearing a relatively small number of speakers). But in our minds there is no magic that allows for such impressive speech recognition - this is an ability that developed with evolution - and there is no principled reason why a person cannot reproduce it through computerized technological means.
Lisa:
I repeat - neither the similarity in the way of operation nor the similarity in the behavioral output justify the claim that it is possible to decipher what is happening in the brain without tuning.
It is not entirely clear to me from your words if you think that this is indeed feasible (it seems to me that this is what you are claiming in the sentence with multiple negatives "I do not think that the scenario in which little knowledge of a person's personal background will not allow us to read a lot of his thoughts.") But if this is your opinion, then in my opinion it is In the wrong opinion that the reasons you gave do not support it.
I suggest you think for a moment about the following question:
Today - even for deciphering what is happening in the visual cortex which, as I mentioned, is the most technical cortex - the one that has an almost direct connection with the outside world - tuning is necessary.
This tuning is necessary to get data that is not available in any other way.
How, in your opinion, will it be possible to obtain this data in the future.
Consider also the following question - which also arises from what I have already said: why does the brain itself need self-adjustment? Why is it that even babies' control of their bodies is not perfect from day one?
This is a phenomenon that at one level or another exists in all creatures with a brain, but in humans it is very extreme and many of the abilities that are pre-wired in the minds of animals are learned actions in humans.
At the same time, it is also important how a device that has never heard of my grandmother can understand that I am thinking about my grandmother.
And in conclusion - think about the question of how a device that watches the mind of the thinker of an idea that has never been used before (invents a new device or discovers a new scientific theory, for example) can understand what it is about.
In my opinion, there is simply no theoretical possibility that any of this will happen.
Mr. Roschild:
In my opinion, understanding the operation of the brain means a **functional** understanding of its actions. Of course, a different content can be assigned to a certain function (and this is what is meant by the influence of personal history on the operation of the brain). Two brains are more similar to each other than a brain and a hand. In what sense are the two minds similar? in the manner of their actions, in the functions performed by different parts of the brain.
This is also what is meant by the separation of the levels I was talking about - such a separation of levels makes it possible to talk about the actions of a certain level without the need to talk about the lower levels (and the possibilities that are different from each other). Two minds can work differently on one level but be the same or similar on another. The extreme example is a language that enables the 'synchronization' of minds, even those that are different from each other.
Regarding the artificial neural networks: it is true that two networks can learn a similar task in a slightly different way - but this is not a fundamental - or functional difference (after all, they both learned a similar task - this is a functional similarity even if the parameters according to which the network was eventually adjusted are different). Here too two networks that have learned the same task will be functionally more similar than a structural network and a different learning model (each from a long list of models used for computer learning)
I don't think the scenario where little knowledge of a person's personal background would make it impossible to read much of their thoughts.
For that matter, I don't think elephants are represented so differently for two people with similar backgrounds (such as those who grew up in the city and their familiarity with elephants is from zoos and television). Of course, there are concepts that are represented in a more different way and there are some that are less
Mr. Roschild:
In my opinion, understanding the operation of the brain means a **functional** understanding of its actions. Of course, a different content can be assigned to a certain function (and this is what is meant by the influence of personal history on the operation of the brain). Two brains are more similar to each other than a brain and a hand. In what sense are the two minds similar? in the manner of their actions, in the functions performed by different parts of the brain.
This is also what is meant by the separation of the levels I was talking about - such a separation of levels makes it possible to talk about the actions of a certain level without the need to talk about the lower levels (and the possibilities that are different from each other). Two minds can work differently on one level but be the same or similar on another. The extreme example is a language that enables the 'synchronization' of minds, even those that are different from each other.
Regarding the artificial neural networks: it is true that two networks can learn a similar task in a slightly different way - but this is not a fundamental - or functional difference (after all, they both learned a similar task - this is a functional similarity even if the parameters according to which the network was eventually adjusted are different). Here too two networks that have learned the same task will be functionally more similar than a structural network and a different learning model (each from a long list of models used for computer learning)
Lisa:
It is true that language allows us to understand each other, but each of us creates the sentences in his own mind - a mind that is already tuned and in tune with his feelings.
