What level is Musk's brain-computer interface?

What level is Musk's brain-computer interface?

This is one of the hot science events in recent days, often make some scientific amazing events Elon Musk, just a few days ago on August 29th announced another news, by his administration of the brain-computer interface company Neuralink, developed a brain-computer interface products ~ a chip with the size of a coin, but also accompanied by a surgical implantation of the chip can be automatically completed! A robot that automatically performs the surgery in which the chip was implanted.

So what's so different about this brain-computer interface that Musk released? Let's discuss.

What is a brain-computer interface?

A brain-computer interface is also known as a brain port or brain-computer fusion perception. It is a modern science and technology that establishes a connection pathway between the human or animal brain and an external device. This interface has one-way and two-way points, one-way brain-computer interface computer only accepts commands from the brain, or the computer sends signals to the brain. Bidirectional brain-computer interface is the brain and computers and other external devices can carry out two-way information exchange.

In fact, the brain-computer interface itself is not Musk's first creation, as early as in the last century, brain-computer interface research has blossomed, and we are also involved in China, "China's Brain Program" project contains such content, is a national key project.

One of the more famous people in this field is the American neuroscientist Philip Kennedy, who is known as a "scientific madman". He began in the 1980s on the brain-computer combination of research, and in 1996 on a dying patient with acromegaly brain implantation of electrodes, and later there have been a number of paralyzed patients, car accident patients implantation experiments, although some progress has been made, but there is no obvious success.

For this reason, the FDA (U.S. Food and Drug Administration) revoked his license to conduct human experiments. When he made a major breakthrough in his research, he could no longer find volunteers. In order not to let his 29 years of research go to waste, he spent $25,000 on a surgeon in 2014, requesting it to implant this electrode in his brain. Before the surgery, he planned everything and was prepared to not wake up. It saved enough money to last a few months and made sure his son knew where he was.

But the surgery didn't live up to expectations. Kennedy woke up from surgery for the first time and at one point lost the ability to speak and respond to doctors. A few months later, he had to undergo a 10-hour second surgery. But his skull has been unable to heal since then, eventually forcing him to remove the electrodes. The electrodes were present in his brain for four weeks, and he used this time to collect a large amount of data on his EEG signals, obtaining valuable first-hand information for future research. For this reason, although he escaped with his life, he was left with a $96,000 bill for the surgery, and only $15,000 was covered by his insurance company.

Some of the scientists were controversial, and many did not approve of this risky approach. But his spirit was widely admired, and he became known as the "father of semi-robotics.

There are many more examples of brain-computer interfaces.

In 2008, neurobiologists at the University of Pittsburgh, through the brain-computer interface, let monkeys through the brain to manipulate the robotic arm will be food to get and send into their mouths; in 2004, the United States Cyberkinetics, Inc. to get permission from the FDA, in a 25-year-old paralyzed man implanted electrodes in the brain, the man through the idea of controlling the computer, to be able to watch TV and send e-mail, and even play computer games, with an accuracy of up to 10 percent, and the man was able to control the computer. The man was able to watch TV and send emails and even play computer games with 70% accuracy by controlling the computer with his mind.

In 2012, at a World Cup match in Brazil, a paralyzed teenager wore a machine battle armor designed by a lab in S?o Paulo, Brazil, and drove a soccer ball with a brain-computer interface and a mechanical exoskeleton; and in 2015, the U.S. Defense Department's Advanced Research Projects Agency surgically implanted a microchip to enable a paralyzed person to fly a simulated airplane.

In 2018, at the International Consumer Electronics Show (CES) in Las Vegas, Japan unveiled brain-controlled car technology, which allows drivers to quickly communicate brain waves to the vehicle without the need for a brain chip implant as long as they wear a special helmet on their head, making driving easy and fun.

