— Can we classify the human brain as a computer?
— The human brain is a computer, indeed. However, it should be noted that our understanding of the brain is always linked to our current level of technological advancement. That is, when electronic circuits first emerged — I'm talking about the popular then feedback loops — people said, "The brain is a feedback system". When computers were invented, people started saying, "The brain is a computer". Now, with the development of artificial neural networks, you will often hear "The brain is a neural network". So, in essence, yes, the brain is a computer in that it has elements that are interconnected, and we can observe how they function, even though we don't fully understand yet how that happens. But without a doubt, the functioning of the brain is somewhat similar to certain electronic circuits.

— But a computer isn't a neural network, right? So equality is kind of ​​nominal, isn't it?
— Of course, if you compare them in a literal manner, the brain is vastly different from the computers we use today, both in the way information is encoded and processed. But then again, there is no rule that computers have to be the way they are now. Neuromorphic computers that mimic the functioning of the brain are being developed.

— When was the concept of a brain-computer interface first introduced and what does it entail?
— The idea of connecting the brain to a computer or an external device was first conceived by science fiction writers. So the idea was floating around. But probably the first demonstration of a neural interface happened in 1963 when William Grey Walter showed that he could connect his patients' brains with electrodes in the motor cortex to a device like a slide projector.
He asked the patients to press a button to switch slides. At that moment, a readiness potential was generated in the brain. It was strong enough that Grey Walter could disconnect the button and connect the readiness potential to the slide projector, allowing the patients to switch slides.

— So they switched slides without pressing the button, just using thinking power?
— Yes. At least according to eyewitnesses – it wasn't published as a scientific paper. The participants were amazed at how the device could anticipate their desire to press the button before they actually did. Their brain activity was recorded using electrodes.

— But it's not that simple, is it?
— Not at all. For example, contemporary deep learning techniques can identify patterns. If   an image is presented, they can recognize what's in it. But why does that happen? Because there is definitely something recognizable in the picture. The situation with the brain is slightly different, as it contains elements that use both electrical and chemical methods of signal transmission. We can attempt to connect to those elements, but it's not the same as connecting to a cable that delivers information in a way convenient for us. The electrode we use is significantly larger than a neuron. There is a delicate network of neurons, and suddenly a large nail-like object is inserted into it, attempting to record things. If we consider exploring a computer that way, it would be extremely odd that we are able to achieve anything at all. It's possible with the brain because the brain is somewhat "friendly" towards researchers. For instance, if I move my finger and certain neurons discharge each time, I can say, "Aha! These neuron discharges are linked to finger movements", and by recording the neurons' activity, I can decode finger movements. The neurons have discharged, which means the finger is bending, or I'm thinking about bending my finger. That's one way to do it. We can also use non-invasive methods, such as placing electrodes on the head. That way, there is no need to drill holes in the skull, and the electrodes will record larger-scale potentials that appear as rhythms: alpha-rhythm, beta-rhythm, gamma-rhythm, and so on. They represent the synchronized activity of a huge number of neurons, so this method differs somewhat from directly recording neuron discharges. Furthermore, the brain's blood flow can also be studied. For instance, when a certain area of the brain is active, the blood flow there increases as well.

— You know, linguists can even see changes in brain activity on MRI scans when listening to nouns. Each case has its own unique brain activity. People actually receive grants to research the areas.
— Yes, MRI is a highly effective method, particularly because its spatial resolution allows us to look deep inside the brain, where surface electrodes can't reach.

— What does this information demonstrate and how is it interpreted?
— It is a major challenge because, as I've mentioned, it's not designed for our convenience to begin with. Scientists are essentially trying to decode it without knowing the code. All methods are correlation-based, meaning if a neuron is activated when I do something, it's likely to be involved in the process in some way.

— And there has to be statistical significance, meaning the same pattern has to be repeated many times, right?
— Yes, either many times or we can prove it by recording patterns from numerous neurons simultaneously, which also enhances the statistical validity. Say, neural interfaces aim to record as many channels as possible to improve decoding accuracy.

— The accuracy of decoding that's still not quite decoding but rather something as close to it as we can get?
— Yes. It's a type of correlation method, but we should keep in mind that correlation doesn't imply causation.