Therefore, this does not indicate the possibility of penetrating our mind and understanding what is happening in it with universal tools that do not require tuning.
If we use the analogy we already talked about - that of a computer - the claim I made is that our brains are different computers - not identical!
This is a private case of the difference between the rest of our body parts.
The body is built according to dynamic construction principles that adapt themselves to the circumstances.
Even if you look at your two hands you will see that the vein systems in them are different from each other and certainly from the vein systems of others.
It is a physical difference that does not interfere with blood flow according to uniform principles, but this difference is essential when it comes to thought processes - on different computers.
Who built computerized neural networks - Neural networks - knows the subject from another aspect.
Machine neural networks are very similar to the brain in certain aspects and one of their properties is that two identical neural networks will learn to perform the same task in different ways!
This is a result of the fact that they adjust themselves according to the input they receive and the required output and the strength of the connection between the various components adjusts itself to meet expectations.
It is very similar to what happens in the brain.
There is no general way to understand what is going on in a neural network that performs a given task.
If you want to understand how the neural network does what it does, you must observe it - and only it - and try to give meaning to the strength of the connections it has chosen for itself based on its experience.
In my opinion - my previous responses on the subject - which demonstrate many aspects of the need for the mind to learn about itself and to know the body it is integrated into and the environment in which it must operate - make it clear why a universal tool that does not depend on personalization - for reading minds - is not possible.
Mr. Roschild:
I tend to agree with you that there are several levels of description for the processing processes in the brain. For example, if we look at the vision system, I assume that there is a level that corresponds to the raw input (the 'pixel' level), above it is a level of basic pattern recognition (such as vertical/balanced lines and different angles), then slightly more complex patterns, etc...
An interesting question is when parts of the brain responsible for processing other types of information (such as hearing or language processing) begin to 'intervene' in the process of processing the visual input.
In connection with the fundamental question of whether it will be possible to understand the processes taking place in the brain by reading the neural activity, I see no fundamental obstacle as to why this is not possible. This will of course require a functional understanding of the information processing processes. I assume that at the functional level it is possible to 'mapping' between brains (ie this part of the brain is responsible for this and that stage in visual signal processing). Exactly in the same way that such a mapping can be done between computers.
Of course, the translation of the letter into semantic content can depend on the person's personal history. As reading the content in a computer's memory will not necessarily be self-explanatory and will depend on the programs installed on the computer that read the same content.
The interesting point is that, despite the great difference between minds, the language makes it possible to do exactly the 'reading of minds' with varying degrees of accuracy. That is, every brain has processes that process the input - but in addition there are processes responsible for 'translating' the internal representation so that it is understandable to others (there are no internal modems).
For example, let's look at a person looking at an elephant - in his mind the vision processing system will be activated until the stage where the visual content is identified with the internal representation of the object 'elephant' in his mind. Later, when another person is met, the translation system will be activated that turns the representation of 'elephant' into a sound signal that will be understandable to the other person (who has not seen the elephant) until finally the same representation of 'elephant' is activated in the mind of the other person. (This is of course assuming the other person does have an 'elephant' representation). This is not fundamentally different for emotions.
That is, the language allows us exactly the reading of the spoken thoughts. If we can reproduce the translation processes that happen in a person's mind when they translate an internal representation into spoken language, we can also read their thoughts.
June:
I don't know why it doesn't work for you.
It works for me every time.
Amadeus:
I agree with the claim that the issue is still unclear.
I just wanted to point out the fact that identification between two phenomena such as a certain activity pattern in the brain and a certain state of mind or thought is not far-fetched and in other areas we do quite well with such identification.
The difference that I didn't want to point out earlier lies in the fact that we really know how accelerating the movement of molecules in the environment causes the mercury to rise or the water to dissolve, while since we don't know what a qualia is, it's difficult to link it to the activity pattern in the brain.