There are also many scientific teams in various countries, are conducting brain-computer interface experiments, made various progress and results, such as a variety of paralyzed patients, through the robotic arm to obtain food, to obtain a drink, as well as through the idea, computer typing, etc. In October 2016, a volunteer named Nathan Copeland, respectively, motor cortex and sensory cortex implanted electrodes, successfully shook hands with then-President Barack Obama using a robotic arm controlled by his mind. The key breakthrough in this experiment was that the volunteer's mind produced the sensation of a "handshake", and Obama was shocked.

So the whole technology of brain-computer interfaces that Musk's Neuralink is releasing is not, in itself, a new thing.

The results of Musk's brain-computer interface.

What was shown at the launch event was a 23mmx8mm chip, which is the latest brain-computer interface chip developed by Neuralink after several years of hard work, named Link 0.9. This chip can sense temperature, air pressure, and read physiological signals such as brain waves and pulse, and it can collect and transmit neural discharge signals over 1024 channels, with a wireless transmission range of about 5 to 10 meters. The wireless transmission range is about 5-10 meters.

The principle of this technology is that through the implantation of the brain chip, connected to external machines, can read human brain signals, and recognize the intention of the person, which will be converted into a command signal, to control the external machinery and equipment to implement human intentions, to complete the task. In this process, the human brain can get feedback on the whole process.

On this display, Musk got a few experimental piglets, and by implanting a chip in the piglet's brain, people were able to see the piglet's brainwave state in various activities.

Musk claimed that the development of this technology, will realize that people with brain consciousness to control the Truss car, playing games, but also able to treat neurological diseases, such as improving the state of paralysis, depression, amnesia and other diseases due to damage to the brain or spinal cord, to help patients with asymptote recovery of sensory and motor functions.

Musk's real "black technology" is the use of "neural lace".

The so-called "black science and technology" was originally a network cartoon non-human strange science and technology, and now also refers to some of the subversion of human cognition of the new technology. Max's black technology is the development of a thing called "neural lace".

What is neural lace? This is a thing with science fiction color, in the science fiction writer Ryan Banks pen, "nerve lace" is through high-tech implantation of the human brain mesh, able to connect the human brain and computer, mutual communication. Musk's team borrowed this concept and called the fiber-like "thread technology" implanted in the human brain "neural lace". The thing about their neural lace is that it's very small.

Neuralink claims that their neural buds are 4-6um wide, about a quarter of the width of a human hair, and that they don't need to be implanted without opening the skull in the future, but only implanted with lasers, which allows them to be implanted deeper into the brain, and reduces the damage to the brain to a very small degree. Even so, there is still a long way to go to achieve good human-machine interaction.

Even Neuralink itself recognizes that 1024 channel signals are not enough to accurately transmit the complex information generated by the 86 billion neurons in the brain, which requires at least 1 million simultaneous neuron recordings, which means that millions of electrodes need to be integrated into the tiny chip.

Thus, even if this neural lace is only one-sixteenth the size of a hair, implanting a million lace is equivalent to 65,000 hairs. Let's think about it: 65,000 hairs implanted in the human brain, wouldn't that be horrifying? Therefore, the current advantage of implanting lace, which is thought to cause very little damage to the brain, will no longer exist.

How to make the "lace" connecting the neurons smaller, these components penetrate the brain tissue without causing damage to the hemorrhage, how to solve the problem of immune system rejection, what materials can be retained in the brain for a long time, how to solve the problem of safe and efficient charging of the chip, and so on, will make Neuralink a big headache, and will take enough time to solve the problem. It will take enough time to solve the problem.

But Musk let the cat out of the bag anyway.

Musk has never been confident in his "black technology", and this time is no different. He even threatened that his development of this neural lace technology will allow humans to live forever in the cloud in the future. From the front of the mechanical hand and Obama handshake, can produce in the paralyzed brain "grip" feeling, this human-machine two-way information exchange has not been a myth, so the ultimate realization of the combination of man and machine, change the form of human existence fantasy is no longer a dream.

