Mind Reading is one step further - neuroscientists can read words before they are spoken

[SvenSvensonov]

If this sounds a bit like mind reading, well, it is. A group of neuroscientists have figured out how to decode a limited set of words "spoken" by our inner voices from looking at brain activity alone.

The hard part, for now, is getting inside someone's brain to record directly from their neurons.A study published in the journal Frontiers of Neuroengineering peers inside the brains of 7 patients during epilepsy surgery (because it's not quite ethical to crack open someone's brain just for science). The patients read aloud a piece of text—the Gettysburg Address, Kennedy's inaugural address, or Humpty Dumpty—while electrodes recorded the activity in their brain. A special algorithm was developed to match up brain activity with the sounds of the spoken words.

Later, the patients were asked to read the text again, this time silently to themselves. Now, the algorithm was used to decode the brain activity, turning it into words. The algorithm was able to identify words in several of the volunteers.

EXPAND

The algorithm certainly wasn't perfect, and it could only work on the limited set of words that the patients were reading from. But this is pretty important step to developing a device that could one day help paralyzed patients speak again—or, more nefariously, one day read the minds of others. In related research, neuroscientist have also reconstructed images, albeit blurry ones, from brain activity alone. Someday, our private thoughts may not be so private anymore.

From This Brain Decoder Can Read the Words of Your Inner Monologue

My Comments:

My M.S. is in Behavioral Neuroscience (my doctoral studies are in biochemical engineering) and I can say with certainty that if we can read your words before you speak them, then by monitoring short nerve-to-nerve (not all signals pass through our brains, some often-used functions such as movement develop neural shortcuts... muscle memories are another example of such shortcuts) or brain-to-nerve signals we could read the actions of a person before they act. It's creepy for sure, but very, very interesting to me.

@Nihonjin1051 - as a fellow man of psych I thought you would be interested.
@Donatello

 

[Donatello]

Thank you for posting this. Indeed, it is very exciting and 'mind-blowing' at the same time. I was a student of Undergraduate Electrical Engineering (Power systems, motors, VFDs etc) but when the time came to do my thesis, i chose neural implants and as such had to do a great deal of literature review on neuroscience/bioelectronics. As the science and electronics mature, we have these new fields emerging, offering exciting new insights. However, at the same time, it is pretty scary. I mean someone can read your mind, what are the ethical/moral problems associated with it? You can have implants that control specific regions of your brain. You can alter you brain networks to work when you want and when you don't want. Imagine, having a bio-implant that boosts your neurons action potential, no need to take those daily Starbucks. No caffeine, no ill effects of tobacco/alcohol. We can create cyborgs within the current human form.

Big ethical dilemma there, no doubt.

 

[Aepsilons]

If this sounds a bit like mind reading, well, it is. A group of neuroscientists have figured out how to decode a limited set of words "spoken" by our inner voices from looking at brain activity alone.

The hard part, for now, is getting inside someone's brain to record directly from their neurons.A study published in the journal Frontiers of Neuroengineering peers inside the brains of 7 patients during epilepsy surgery (because it's not quite ethical to crack open someone's brain just for science). The patients read aloud a piece of text—the Gettysburg Address, Kennedy's inaugural address, or Humpty Dumpty—while electrodes recorded the activity in their brain. A special algorithm was developed to match up brain activity with the sounds of the spoken words.

Later, the patients were asked to read the text again, this time silently to themselves. Now, the algorithm was used to decode the brain activity, turning it into words. The algorithm was able to identify words in several of the volunteers.

EXPAND

The algorithm certainly wasn't perfect, and it could only work on the limited set of words that the patients were reading from. But this is pretty important step to developing a device that could one day help paralyzed patients speak again—or, more nefariously, one day read the minds of others. In related research, neuroscientist have also reconstructed images, albeit blurry ones, from brain activity alone. Someday, our private thoughts may not be so private anymore.

From This Brain Decoder Can Read the Words of Your Inner Monologue

My Comments:

My M.S. is in Behavioral Neuroscience (my doctoral studies are in biochemical engineering) and I can say with certainty that if we can read your words before you speak them, then by monitoring short nerve-to-nerve (not all signals pass through our brains, some often-used functions such as movement develop neural shortcuts... muscle memories are another example of such shortcuts) or brain-to-nerve signals we could read the actions of a person before they act. It's creepy for sure, but very, very interesting to me.