There is actually no serious problem with the desire. The problem is with things like experiences and semantic content. What I'm trying to say is that it might be possible to say that things like pain are identical to certain neural activity, but no one has yet provided an explanation of how this identity exists, similar to the explanation that exists regarding heat and the movement of particles. In the meantime, it is impossible to translate the world of mental phenomena into the world of physical phenomena, and just saying that the two things are the same is not enough to make a successful reduction (besides, no one has yet proven that the things are the same, it is always possible to explain any observation with the help of correlation).
Amadeus:
That's not what I'm claiming and I don't think that's the problem that bothers you.
Although today there are ways to indicate certain patterns of brain activity that can be identified with certain sensations and even if it is often quite "approximately" then other times it works really precisely (as in the cases described by Sanampolinsky).
In other words - regarding the feeling of "the desire to move the hand" one can identify exactly a nerve cell that if stimulated feels such a desire and even thinks it is free will.
The problem is that it doesn't put our mind to rest and I ask myself "why?"
Note that even in the link between the average kinetic energy of molecules and what we call "heat" there is a transition between two different worlds and you can quietly ask yourself "What is this? But how does the movement of the molecules increase the heat?" And not to agree with the identification that claims that the movement of molecules is the heat.
And yet - we have no problem with identifying the world of molecules and heat and we have a problem with identifying the electrical excitation of a nerve cell with the desire to move the hand.
I know there is a difference but I am deliberately not detailing it because I think it is important to think about the question.
Michael, what are you actually claiming? Is there today an explanation for the identity of pain = nerve stimulation in the same way that there is an explanation for the identity of heat = movement of particles? I do not think so. Perhaps in the future it will be possible, but it may be that this is impossible at all, or that the explanation should be sought instead in nerve blocks, in something else (and I do not claim that specifically in non-material spiritual objects).
Amadeus:
I agree with you that the matter is difficult to understand intuitively and perhaps to digest, but it is possible that the mental exercise I am doing for myself in this matter will work for you as well:
Think about heat.
I am not talking about the qualia of heat but the physical phenomenon of heat - the same phenomenon that causes the mercury in the thermometer to rise and the snow to melt.
This phenomenon has been known for many years. Long before modern physics.
And yet - today we have no problem identifying the heat of a substance with the average kinetic energy of the molecules (in the substance) or with the energy of electromagnetic waves.
I agree with Michael that there are two levels of calculation in the brain. The intuitive and aware.
When we throw a ball into the basket, we perform a very complex physical calculation in an intuitive way whose essence is not clear. However, if we want to build a device that will fall into the basket, we will have to use mathematical formulas and we do not yet have a way to apply the algorithm applied in the brain for a technological need. I'm also not sure that today it is possible to build a device that will compete with a person in shooting a basket while moving in three dimensions (including jumping).
To Michael - you are probably right, the feelings and all the other things are material phenomena, but the really interesting question is how this is possible, and how one can be explained by the other. There is a huge gap between the things that matter knows how to do, and the phenomena that happen in our inner psychological world. How do you explain this gap? Because the material remains dark for the time being, despite all the promises and speculations.
Michal I tried but for some reason the response is not sent (it is not even written that it is waiting for approval... the page simply refreshes and the response does not appear)
I remember that in the past I tried to copy such a link from the view source and even then it didn't work...
There is a difference between creating experiences and feelings in the mind and submitting them in practice.
The personal experience is the dark matter.
Without the personal experience it would indeed be another machine.
June:
Basically - you have to write in html
It's hard to demonstrate it directly because if I write what needs to be written to create a link…..a link will be created and again you won't see how I did it.
Here is a general tip:
Whenever there is something you want to learn how to do, use the view source option of the explorer to see exactly what is written on the page.
If you make it to the current page you can see the link I made.
Michal, how did you make a text link inside the message?
I also want to know.
You have to take all the excitement and celebrations and advertising noise a little out of proportion.
Because it is possible to sample different electrical signals from the human body. show him images while sampling, and then identify patterns in the sample that will match the images that will be shown to him again.
There is no particular advantage in sampling directly from the brain using MRI.