Some people say that Musk is a big talker, and even many people are disgusted by his big talk at times. But Musk is a strange man, so far, his big words are based on the spirit of science and their own relentless struggle, and one by one, those big words have become a reality or are becoming a reality.

For example, the lightweight Dragon spacecraft developed by his team successfully carried people to the sky, realizing complete automation of docking with the space station; the spacesuits developed by them are lightweight and beautiful, and can be completely fashionable; he successfully realized the rocket recovery technology, which greatly reduced the launch cost; he tore open the iron curtain of the U.S. and even the global space field of the country's dictatorship system, and became the first private company to undertake a national space mission; he also threatened to soon send a new spacecraft to the world's most important cities and towns. He has also threatened to send people to Mars soon, and his Tesla electric car has taken the world by storm.

And he's just an entrepreneur and a private businessman, but people call him the Iron Man. So I choose to believe in the "big words" of the Iron Man. But I also believe that it will take a long time to realize these big words, and we need to be patient.

That's it. Welcome to the discussion, and thanks for reading.

If you usually pay attention to the progress of technology, I believe you must have noticed a heavy news last weekend. On August 29th, Elon Musk held a press conference for his brain-computer interface company, Neuralink, to show off the latest version of its brain-computer interface product: LINK V0.9, a miniature brain-computer interface device that looks like a coin and can be implanted in a person's head to capture brain signals around the clock.

I've noticed a lot of skepticism since this story was released. Some commented that Musk this time is just a product integration of technology that was already available in the field of brain-computer interfaces 10 years ago, with no essential innovation. Some people also pointed out a detail that the number of electrodes used to capture brain waves in this LINK V0.9 version is only 1,024, which is more than 2,000 fewer than Neural Link's own product from last year instead, suggesting that the launch may have been an exaggeration of publicity.

However, Mr. Peng Tianfang, who got on the show, had a different view. As the organizer of the Hard Tech Report, he has been paying close attention to the progress of brain-computer interfaces, and after the conference, he asked Mr. Sun Yu, the organizer of the Brain-Computer Interfaces course, as well as researchers from the Institute of Brain and Cognitive Sciences at Tsinghua University, about the progress of brain-computer interfaces.

He believes that Musk's actions in this area of brain-computer interfaces all point in one direction, which is to lower the threshold for collecting human brain data. And this is of great significance for brain-computer interfaces to move out of the lab and into the commercial market.

In today's content, I'm going to share Mr. Peng's views with you, and here are Mr. Peng's original remarks, which I'll paraphrase for you.

Hi, I'm Peng Tianfang.

After the launch, it's natural to have questions like, what is the level of the brain-computer interface product that Musk is releasing this time? What are the real breakthroughs? How far away from being able to let us directly use brain waves to manipulate household appliances, play games, and even realize the same dream we had when we were kids***: directly using a brain-computer interface to transmit knowledge to our brains?

To understand Neurolink's stage-by-stage progress, we have to start with the ultimate ideal form of all brain-computer interface products.

We know that the human brain is a complex system of tens of billions of interconnected neurons. With our understanding of the human brain today, every thought we have, every memory we have, exists in the form of an electrical signal that is strong or weak inside the neuron connections. So naturally, an ideal brain-computer interface should be able to measure the activity of all these tens of billions of neurons at the same time, and have the ability to control and modify the signals of any one of them. This ultimate form of brain-computer interface is also called a so-called "whole-brain interface.

Obviously, a "whole-brain interface" is too good an ideal for most of us to see in our lifetimes. But we don't have to be discouraged. It's as if all the computers we use today aren't the universal machines that were first conceived in the 18th century, but rather the so-called Turing machines of today. Scientists have always been able to find a workable solution to an engineering problem that is close to the ideal.

And the scientists at Neurolink believe that this workable solution for a brain-computer interface only requires that the interface's bandwidth be on the order of roughly 1 million neurons. That is, when a brain-computer interface product is able to monitor the real-time signals of a million neurons at once, that's enough to generate a lot of disruptive functionality.