— As far as I understand, Elon Musk's current high-profile neurotechnology projects (I'm referring to Neuralink) are genetically linked to your work, and his current employees used to work in your former lab, correct?
— Yes, I've worked with some of them, and two of them have since left Musk due to their own reasons. But his ambitions are grand. Generally, Musk is replicating what we already did a decade ago, but with significantly superior technology. Therefore, by investing assets and recruiting engineers, he has quickly made those interfaces truly practical. They are small, fully implanted, and communicate wirelessly with the outside world. He recently demonstrated a monkey, and it's not even apparent that the monkey has an implant, which is a considerable achievement. As for what exactly he intends to do, I believe he himself has very little idea. However, I don't personally know Musk, so I can't say for certain. He clearly senses a significant potential for growth in this area. Someone once asked him about technologies of the future. And instead of saying space or electric cars, he responded, "The brain. The brain is the future."

— Do you know if his chip will have some fundamentally new features that others lack?
— There is nothing fundamentally new there yet, but everything is done very skillfully. It's a greatly enhanced version of our old ideology that if you insert as many electrodes as possible into the brain to record the activity of individual neurons, you can successfully decode anything. Any thoughts. And then, since we have such access to brain activity, we can control neuroprosthetics, treat people, and even enhance our existing brain.

— As far as I understand, there is something curious about the way his chip processes signals, right?
— Musk isn't unique in that regard. In an interface, you need to decide how much processing to allocate to the chip itself and how much to delegate to external devices. Because if the chip merely records neuronal activity and transmits it in its raw form, it consumes a lot of energy and places high demands on the communication channel. That doesn't work. So, the chip needs to do some of the work itself. In Musk's project, the chip performs spike sorting, meaning it observes the discharges of neurons and identifies when a particular neuron has discharged. So instead of recording the entire discharge pattern, it only records when and which neuron has discharged and transmits that information outward. Then, decoding algorithms located outside the brain on an external computer take over.

— Yes, Nina Sadykova, who used to work at the Institute of Plant and Animal Ecology of the Ural Branch of the Russian Academy of Sciences, organizes birdwatching tours around Yekaterinburg. She shows the participants around and educates them about the local environment...
—Yes. There is a lot of this happening now. In Moscow, for example, an organization called Ptitsy I Lyudi [Birds and People – editor’s note] does this. There is Birdwatching Moscow, which offers training through the Norwegian BirdID program. Andб of course, the development of phones and apps helps. Someapps make bird identification easy. Competitions are held to see who can spot the most diverse range of birds in different locations, adding a competitive element to birdwatching. A lot of this is happening in Russia now, but it doesn't compare to the scale of it in the West.

— As far as I know, ornithologists often enlist birdwatchers to monitor populations, say, nightingales in the floodplain of the Moscow River, but this isn't exactly avant-garde ornithology...
— Musk isn't unique in that regard. In an interface, you need to decide how much processing to allocate to the chip itself and how much to delegate to external devices. Because if the chip merely records neuronal activity and transmits it in its raw form, it consumes a lot of energy and places high demands on the communication channel. That doesn't work. So, the chip needs to do some of the work itself. In Musk's project, the chip performs spike sorting, meaning it observes the discharges of neurons and identifies when a particular neuron has discharged. So instead of recording the entire discharge pattern, it only records when and which neuron has discharged and transmits that information outward. Then, decoding algorithms located outside the brain on an external computer take over.

— To wrap up the technical part: am I correct in understanding that such interfaces are primarily invasive, meaning they need to be inserted into brain tissue?
—It depends. There are invasive interfaces where electrodes are inserted into the brain, but that raises biocompatibility issues. The brain isn't gullible. It recognizes when an electrode has been inserted and tries to protect itself from it — first through glial cells that surround and seal off the electrode, and then by wrapping everything in such dense connective tissue that it affects recording quality.
By the way, if you use that electrode for brain stimulation, it will still be functional even if it becomes encapsulated. However, for recording purposes, its behavior is highly unpredictable. It might work for two weeks and then stop, or it might work for many years. Some monkeys in our experiments had electrodes for up to eight years, and the recordings were quite satisfactory.