And there is nothing to fear that your thoughts will be stolen. Because in the same way it is very easy to test a lie detector with different methods, the MRI can also be tested
age:
And yet - the material is the one that feels and the material is the one that creates the emotions.
The fact that you claim the opposite does not constitute a counterweight to facts that are easy to notice (such as the ability to create experiences and emotions in the mind through material manipulations).
The facts, for example, are the ones that convinced Prof. Haim Sanampolinsky of the complete opposite of your words and as you will see in the link this, Haim Sompolinski was not supposed to be a person who was easy to convince on the matter.
In my opinion, he exaggerates a bit in the opposite direction, but for our discussion here it is not important.
Emotion is the dark stuff of the brain
It is linked and influenced by brain activity but is not part of it.
Matter can transmit signals and messages, it can store and process information
He can change and react, but he cannot feel.
Amadeus, not true.
It's not a qualia, it's a zirconia. Prove me wrong.
Leyar, an unsuccessful comparison. The XNUMXD of the computer is an illusion. There aren't really cars and people inside the computer, all the computer does is illuminate pixels of different colors on the screen at the right pace and in the right place. It's not a particularly impressive thing if you think about it, because it's not fundamentally different from those Las Vegas neon lights that create the illusion of movement by turning them on and off at the right time. The cars and people in XNUMXD are only in our heads, and all the real "magic" happens in our heads and not on the computer. So that explaining our experiences with an analogy to computer simulations is actually a circular explanation.
On the other hand, our experiences inside the head do not consist of intermittently illuminated pixels, but of very special psychological entities called qualia, the distance between them and a binary code is light years, and as I argued, no one to this day has been able to reduce them to any known physical, digital, mathematical phenomenon, And I doubt if it will happen in the near future or at all.
As for the robotic arms, it's certainly impressive, but it doesn't really prove anything, because we learned where there are electrical currents in the brain when we make this or that movement. This does not mean that these currents are the thing itself, but can only be an accompanying thing that always appears in correlation with the real phenomenon whose nature we do not yet know.
Year:
You are simply wrong and I explained exactly why.
It turns out that you didn't understand my explanation because I didn't talk about any mysticism in it.
On the contrary - of all the visitors of this site, it seems to me that I am the most determined fighter against mystical concepts.
If you read my words and understand them - I have no doubt that you will be convinced.
If you think I am talking about mysticism - I have no doubt that you have not understood them even partially.
Lisa:
I assume that the number of levels differs from field to field and even in a single field each level also has several parallel sub-levels.
For almost all the senses, there is a possibility of losing part of the function without losing the whole - that is - reaching the "higher" level without all the work in the levels below it having been carried out correctly.
It can be the loss of the link words in the language or the loss of the sense of movement in the sight.
There are also situations in which it is possible to discover that lower levels worked while the upper level failed (as in cases of blind sight).
Looking at learning, one can also distinguish at least two levels, one of which looks like a sequence of many sub-levels.
The sequence is "below" and it expresses "sensory" learning and the super level is "above" and it is expressed in declarative knowledge.
In order not to remain in the words of a pharmacy I will demonstrate this on the subject of cycling.
The sequence is the one in which a child who has never ridden a bicycle improves his ability in the field gradually until full control.
During all that time he is usually not really aware of his activity. His body learns to act instinctively.
The declarative level is reached when the child knows how to formulate rules by which he explains to someone else how to ride a bicycle.
The expressive level reaches perfection when the child knows how to program a robot that will maintain the balance of the bicycle.
You're just wrong and I'm not saying this with the intention of hurting or fighting with you, I'll try to think later how to explain it to you in a more logical and detailed way, but the bottom line is that there is no mysticism and no magic in the brain, it's all just physics, just like information stored in a computer , and because we are all human beings, our brains at the basic level encode information in the same way, and once we understand how this encoding works, we can carry out a reverse process and from the structure of an existing brain restore everything that is inside it, in the same way that we know how to restore information found on a hard disk even though Hard disk the information is different from that on other disks.