@Nihonjin1051 - as a fellow man of psych I thought you would be interested.
@Donatello

Very interesting developments indeed. You read into some research in developments of quantifying the unconscious? Really amazing read, too. Freud’s idea of unconscious has definitely been one that has been accepted and implemented within the school of psychology, sociology, neuroscience as well as even in the field of medicine, especially in psychiatry, which embraces many concepts taught in psychology. Its interesting to find out that the field of physics has been interested in charting the unconscious (Martin, Carminati and Carminati, 2013), as physicists try to unify all quantum fields of matter in a unique quantum field, Martin et al (2013) tried to unify the unconscious fields and describe above with this unique quantum field of matter and obtain a unique quantum field describing both matter and mind. The development of the Unconscious Theory Thought, which states that unconscious thoughts are unconstrained by cognitive capacity, thereby enabling people to make capacity-free choices when information can be aggregated unconsciously. And this has led to many in the field of psychology to argue if whether or not the Unconscious Thought Theory is a plausible model.

Reference:
Martin, F., Carminati, F., & Carminati, G. (2013). Quantum Information Theory Applied to Unconscious and Consciousness. Neuroquantology, 11(1), 16-33.

However, at the same time, it is pretty scary. I mean someone can read your mind, what are the ethical/moral problems associated with it?

Definitely very exciting to see how these new developments can unravel and chart new parameters. One thing that fellow behavioral scientists have discoursed about is the plausibility of preserving the unconscious , hypothetically, even after physical death. One of the new developments in Unconscious Thought Theory. One aspect of 'Immortality' that's worth pondering.

 

[somsak]

Very interesting.
Can the reconstructed spectrogram be represent in Sound? If so, do they sounds like those original words?

Last time I got an interesting news from Youtube. You may be interested in.
"Cat brain visual image reconstruction"

 

[somsak]

Have you ever ask a question: Are we robots?
or question: Is the brain a type of computer, which we can imitate?
Is there any scientific theory or hypothesis to answer these questions?

When I was young, I answer that "No, we cannot. Soul is real. There must be something unexplainable in our brain that is god's element. There must be an element in our brain that gives consciousness."
Then it goes through philosophy what consciousness is. I stop right there and never think more of it.

Currently there is a new branch of science called computational neuro-science. These people are using mathematical model to try to explain how the brain works. They are people who have access to both Math world, programming world, and Neuroscience world. As more and more result from them are out, now I start to believe that we are robots and that brains are computers.

 

[Developereo]

the patients were asked to read the text again

Not sure I buy it yet.

I would have been more convinced if they had chosen an unemotional text (are they detecting words or emotions?). Also, it is not clear that "detection" (even if we accept it) in one person would transfer to other people. So, it may be less a case of reading strangers' thoughts, and more of training a pattern recognition system (e.g. voice-recognition system) to recognize one's own characteristics.

the unconscious fields

Admittedly I haven't read the paper, but this sounds like pseudo-scientific crap dressed up in scientific lingo.

All they are doing is rephrasing the conventional wisdom that consciousness affects quantum events (in other words, we haven't a clue what's going on) by saying that there's a field, see, an unconscious field, that interacts with matter/energy at the quantum level. It's not saying anything new, just wrapping a mystical statement into equations.

Nobody knows what a "field" is -- whether it's electrical, magnetic or gravitational -- it is just a mathematical construct used to describe action at a distance. Quantum mechanics tries to get rid of fields and proposes that "field" interactions can be described as an interchange of photons between the participants.

 

[William Hung]

Have you ever ask a question: Are we robots?
or question: Is the brain a type of computer, which we can imitate?
Is there any scientific theory or hypothesis to answer these questions?

When I was young, I answer that "No, we cannot. Soul is real. There must be something unexplainable in our brain that is god's element. There must be an element in our brain that gives consciousness."
Then it goes through philosophy what consciousness is. I stop right there and never think more of it.

Currently there is a new branch of science called computational neuro-science. These people are using mathematical model to try to explain how the brain works. They are people who have access to both Math world, programming world, and Neuroscience world. As more and more result from them are out, now I start to believe that we are robots and that brains are computers.

I'm opposite to you. I used to think that a human brain is like a computer. But now I believe a human brain is not the same as a human mind and that a computer cannot imitate a human mind. I've come to this conclusion because of mathematics.