For example, you can use brainwaves to play games, control airplanes and ships, and even discard the mouse and keyboard between humans and computers and use the brain-computer interface to transmit information directly.

So you might be asking, how far away is Musk from this simplified version of the program? Well, you'll find that the answer to that question is probably a lot more optimistic than the answer to the ultimate ideal of a "whole-brain interface". Here's a simple model to estimate:

Since the bandwidth of a brain-computer interface is roughly proportional to the density of the electrodes on a sensing chip, we might assume that the bandwidth of a brain-computer interface will grow at the rate of Moore's Law for chips in the future, which is to say, it will double every 18 months. Taking the August 29th launch with 1024 signal channels as a starting point, and assuming that each channel measures only one neuron's activity, it will take about 15 years, or roughly 2035, to get to the ideal brain-computer interface with a bandwidth of 1 million neurons.

That 15 years still sounds like a long time, but it's a much better and more predictable outcome than the "lifetime" answer.

But I think some people might be saying that Peng Tianfang is helping Musk to brag. I've been talking about the future for a while now, but what are the real breakthroughs in Musk's launch?

Indeed, the estimates just made are just pointing to a future that is feasible in principle, using the past experience of the development of similar technologies. And where Musk really excels is precisely in his ability to chart a path from reality to the future. This has been proven over and over again with his past successes in driving disruptive innovations in commercial spaceflight and electric cars.

We won't expand on these past successes for the sake of space, but let's get back to brain-computer interface technology. We know that today's brain-computer interface technology, whether in terms of product bandwidth, recognition accuracy, or biocompatibility, etc., there is still a great deal of shortcomings from the ideal program just mentioned. So how does Musk intend to drive the development of brain-computer interface products in the 15 years from today to 2035?

From this launch, I think that in Musk's view, the key to driving the maturity of brain-computer interface products is likely to lie in the word data.

Brain-computer interfaces, despite their rapid development in the past, have been very limited in terms of access to data and the amount of data available. For example, we've seen that most of the brain-computer interface research in the past required either recruiting volunteers or finding people who happened to have a brain disease. In short, it's been a small amount of data acquired over a relatively short time span, on a limited number of people, with equipment in a lab.

And as we know, electronic products for the masses often require a lot of user data as the basis for product design and optimization before they can finally mature and be brought to market. Not to mention the need for such electronic devices to be embedded in the human head. Due to the limited size of the data collected in the past, it was actually very difficult for product developers to produce results that would be useful for the functionality of a brain-computer interface product, except for published papers.

Thus, a positive feedback flywheel from product to data, to better product, to more data, is hard to spin. Not to mention the fact that we don't really know much about the exact meaning of what a large number of brainwave signals are saying, which requires massive amounts of data to back up the research.

And this time Musk released a product that, in my opinion, mainly tries to optimize the product experience of brain-computer interfaces, and solve the problem of the high threshold of data acquisition.
We can see that, compared to previous brain-computer interface devices, the most obvious change in this product is not a breakthrough in the technical underpinnings, but rather that it has made great improvements in the user experience.

In the past, when brain-machine interface devices were implanted in a person's head, their design was mostly taken out of the lab, and they were not only bulky, but also often needed to be connected to a few wires from the inside of the person's head, which made them look like medical instruments, which was inconvenient to wear and use, and might even have a certain impact on the user's self-esteem.

And this release of the product is only a coin size, the future plan is to embed a small surgery in the human skull to collect brain signals, and use the wireless way to transmit data remotely, and even through the way similar to the wireless charging of cell phones for charging. In terms of experience, this product is not like a medical instrument, but more like a health bracelet. I've put a photo of the product in my hand in the press release area, so you can take a look if you're interested.

According to Musk, the product will look almost indistinguishable from a normal person's head when it's mounted on a person's head. You can see that all the details of this product release are preparing for optimizing the user experience and lowering the threshold of data collection.