You don't need to be a neuroscientist with a professor's degree to understand such a simple principle like this, you are getting too caught up in the details instead of internalizing the principle.
Joseph:
Why do you direct your words to Ehud? He didn't say anything to contradict that.
Anyone who did say something that contradicts Yossi's words, my words, Ehud's words and the words of others, is Yair.
Year:
You accuse me of short-sightedness based on meaningless analogies.
There is a difference between predictions that are based on nothing and reasoned predictions.
Note that the experiment described here also confirms my words!
What is described as progress towards the goal that you, in your surprising myopia, think can be achieved, shows that, in fact, even at the level of the visual cortex, which is one of the most "technical" mechanisms and the furthest from consciousness, personal tuning is necessary to succeed in the task.
What it means?
It means exactly what I claimed, which is that different minds do the work differently.
Anyone who understands something about the structure of the brain - as well as anyone who just thinks about the function of the brain - could know this even without an experiment.
After all, the brain changes physically as a result of studying.
A brain that has learned one thing is different from a brain that has learned another.
Parts of the brain change their function routinely and this is very noticeable in the case of "ghost limbs" that are created following the amputation of a living organ. In such cases - the part of the brain that was used by the severed hand may begin to be used by the face.
It is so crazy that not only a computer but even the person himself does not know what he is feeling and when his face is touched he feels the touch of one of the fingers of the severed hand.
So tell me, Yair: if the person himself cannot decipher his feelings - how will a mechanized system do it?
In fact, every person is constantly studying his body and this study is important because changes are constantly taking place.
People who wore glasses with prisms that invert the image were confused at first but later they got used to it and saw normally. Then - when they removed the glasses - they were confused again and after another time they got used to it.
This also indicates the dynamism of the mind.
People must also make the same adjustment they made in the current experiment to the software.
Moving artificial limbs through thought - the very example you gave to support your point - actually disproves them.
In fact, it is about the brain learning to manipulate an extended body - one that no animal naturally has.
He learned it through a long process of trial and error and you expect a mechanized system to figure out by itself and without corrective feedback what is going on.
I'm sorry, but it seems strange to me.
Did you know that cats that were kept in a room with only vertical lines when they were young could not recognize horizontal lines when they were taken out of the room?
The opposite was also tried and the result was appropriate.
All the facts indicate that the mind itself must learn to decipher what it perceives.
This means (as can be expected from a system whose role is to be dynamic and to understand a changing environment) that the activity of the brain is not wired in advance, but rather it develops as a result of the personal experience of the person or the animal.
It is different from any other system in the body, most of its activity is already written in the genes.
The genes determine what the pancreas will do but they do not determine what the brain will think.
Extreme myopia is not the inability to see the future but the inability to see the present.
sympathetic:
It should be taken into account that the brain adapts itself easily to different situations. In all of these experiments there was actually an element of cooperation with the subject that actually helped the software identify patterns.
A neutral group of people who were not told anything before the test was not tested. It is also not feasible at all because the equipment is so cumbersome and there is no way to check without the cooperation of the subject. It should be remembered that this is equipment that takes up an entire room. And that the subject is put into a very narrow hole in a noisy machine, clearly not a comfortable place for the subject unless he mobilizes his good will to cooperate. If there was a way to create a control group like they do with drugs, the results would be very, very weak. From my personal experience, programs that pick up different physiological signals from subjects work well when there is cooperation. and give bad results when the subject shows no desire.
Although it depends on the type of signal and the way in which it is sampled.
Mr. Amadeus, I really don't agree with you, only the famous experiment from a while ago in which a monkey (and then a human) was able to move a robotic arm and feed itself using only thought only shows you how advanced the research on the subject is and that we have a practical and not just a theoretical ability to translate already today Some of the thoughts that pass through our brain into actual physical action in the real world, to say that our thoughts cannot be reduced to raw code (connections between the neurons, and the electric currents between them) is like seeing an amazing computer game in XNUMXD with magnificent colors and with shading and depth and action and saying that it is impossible to reduce it to electric currents that pass inside the electronic cards of the computer....