— What about non-invasive interfaces?
— Non-invasive means we don't penetrate any part of the body. The most common method is to place electrodes on the scalp and record electrical activity of the brain. The brain is a source of electromagnetic waves, which we have been recording for well over 100 years. This type of recording is highly informative and suitable for many purposes, such as rehabilitation. For instance, it is extremely useful when we are rehabilitating a patient with neurological damage and need to understand how their activity changes based on the therapy we administer.

— Would invasive methods be preferable for an experimental researcher?
— Yes, invasive methods are indeed better from the perspective of a researcher. However, if we aim to develop a practical system, we must consider how many people would be willing to have electrodes implanted in their brain. If we can develop new non-invasive recording methods that could benefit hundreds of thousands of people, then perhaps that's the direction we should pursue.

— You conducted unique and pioneering experiments on monkeys and rats in the U.S. What were your findings that elevated this field to a fundamentally new level, influencing its future development?
— We always aimed to conduct pioneering research, striving to innovate rather than replicate what had been done before us. Our first notable demonstration was creating a neural interface that controlled typical movements for primates — reaching and grasping — which are also common to humans and many other animals.

— Was this achieved through the power of thought?
— Yes, it was achieved through thinking power. We recorded activity in the motor cortex and the action was performed by a robot. It was quite a famous study. We then wondered if we could actually add sensation to that movement. By that I mean that we control a hand that reaches out and touches objects, but what use is it if it lacks sensitivity? But if it can touch an object and feel it, then...

— How would you know if it felt something? Does it react to, say, hot objects?
— Something like that. We place sensors on that mechanical hand and they sense object properties: shape, texture, temperature, etc. By the way, it's not easy to develop a skin for the robot that can sense temperature, touch, and so on. It was our second demonstration where we showed the brain-machine-brain interface that not only controlled a virtual hand but had that hand send back signals telling us what it was feeling though its artificial tactile sensations.
Then we made some progress using such interfaces for bimanual tasks. Because even though experimenters love tasks involving only one hand or finger, in reality, we use both hands for most tasks. So, we developed an interface where a monkey could control two virtual hands independently using a neural interface, thus proving that it was actually possible. Interestingly, to achieve that independence, we didn't need to record from different parts of the brain. We recorded from the same set of neurons, and that ensemble enabled independent control.
Then we thought, we've done enough with hands, so let's focus on legs. So we developed an interface where a monkey walked on a treadmill and we decoded the walking pattern.  We monitored brain activity and decoded the stepping movements in such a way that, just by observing the discharge of brain neurons, we could determine the type of stepping movements the monkey was making. However, that wasn't enough. We needed to demonstrate why it was necessary at all.
We linked our monkey in America, to a humanoid robot in Japan, and it moved in sync with the monkey. Then we halted the treadmill, and the monkey, seeing the robot's image on the screen in front of it, continued to control it while standing.

— What was the practical purpose of that?
— The implication is that, generally speaking, a person with paralyzed legs could potentially control an exoskeleton attached to their legs. We demonstrated it both in Brazilian research and at the Higher School of Economics where I had a megagrant. We showed that a robot's stepping movements can be initiated through electroencephalograms, which is particularly useful for paralyzed individuals as they need to achieve Hebbian plasticity. The exoskeleton takes a step, triggering a receptor discharge that travels to the brain via any remaining fibers post-spinal injury. Then, that sensory influx synchronizes with the brain activity, initiating the healing process.

— Are those the so-called artificial sensations?
— Yes, it's an artificial sensation combined with synchronization of brain activity, which leads to neurorehabilitation.

— Let's talk about it in a little bit more detail. What exactly is an exoskeleton?
— Our regular skeleton is an endoskeleton, meaning it's located inside us. For instance, a cockroach has an exoskeleton because its supporting structure is on the outside. Therefore, an exoskeleton for a human is a device that attaches to them and can move their arms, legs, torso, and essentially their entire body in a manner that suits the individual. This is useful in industries where a person wants to perform an action but needs to amplify it: for example, lifting five tons with the tip of a finger or moving that weight from one place to another. Such industrial exoskeletons do indeed exist. The other use they have is for rehabilitation purposes, which we've mentioned before. We have a paralyzed patient who can't move their own body, so we equip them with an exoskeleton, enabling them to move. There are various types of exoskeletons, ranging from those that attach to a single finger or hand, capable of moving fingers or bending and extending them, to those that cover the entire arm or even legs, enabling the person to walk. There are even full-body exoskeletons that can move both arms and legs. In some respects, this is a partial solution. We're not fully curing the person but transforming them into a sort of cyborg. Nevertheless, in certain cases, it represents a significant advancement. For instance, a patient with a spinal cord injury who spends their entire life in a wheelchair can be placed in an exoskeleton. This not only allows them to stand upright but also to walk down the street, essentially making them like any other normally walking individual.