When you think about mathematical topics like formal system, provability, computability, completeness/incompleteness, you face really puzzling issues. Well, maybe I'm not a math major so I find it puzzling.

For example, how does mathematicians discover/come up with mathematical Axioms? These axioms does not come about through a mechanical process (i.e. step-by-step process) that a Turing-machine type of computer can replicate. Some say it's a special "intuition" that mathematical genius people possess. You cannot use a mathematic model or computer to replicate or explain this "intuition".

I'm not a math major, but maybe @Gauss can better explain this.

Also, computer software language are based on classical logic (logic calculus), but human do not use any form of formal logic that are derived from this classical logic. A human mind seems to use a logic system, if any at all, that logicians are still puzzled with. So all the computer system that we have today, does not work the same as a human mind.

And when people say "be logical" or "only logical arguments are good", they are actually wrong. A computer is "logical" but a human mind must not be logical.

 

[SvenSvensonov]

Not sure I buy it yet.

I would have been more convinced if they had chosen an unemotional text (are they detecting words or emotions?). Also, it is not clear that "detection" (even if we accept it) in one person would transfer to other people. So, it may be less a case of reading strangers' thoughts, and more of training a pattern recognition system (e.g. voice-recognition system) to recognize one's own characteristics.



Admittedly I haven't read the paper, but this sounds like pseudo-scientific crap dressed up in scientific lingo.

All they are doing is rephrasing the conventional wisdom that consciousness affects quantum events (in other words, we haven't a clue what's going on) by saying that there's a field, see, an unconscious field, that interacts with matter/energy at the quantum level. It's not saying anything new, just wrapping a mystical statement into equations.

Nobody knows what a "field" is -- whether it's electrical, magnetic or gravitational -- it is just a mathematical construct used to describe action at a distance. Quantum mechanics tries to get rid of fields and proposes that "field" interactions can be described as an interchange of photons between the participants.

This type of research and my former job in the USN have a lot in common. EMSEC involves pulling electronic information from the air and piecing it back together in a readable picture. Likewise, our brains are putting out enough electonic leakage that we can not only detect, but with the relevent tech, piece it back together. The problem was that we never could do cypher the picture we got. That is were algorithms come into play to predict behaviors. I did my Masters studies while still in the Navy and while still an EMSEC tech, this helped me a lot as I already had a great deal of experiene with electronic signals.

 

[Developereo]

This type of research and my former job in the USN have a lot in common. EMSEC involves pulling electronic information from the air and piecing it back together in a readable picture. Likewise, our brains are putting out enough electonic leakage that we can not only detect, but with the relevent tech, piece it back together. The problem was that we never could do cypher the picture we got. That is were algorithms cone into place to predict behaviors. I did my Masters studies while still in the Navy and while still an EMSEC tech, this helped me a lot as I already had a great deal of experiene with electronic signals.

The chemical and electrical mechanism of neuronal functioning have been known for a while and electrical signals from the brain have been mapped before in very general terms.

The researchers are detecting a particular pattern of brain activity and saying that the pattern corresponds to a specific word.
Does it?
Or does it simply indicate a particular emotional state evoked by that word in that context?
Would the same word in a different context produce the same pattern of electrical activity? The research doesn't seem to tell us.

 

[kbd-raaf]

Very interesting developments indeed. You read into some research in developments of quantifying the unconscious? Really amazing read, too. Freud’s idea of unconscious has definitely been one that has been accepted and implemented within the school of psychology, sociology, neuroscience as well as even in the field of medicine, especially in psychiatry, which embraces many concepts taught in psychology. Its interesting to find out that the field of physics has been interested in charting the unconscious (Martin, Carminati and Carminati, 2013), as physicists try to unify all quantum fields of matter in a unique quantum field, Martin et al (2013) tried to unify the unconscious fields and describe above with this unique quantum field of matter and obtain a unique quantum field describing both matter and mind. The development of the Unconscious Theory Thought, which states that unconscious thoughts are unconstrained by cognitive capacity, thereby enabling people to make capacity-free choices when information can be aggregated unconsciously. And this has led to many in the field of psychology to argue if whether or not the Unconscious Thought Theory is a plausible model.

Reference:
Martin, F., Carminati, F., & Carminati, G. (2013). Quantum Information Theory Applied to Unconscious and Consciousness. Neuroquantology, 11(1), 16-33.