Apologies for all my typos in the previous message, it's probably too late.
"It's a little creepy to think that our thought processes can be reduced to binary code in this way"
Big mistake, thoughts cannot be reduced to binary code. The binary code represents patterns of blood flow in the tissues. But patterns of blood flow are not thoughts, just as the flow of water in a faucet is not a thought. Thought is a complex and mysterious thing on several levels than anything we know, and no scientist currently has any idea how it is possible to reduce thought to scientific or mathematical concepts (similar to how you can reduce, for example, water to interactions between H2O molecules).
You should be very skeptical about all the imaginary and crazy promises of neuroscience and artificial intelligence people, in the style that here in two years we will be able to copy our personality into a computer drive... In the meantime, none of them have really managed to even touch the real problems of the brain, the HARD PROBLEM of consciousness, of experiences , of contents, etc. This is not only the problem that we currently do not know for the time being how things work, but that we are not even able to point to even a possible theoretical explanation that can somehow begin to explain the ear. We simply have no idea what kind of processes from the whole world of physical phenomena we know are capable of creating experiential experiences or having semantic content. We only assume that the lump of flesh called the brain is responsible for all the wonderful things our MIND does, and that there is a correlation between such and such patterns in the MRI and certain functions, but no one is able to make a real reduction from one thing to another. Neuroscientists only know in detail which role is performed by which part of the brain, but they have no idea how it does it, or why this part and not the other - just like we know why kidneys filter the blood, and not the heart for example, but we don't know Why exactly is the visual lobe responsible for creating images, and the auditory lobe for creating sounds, we only know that it is so because that is what came out in the experiments, and not beyond that.
I suggest everyone watch this fascinating course on the brain:
http://www.youtube.com/view_play_list?p=D8C99F67C81778E8
(on behalf of the Hebrew University)
Mi-Kal, I think you are very wrong and in fact you quite surprise me with your myopia on the subject (your words remind me of many people in the past who claimed with great confidence that "machines heavier than air will never be able to take off into the air!" and other similar claims) our minds are indeed different from each other and depend on the experiences we've had in our lives and the things we've learned, but still their basic structure is the same, and once we manage to identify the way in which images and knowledge are physically encoded in the brain, there is no reason why we can't do the opposite and from the need for connections and electrical currents to deduce which image the subject was viewing or which Knowledge kept in his head.
Here is just one example of our progress on the subject:
http://www.tapuz.co.il/blog/ViewEntry.asp?EntryId=1399918
In any case, to say that we will never know what a person is thinking about or what picture they are looking at because every brain is individual is like saying that a doctor will never be able to identify what problem you are suffering from because every body is different and there are no 2 patients whose bodies are the same, there are thin people and there are Fat people, some are short and some are tall, there are those with a lot of tools as well and some with less tools as well, some are hairy and some are slippery... It's like saying that a car mechanic will never be able to fix a fault in your car because no two cars are completely identical to each other, every car has a different color, every car suffers from a different level of wear and tear, in every car the parts are a little different probably at the molecular level and so on... but a fact , as soon as a car mechanic or a doctor knows the basic structure of a particular vehicle or of a human being, they are able to treat a very wide variety of vehicles or people whose structure is such that they have never encountered it before, not even within the framework of medical and mechanical studies.
In short, what I want to say is that once you understand the basic structure of the brain and how it basically encodes information of any kind, then there is no reason in the world that you cannot perform reverse engineering and by observing the brain know what information a person has in his head, what memories he has and what image he has Look at a given moment.
Or as Ray Kurzweil says, as soon as we reach the age of nanotechnology (a topic that is starting to gain momentum in recent years), that is, within two decades or so, we will be able to inject tens of millions of nano robots into our bloodstream (in fact, every white blood cell is a nano robot created by evolution and fighting against bacteria and infections in the body) and as soon as these nano-robots enter the brain, they will be able to individually and precisely scan each and every neuron on all its connections, and all the currents that pass through it every moment, and transmit this information out to a computer that will build a three-dimensional model of the entire brain, then From that model it will be possible to extract any type of information that exists in the brain, everything including everything.