— So, they can move from point A to point B independently, without any external help.
— Precisely. This is beneficial both socially and health-wise, because when the legs start functioning in an exoskeleton, the overall physiological condition improves. Furthermore, neuronal plasticity mechanisms are activated, potentially restoring the ability to move. This has been tested for various conditions, and the effect is noticeable in cases of spinal cord injury, stroke, and cerebral palsy. This type of recuperative walking is extremely beneficial. Although it's not as straightforward as it appears. For instance, when we perform a movement for someone, it can paradoxically teach them not to do the movement themselves. There are exoskeletons where the individual is fully suspended and a robot moves their legs. Surprisingly, this method is less effective than the one where they're placed in an exoskeleton and given crutches to prevent them from falling, meaning that they will simply topple over if they don't put in the effort. So, the most effective approach turns out to be the necessity to do something on your own.

— Discussing the capabilities of contemporary neuroscience, neurophysiologists cite something referred to as the "Jennifer Aniston neuron" or "grandmother cell", sometimes also called the "Eiffel Tower neuron". What is that, exactly?
— Actually, that story goes way back. In the beginning of the 20th century, Santiago Ramón y Cajal and Camillo Golgi were awarded the Nobel Prize for observing the brain under a microscope and identifying interconnected neurons. Santiago Ramón y Cajal exclaimed, "Behold, a neuron! It is a fundamental unit of the brain." And Camillo Golgi said, "There are no fundamental units here, only a functioning tangled network of sorts." This discovery essentially revealed two distinct schools of neuroscientists. One group consistently highlighted the neuron and its visibility... Read any popular science literature that says, "A unique neuron has been discovered, responsible for..." or "A mirror neuron has been discovered. It acts like a mirror." The other group stressed the networked nature of the brain. So, the "Jennifer Aniston neuron" or "grandmother cell" is a highly specialized neuron that only reacts to Jennifer Aniston. However, the truth likely lies somewhere in between. On one hand, there are indeed highly specialized neurons that reside in the nodes of that network, composed of a vast number of neurons, and provide specialized information. Meaning, if you need to identify Jennifer Aniston, you will receive a clear signal somewhere telling you that it's in fact  her. But there is also a massive network that identifies all of this, and it closely resembles artificial neural networks.

— Let's get a little more practical. Are there any companies currently making profits from the development of neurointerfaces?
— We need to consider practical outcomes here, and the most significant practical outcome is the cochlear implant. It's implanted in hundreds of people with hearing impairments, and all companies operating in this field are generating real profits and achieving tangible practical results. Although there is plenty of room for improvement, both in the quality of the electrode and the decoding process.

— Is there anything else? Visual prosthesis, perhaps?
— There is actually a new wave of research on this topic. Restoring vision is a much more complex task, but it's great to see that this field has also entered the business realm. I believe that we will see more and more blind individuals whose sight has been restored. Even if it isn't restored fully, the difference between complete blindness and some visual signals is significant.

— And the futuristic part probably has to do with mind-reading, right?
— Yes, mind-reading. And it's not as futuristic as you think. Even a simple method like evoked potentials can evaluate preferences you might wish to conceal. This immediately brings up an ethical issue, although neuromarketing would certainly like to exploit this opportunity — any business benefits from knowing what a person truly desires if it wants to boost its sales.

— But I don't want the salespeople to know that I liked that can of baked beans!
— You may not want them to, but they will know. That, or you need to have a really good poker face. But even then, many of your preferences can be discerned just by observing your eye movements. So, yes, to maintain some level of privacy, we need to consider the ethical implications of these technologies.

— Wrapping up the topic of futurism, who exactly are human cyborgs? In 1998, we hosted Kevin Warwick, a human cyborg who implanted a chip in his forearm allowing him to control lights in his house, open electronic locks, and essentially integrate his nervous system with a neural interface to convey his emotions to another cyborg, his wife Irene. How many human cyborgs exist today?
— There are actually quite a few of them.