Definitely very exciting to see how these new developments can unravel and chart new parameters. One thing that fellow behavioral scientists have discoursed about is the plausibility of preserving the unconscious , hypothetically, even after physical death. One of the new developments in Unconscious Thought Theory. One aspect of 'Immortality' that's worth pondering.

You've got to be the only guy on fora anywhere in the world that provides in text citations to peruse at our pedagogical leisure.

 

[Developereo]

You've got to be the only guy on fora anywhere in the world that provides in text citations to peruse at our pedagogical leisure.

There was another member @Wholegrain who was a top notch poster and provided references for his claims. Unfortunately, he got bullied off the forum by some members.

 

[kbd-raaf]

There was another member @Wholegrain who was a top notch poster and provided references for his claims. Unfortunately, he got bullied off the forum by some members.

Habits we should all take after, perhaps.

 

[SvenSvensonov]

The chemical and electrical mechanism of neuronal functioning have been known for a while and electrical signals from the brain have been mapped before in very general terms.

You seem to be confusing too different concepts. The electronic signals of the human body have been mapped many times before and for a relatively long time using EEG and EKG machines. The problem with these types of sensors is that they tell researchers what points of the brain (in the case of an EEG) are activating, but not what is actually happening. These signals tell the researcher if damage has occured, if there are deficiencies or excesses in the amount neural chemical being produce, etc., but not what is happening. The new studies being conducted (the OP) is not about detecting signals, but rather detecting signals and putting them together to determine what has happened in the brain. Not only do they now know that the speech center of the brain has activated, what an EEG would tell you, but they know exactly what words are being produced, something the typical mapping methods cannot do.

Also your skepticism is healthy, but might be misguided. Piecing together the signals received into readable pictures is maturing, slowly, but progress is being made. These types of studies are not new and I'll post a few links to past studies of the same nature.

Mind-reading research: the major breakthroughs - Telegraph

Mind-reading Technology Speeds Ahead - Scientific American

Method of recording brain activity could lead to 'mind-reading' devices, scientists say | News Center | Stanford Medicine

Neuroscience: The mind reader : Nature News & Comment

@Nihonjin1051 - have you come across any additional studies, have any info of your own or your own research, or conversed with anyone who has additional information?

 

[FairAndUnbiased]

Nobody knows what a "field" is -- whether it's electrical, magnetic or gravitational -- it is just a mathematical construct used to describe action at a distance. Quantum mechanics tries to get rid of fields and proposes that "field" interactions can be described as an interchange of photons between the participants.

Depends on what quantum mechanics you want to talk about. Regular quantum mechanics usually has quantized particles moving in a classical field. 90% of real world materials can be described by classical fields with either quantized or classical particles moving in them. Only when you get into quantum field theory do you have to think about quantized particles and quantized fields (second quantization).

Also, there is indeed quantum mechanics in biology, but they're limited to photosynthesis. There is pretty much no way that quantum mechanics has anything to do with the brain except perhaps magnetosensing in birds.

 

[Developereo]

You seem to be confusing too different concepts.

Oh, I understand perfectly well the difference between detecting vague waves, and mapping patterns in the brain. What the current researchers are doing is no different, qualitatively, from mapping Broca's region or Wernicke's. They are just doing their pattern mapping at a finer grained level.

I will remain skeptical until they can demonstrate their algorithms across subjects. Most (not all) of these studies prime the system using one subject and read back the same subject's patterns. While I admit its usefulness, it's a far cry from general mind reading of random humans. The Scientific American article states as much, saying matching patterns across individuals remains elusive for the time being.

Depends on what quantum mechanics you want to talk about. Regular quantum mechanics usually has quantized particles moving in a classical field. 90% of real world materials can be described by classical fields with either quantized or classical particles moving in them. Only when you get into quantum field theory do you have to think about quantized particles and quantized fields (second quantization).

Also, there is indeed quantum mechanics in biology, but they're limited to photosynthesis. There is pretty much no way that quantum mechanics has anything to do with the brain except perhaps magnetosensing in birds.

What I was saying is that there's no physical analog to the field. A "field" is a mathematical construct only. It's a fancy way of saying object A affects object B at a distance, we're not quite sure how. One explanation is that such interactions are mediated by exchanging photons. The exchange of photons defines a "field" -- the "field" has no physical existence otherwise.