To some people it still seems like science fiction, to me it seems inevitable and I have no doubt that these things will happen faster than people expect or think (this is part of the trend of accelerated technological development, and our understanding of our bodies and minds is also advancing at a rate that is only gaining momentum in recent years) That is, within the next 25-30 years.
Probably everything that was considered, then and now, mysticism, is simply a kind of science that has not yet been revealed to us. Who would have believed that mind reading could become - even only potentially - an actual possibility. Well done.
Michael
In the computer analogy I brought, I did not ask to repeat your claims again, but to provide another reason. In my opinion there is also a problem in reading the signals from the brain because as far as I know signals can only be read from the outer cerebral cortex the depth of penetration of the magnetic fields from the brain out is low. But it may be enough to look at the outer shell from a neurological point of view. I agree that when it comes to consciousness, people are completely different machines and therefore there is a fundamental difficulty in reading minds. Regarding image identification, which is done almost without processing, it is possible, as stated in the article, to identify the objects a person is looking at, but for me, this is a step very close to an experiment where the information is read directly from the retina. Reading the information straight from the retina would not have surprised anyone (if it was possible to execute a program) the more the input goes through processing: objects are identified, adding an emotional charge comparing them to other objects that the person has seen thus making the task of reading more difficult.
Joseph
It is not clear how complicated the problem "knowing ourselves" is, therefore it is not clear if a quantum computer will be able to help solve it. An interesting fact is that Roger Penrose, one of the great mathematicians of the XNUMXth century, believes that there is a connection between human cognition and quantum mechanics (I personally do not think so), he wrote several articles on the subject and I think also a popular science book.
To all commenters:
The following questions are interesting:
If we continue with the computer-brain analogy that came up here, then in the computer there is a 'separation of description levels' of the programs running on the computer. I mean that when you ask 'What program is currently running on the computer?' This question can be roughly answered on several levels. The basic layer is that of the hardware and the electrical signals, above that is the representative layer of zeros and ones, above that is the assembly language, then you can talk in terms of slightly more 'high-level' languages such as C, above them there are even more high-level object-oriented languages such as C++, Java etc... This separation is extremely powerful (so much so that you can write a Java program without knowing the machine it will run on!).
And to the question: Do you think there is such a hierarchy of description languages in brain activity as well? That is, is there such a separation of levels in our minds as well? If so, what is your assessment of the number of description levels that exist in the brain?
Roughly speaking, there are certainly two levels - that of the neuronal activity and that of the thoughts or the language we speak...
When I wrote difficulty in reading and processing the same I meant decoding the signals.
I also think the same way as Michael about:
Michael: "I actually think that it is possible to decipher the software running on a computer if you measure what is happening in its electronics and know its structure and how it works."
Today there may be difficulty in reading and processing the signals that the brain produces, but when there is a broader understanding of how the brain works then learning algorithms will eventually be able to read thoughts, feelings, retrieve images and events from memory, etc. just like in science fiction movies and maybe even transmit information to the brain ( Like in the Matrix and load it with the XNUMXst, XNUMXnd, XNUMXrd grade, chess, flying airplanes, etc., although I'm quite skeptical about the fact that the brain will be able to process so much information in a relatively short time).
Now reading minds is an operation that seems like a distant dream, but in a good few decades it might already be possible to do it well.
sympathetic:
By the way, this is an intriguing question if the human mind is able to crack how it itself works.
Will a quantum computer in the future be able to crack such a question.
sympathetic:
A real breakthrough will be when they manage to find the algorithm that the brain uses to identify and present the visual information or any other information. No matter how sophisticated pattern recognition is, it doesn't help enough to decipher how the brain computes.
fresh:
The claim that math is just a language is one of the sillier claims that people won't let die.