— There are actually quite a few of them.
— I've already mentioned people with cochlear implants...

— I'm specifically referring to people who want to enhance their capabilities and, say, open a door with a chip in their elbow instead of a key.
— In essence, any modification (and there are many of them) gradually transforms a person into a cyborg. Take Parkinson's disease for example. Patients are implanted with deep brain electrodes that stimulate them and alleviate the symptoms of the disease. However, it also alters their personality. One side effect could be that they become happier as the nearby areas in the brain associated with happiness are stimulated. So, any such implantation will inevitably change the personality. Regarding such implants for practical purposes, like not wanting to use a credit card and opting for a chip implant instead, it might be convenient, but it's not exactly a neurointerface. It's more of a rudimentary application. Pets are implanted with such chips and they don't mind. They're not even aware of it.

— How long can a chip remain in the body? I haven't kept in touch with Kevin Warwick, and I wonder if he still has the same chip or if it needs to be replaced from time to time.
— A chip can stay in the body for quite a long time. The only thing is that it will naturally be encapsulated, meaning the body will protect itself from it. If the body determines that sufficient protection is achieved, the chip will remain in place indefinitely. Generally, the body is so smart that it attempts to expel any implant, much like it does with splinters. In the Soviet Union, they sent monkeys into space with small implants, and those implants were suppressed in such a way that scientists struggled to quickly locate them.

— You mentioned a certain Philip Kennedy in one of your lectures. What is he famous for?
— Philip Kennedy's main idea was to create a neurotrophic electrode, an electrode containing a nerve growth factor. The concept was that neurons generally dislike foreign bodies, but this electrode contains a chemical that attracts them, causing them to integrate into it and function optimally. Philip conducted research on individuals with severe paralysis and achieved some positive results, but progress eventually stalled. His methodology didn't gain traction in other labs, and it appears he lost his funding. Ultimately, he took the drastic measure of implanting the electrode in himself. Unable to legally do it in the U.S., he traveled to Central America, where a neurosurgeon performed the implantation. However, there were complications, including various side effects and even loss of speech.

— Did he intentionally implant the electrode in the language center?
— Yes, he was specifically interested in speech mechanisms and wanted to test the electrode's effectiveness on himself. Luckily, he was able to recover. I don't know him well, but I saw him at a conference and he seemed fine. During the experiment, he even kept some records of his own experience, but it didn't result in any significant publications.

— Was he trying to observe how the brain processes speech signals?
— Precisely, yes. He wanted to generate speech through thought, as if a person lost the ability to speak but could still formulate words, thinking about what to say. The interface would convert those thoughts into sound signals, and the speaker would generate speech.

— Didn't Stephen Hawking have something similar?
— Not exactly. When I  worked in the States, I heard discussions about whether to implant such an interface in his brain. However, they decided not to risk such a valuable individual and found a non-invasive solution instead. He retained the ability to move his cheek muscle, which he used to control the speech synthesizer.

— Do we actually understand what consciousness is? Why do we believe that humans possess it but animals don't?
— The subject of consciousness is indeed very important, and while many scientists won't openly admit it, it is likely their main motivation for studying the brain. The renowned Ivan Pavlov wrote extensively on this topic, albeit from a contrasting perspective. "We are studying an animal, but should we consider its emotions or thoughts? Absolutely not, as it would not constitute real science. We must study the brain objectively, and pondering such questions would introduce subjectivity." Indeed, his expressive discourse on the subject reveals that he had clearly pondered over it himself. However, he concluded that to truly be a scientist, one must focus solely on studying the brain. So, what can be said about it today? By delving into the study of the brain, we can consistently progress, enhance our understanding, and eventually explain almost any phenomenon through the workings of the brain. We can examine even the most intricate thought, trace its origin, and identify the specific area of the brain responsible for it. We can understand what calculation that area performed and how it resulted in a particular signal that was transmitted to the language center, leading the person to say certain words. That would indeed provide a comprehensive explanation, but it raises a significant question: why does a person need to be conscious in that process? After all, any robot could perform the same tasks, operating in complete darkness like a mechanical or electrical device, speaking and appearing intelligent, intellectual, and rational, yet devoid of consciousness!
I have personally conducted an informal survey among neuroscientists and I can assure you that no one has a clear understanding of where consciousness originates from. Nobody understands where the subjective aspect of consciousness stems from. There is no answer to that question, and perhaps we may never find one. Therefore, it's a perfect question for philosophers as they love dealing with issues that are either difficult or impossible to explain.