In addition, due to the lack of maturity of the brain-computer interface products, the past products can be said to be a variety of products, unlike the smartphone has a similar touch screen, camera, such as the industry's default design model. This is not conducive to iterative optimization of products based on existing standards.

And this time Neural Link released a product that is itself more productized and integrated, and even intentionally compressed the number of signal channels from the past 3072 to 1024 for the sake of a better design sense of the product. This is to sacrifice a certain amount of data bandwidth for the product experience.

Moreover, Neurolink has designed a surgical robot specifically for the implantation of this brain-computer interface. In the future, it is expected that the implantation surgery can be reduced to one hour. This is not only convenient for the user, like laser myopia surgery; for product development, it is also helpful to clarify the size of the product, the type of interface, and other development standards, so as to improve the speed of future product optimization and iteration.

All of this is to say that we've seen Musk's planning and ambition for future products reflected in this launch. But it has to be said that there are still a lot of obvious flaws in the product that need to be improved.

First, from the naming of the product, it's called LINK V0.9. This means that this product is not yet a brain-computer interface device that can be safely and legally tested and used on the human body. In fact the product demonstration at this launch event only used three piglets as a case study. Only one of the piglets was implanted with the BMI device on the day of the demonstration, and the other two were used as controls. This somewhat discounts the persuasiveness of this product launch.

Second, this release only demonstrated the ability to read brain signals, not write electrical signals to a neural network. That is to say that this time the brain-computer interface only reads signals of neural arousal when the piglet moves or smells a specific odor; but it does not have the ability to output signals into the piglet's neural network to influence or change the piglet's behavior. Although Musk said at the conference that the product could output electrical signals to neurons in the human brain in the future to treat certain mental illnesses or enhance human brain function, this function should be immature from the point of view of this product demonstration.

Third, although this brain-computer interface device reduces the threshold for data acquisition through a more mature product design, there is still a lot of research to be done in the future on how to effectively encode and decode the collected brainwave data itself to analyze the intrinsic meaning of the data.
Like what Musk said about treating mental illnesses, improving learning efficiency, and even uploading people's memories and other cool features, there is still some distance to go before it can be realized.

In addition to this, there are concerns about the long-term implantation of this product generated by the reaction of rejection; the need to frequently replace the hardware due to the renewal of electronic products. These are all areas that Musk's Neurolink will need to improve in the future.

But I think all of these flaws are justifiable for this launch. That's because the main purpose of the event wasn't to unveil a product that's already mature and ready to go on sale, but rather for Musk to show it off and recruit people for his three-and-a-half-year-old startup. At the end of the presentation, he said the company, which currently has just over 100 people, hopes to expand to about 10,000 in the future.

In the end, I'd like to say that brain-computer interfaces are a very complex field that still faces many technical and ethical challenges, and if you're interested in learning more about them, you can take the Brain-Computer Interfaces course that we've got with Mr. Sun Yu, and I'll be keeping an eye on the progress of this field for you.

Well, the above is the analysis that Mr. Peng Tianfang brought us. Brain-computer interfaces are a very complex but also very fascinating field. What attracts me most about this field is that it may create a completely new communication interface for us. If we don't need to use language to communicate, but can exchange consciousness with others through brain-computer interfaces, realizing "brain-brain interaction", what kind of world would it be? We welcome you to share your thoughts in the comments section.

Feeling that the brain-machine chip is a scam, should not be called a brain-level chip, should be called, muscle perception chip, to see is through the brain waves, cracked cognition, is a physical vibration, not cognitive signals. Called brain, inappropriate, called muscle perception is more accurate.

Using muscle perception to obtain human cognition, the amount of experimentation should be very huge, which is the same as collecting big data.

This is big data collection and application, not high tech.

The real brain chip should not be like this, the real brain chip should be active, passive signals of the two-way acquisition, physical signals are only a part of it, the central nervous system temperature perception should also be there, the human body's temperature is changing all the time, cracking the human body temperature information in order to get close to the real brain chip,

.