I am ready to act as your interpreter for this language. Just tell me which word you want to translate between math and Hebrew (in whichever direction you like) and answer the following question:
It has been proven that for every group of N natural numbers between 1 and a thousand - if the least common multiple of every pair of numbers in the group is greater than a thousand - then the sum of the inverse of all the numbers in the group (the inverse of X is one divided by X) is less than 1.5
sympathetic:
My argument is different.
I actually think that it is possible to decipher the software running on a computer if you measure what is happening in its electronics and know its structure and how it works.
The truth is that there were already things from the past and the activity of systems was already deciphered using much less information.
The point is that if you make an analogy to a computer, then the situation is that the computers of different people - are different from each other - and therefore it is impossible to adapt one solution that will suit everyone.
This is the reason why they have to adjust the system for each individual subject (and the fact is that after the adjustment they do succeed in the task; currently only in the visual matter, but in principle one can think about generalizing the matter. The problem is of course that in order to adjust the system for a specific person in deeper areas of cognition it will be required , apparently, a period exceeding that person's lifetime)
Michael
For a change I completely agree with you. Many problems have been encountered with simple tasks of orders of magnitude, for example the recognition of human handwriting by a machine. For the sake of analogy, I do not believe that by the list of electromagnetic signals coming out of a computer it is possible to know what a computer is. The human brain is much, much more complex than a computer, even though it is believed to have different areas for processing information, for example the visual processing area.
Gallant's research deals with decoding vision and not reading minds.
If I understood correctly, they decipher the thoughts that the researchers themselves make the subjects think, through the images?
"It's all about numbers," says Gallant. "The trick is to do proper accounting"
I actually think that everything is physics, and mathematics is just a language. Everything is elementary particles that obey deterministic physical laws, which also create consciousness.
Friends:
I think you can remove worry from your heart that the futuristic part of the story will never come true.
Different brains perform similar tasks in different ways.
This is certainly true regarding the deeper layers of consciousness - layers where we even use different terms and the thought of my grandmother is not something that can pass through the mind of a person who has never known her but - despite the more external layers - those that deal with sensory input - the similarity between us is greater - still the difference Makes even the relatively simple task the researchers here tackled - that of image recognition - impossible without the cooperation of the person whose mind is being scanned.
Note that the facts from which I draw this conclusion are written in the paper itself: "To calculate the probability that patterns in someone's mind represent a certain character, the researchers need to first equip their special equation with a significant sample of data, and input 1,750 fMRI scans of the subject."
In other words - in order for the method to enable the decoding of the images that the person is viewing, he must first help himself to adjust the system specifically for him.
In the narrative part of the article this is expressed in the following sentence: "Galant and his team did not have enough time to create enough scans of my brain to make their algorithm work, but they showed me convincing results from other volunteers."
Again - in other words - the narrative part is not complete - the author did not tell us about the tuning process in this part - probably because she realized that it spoils the drama of the matter.
These reasons are also why I say that telepathy is not possible:
https://www.hayadan.org.il/answers-to-global-warming-deniers-1103103/#comment-263442
Roy, if the research is successful, you may see MRI on Key in the future...
Regarding the displacements, there are already today cameras with the possibility to correct displacements.
To read your mind, you have to be put in an MRI machine, and each scan takes a few seconds. It is enough that you move your head as much as possible so that the resulting images are blurry...
I really hope that the research will not succeed - because if it does succeed, it will be the absolute end of human privacy.
Think what governments and negative actors could do if they had the ability to read minds.
Think about what would happen if your neighbors/friends/family/boss could read your thoughts.
You should read 1984 Big Brother and the Thought Police...
There are a lot of hopes, beliefs, plans, etc., very few actual achievements in practice. I am afraid that the actual application of everything that these neuroscientists are trying to develop will encounter many practical difficulties along the way that will significantly delay progress in the field, and until humans have a practical ability to read minds and make practical use of everything described here, there is still a long way to go. But what can I say? I wish! safe trip.
http://www.youtube.com/view_play_list?p=D8C99F67C81778E8
Hello friend, I would love to hear your personal and honest opinion regarding the discussion that begins here starting with comment 3:
https://www.hayadan.org.il/protecting_our_privacy-3101102
Amazing. No words