— Do you agree that our brain makes all decisions for us, a viewpoint often supported by references to the well-known Libet experiment? Do you believe it's possible for our brain to control us without our awareness?
— You're prompting me to discuss the concept of the soul from a philosophical standpoint, so let's refer to Descartes. He described the soul as our spiritual component and the brain as our mechanical and materialistic component. Descartes also devoted a substantial amount of writing to theorizing how the soul could be connected with the body. Of course, he understood that if we possess a soul that links with the body, they must interact in some way. However, any interaction between the immaterial and the material results in a violation of physical laws, which is impossible. Most likely, this way of posing the question is simply invalid: it's us against our brains. We must acknowledge that we and our brains are, to some extent, one and the same, right?
Upon closer examination, Libet's experiment falls apart. Having conducted numerous experiments on monkeys, I have essentially replicated Libet's experiment. I could observe that before a monkey performed an action, it had the intention to do so. For instance, if a monkey was holding a lever and needed to randomly turn it right or left, I could predict its intended direction based on its brain activity. The monkey itself may not have been aware of it. Perhaps it was, say, a form of habit.
Now, let's consider a broader perspective. It's a well-known fact that we perform many actions automatically without the involvement of our conscious mind. For example, if your knee jerks when hit with a hammer, it's due to a coordinated activation of certain muscles. I have no idea which muscles are activated or in what sequence — my brain makes that decision for me. The brain performs many, many tasks subconsciously.

— But if you were to pick up this glass, here, it would be your intention, right?
— Yes, but that doesn't mean that our conscious and subconscious mind are two entirely different things. The extensive processing performed by the brain serves as a foundation for consciousness. If you remove it, our consciousness will vanish entirely, suggesting that it is a higher construct built upon brain activity. However, it's incorrect to distinguish between consciousness and subconsciousness, as all subconscious elements contribute to consciousness. Libet's experiment is flawed because it is excessive. Its subjects were instructed to make a movement but had to evaluate their desire to make the movement before doing so. Any mental task of this nature is associated with some form of brain activity. It's quite clear that the subjects began contemplating whether they wished to make the movement or not, and their brain activity increased.
There are other experiments claiming that an action can be predicted 11 seconds in advance. But sometimes, you don't even need to examine the brain. You can predict the number a person will think of by simply asking them to think of random numbers. This is because humans are incapable of generating random things on demand and always follow some algorithm, which a mathematical algorithm can detect and use to decode what the person is thinking.

— As we are nearing the end of our conversation, I'd like you to comment on another trending concept or phenomenon, the "super brain". Can people communicate solely non-verbally and how many individuals can be connected to such a system?
— Let's begin from the second part. You can connect as many people as you like, but you will need to come up with a working paradigm. For now, the ways to connect one brain to another are quite rudimentary. For instance, I can generate something with my brain, and another person can be stimulated with transcranial magnetic stimulation, causing their finger to twitch. So far, all existing articles essentially state that I wanted something and the other person's finger twitched. But that's an obvious outcome. When they devise something less predictable — connecting one brain to ten, a hundred, a thousand, or a million others — it could take on intriguing forms. That could actually be called a "super brain", where each individual is solving their own task, unaware of what the "super brain" is doing — which could be solving a super task.

— To finish off our conversation, I have a very simple, practical question. There was an exceptional physiologist, Ivan Pigarev, who tragically passed away recently. He worked with cats. He had them scurrying around his lab with electrodes attached to their heads. He was very fond of them and referred to them as his full-fledged colleagues, calling them "the cats who work in our lab". What is a researcher's relationship with monkeys?
— Each researcher forms a unique relationship with monkeys. I've observed a wide range of relationships, all the way to researchers forming friendships and trust with their monkeys. Monkeys have very distinct personalities, you know. And then there are scientists who don't even look at the monkey when recording it. Instead, they look at neuron discharges on the oscilloscope. That's what they're interested in. So there is no definite answer here.

— And you?
— I guess I fall somewhere in the middle of those two extremes.