Eccentric in the best sense of the word entrepreneur, playboy, philanthropist Elon Musk known around the world. He decided to bring humanity into space, to colonize Mars, to abandon a disposable rockets. He decided to make the world cleaner, transplanted us from cars with internal combustion engines to self-driving cars. While unfolding these businesses, he does not sit idly by. He conceived Neuralink that will help us become new people. Without borders and without weaknesses, as it should be in the new world (Elon musk). To document the crazy ideas Mask, as always, volunteered Tim urban from WaitButWhy (he wrote about artificial intelligence, colonization of Mars and SpaceX). We present one of the best works of modern popular science journalism. Further, in the first person.
Part 1: Colossus Of Human
Part 2: Brain
Part 3: Flying over the nest of neurons
Part 4: brain-computer interfaces
Since I already wrote about the two companies Elon musk — Tesla and SpaceX — I think I understand his formula. It looks like this:
And his first thought about the new company just starts right and goes left.
He decides that certain changes in the world will increase the likelihood that humanity will have a better future. He knows that large-scale change in the world is happening most rapidly, when the whole world — Colossus of man — working on it. And he knows that Colossus of the Human will strive to achieve the goal, and then only if there is economic driving force if the process of spending resources to achieve this goal is good business.
Often, before a thriving industry gain momentum, it all seems like a pile of logs, all the components for a fire place all ready to go, but no matches. There is some technological deficit, not giving off the entire industry.
So when Elon creates the company, its main strategy usually is to create a match that will ignite the industry and make Colossus a Human to work on it. This, in turn, according to Elon, will lead to events that will change the world in such a way that will increase the probability that humanity will be a better future. But you need to look at his company from a bird’s flight, to understand it all. Otherwise, you will mistakenly believe everything he does, business as usual — whereas in fact what looks like business will be the mechanism for the maintenance of the company implementing the innovation to create an important match.
When I was working on articles about Tesla and SpaceX, I asked Ilona, why he climbs into engineering, not in science, and he explained that when it comes to progress, “engineering is a deterrent”. In other words, the progress of science, of business and industry — all this happens with the permission of the technical progress. And if you look at the history, it makes sense — since each of the greatest revolution in human progress — this is a technical breakthrough. Match.
So, to understand the company Ilona Mask, you need to think about the match that he’s trying to create — along with three other variables:
And when I started thinking about what a Neuralink, I knew which variables I need to put down. At that time I had a very vague idea about one of the variables that the company’s goal is to “accelerate the emergence of cerebral neural interface”. Or a magical hat, as I called it.
As I understand it, the General interface of the brain was supposed to represent a brain-computer interface in an ideal world — super-duper-advanced concept, when all the neurons in your brain can invisibly communicate with the world outside. This concept was based on the sci-Fi idea of a “neural lace,” the “Culture” of Iain banks — weightless, intangible interface for the whole brain, which can be beamed into the brain.
Questions I have had enough.
Fortunately, I was headed to San Francisco, where they were to sit down with half of the team of founders Neuralink and be the dumbest person in the room.
Digression on why I’m not exaggerating, calling himself the dumbest person in the room, just see for yourself.
Team Neuralink:
Gender Merollawho spent the last seven years in the role of lead designer of the chips in IBM’s SyNAPSE program, where he led the development of chip North, one of the largest CMOS devices in the history of the number of transistors, if that. Paul told me that his work area was called neuromorphic, and the goal is to create solid state circuits, based on the principles of the architecture of the brain.
Vanessa Tolosa, an expert on microspore team Neuralink, one of the leading researchers of biocompatible materials in the world. Vanessa’s work includes the design of biocompatible materials on the basis of principles of industry of integrated circuits.
Max Hodak, who worked on the development of several innovative technologies NCI in the laboratory of Miguel Nicolelis at Duke, and twice a week went to College for the launch of Transcriptic, “a robotic cloud laboratory for life Sciences”, which he himself founded.
D. J. SEO, who in his more than 20 years, developed at the University of California at Berkeley ultra-modern new concept NKI called “neural dust” — tiny ultrasonic sensors that can provide a new method of recording brain activity.
Ben Rapoport, an expert on surgery in Neuralink and a leading neurosurgeon. He has a doctorate in electrical engineering from the Massachusetts Institute of technology, allowing him to skip his job neurosurgeon “through the lens of implantable devices.”
Tim Hanson, whose colleague was introduced as “one of the best engineers around the world on the planet.” He is self taught, but thanks to their knowledge of materials and methods of microfabrication, he managed to create some of the key technologies that will be used in a Neuralink.
Flip Sabes, leading researcher, laboratory of which the University of California in San Francisco, broke new ground for the NRI by combining the “cortical physiology, computational and theoretical modeling, Psychophysics and physiology of man.”
Tim Gardner, a leading researcher at Boston University laboratory which is working on the implementation of CQM in birds to explore “how complex the songs are collected from the elementary neural units” and learn “about the relationships between patterns of neural activity in different time scales”. Tim and Flip left their positions to join the team Neuralink.
Well, he Ilon, CEO and member of the team. The post of Director General allocates the project to the rest that he recently launched, and puts Neuralink in the highest priority for him, where there is only SpaceX and Tesla. When it comes to neurology, Elon has the lowest technical knowledge in the team — but he started SpaceX without special technical knowledge and quickly became a certified rocket specialist, reading and asking questions of the experts in his team. The same thing will happen here.
I asked Ilona, as he gathered his team. He replied that he met only with 1000 people to gathered this group, and part of the task was a huge number of completely separate expert areas that you should consider: neuroscience, neurosurgery, microscopic electronics, clinical trials, etc. Because it is an interdisciplinary field, he was looking for interdisciplinary experts. And it shows in their biographies — all members of the group possess a unique combination of knowledge that intersect with the knowledge of other group members and together constitute a kind of magexpert. Elon also wanted to find people who could look down on the mission — which was more focused on industrial performance rather than the production of paper. In General, it was not easy.
But here they sat at a round table and looked at me. I was a bit shocked, because I had to spend a lot of research before coming here. I have extracted the thesis, they picked it up and multiplied it four times. And while the discussion continued, I began to understand what was happening.
Over the next few weeks, I met with other founders, each time playing the role of fool. At these meetings, I concentrated on attempting to provide a comprehensive picture of the challenges facing us and how it will look in the path to magical hat. I wanted to understand these two boxes:
The first was simple. The business portion of the Neuralink is a company engaged in the development of brain-computer interfaces. They want to create cutting-edge NCI — some of them are “devices of micron sizes.” This process will support the growth of the company and will be a great base for the implementation of the innovation (like SpaceX uses its launches to maintain the company and experiment with the latest engineering developments).
As for the interface, over which they intend to work, here’s what Elon said:
“We want to hang out on the market something that will help in certain serious brain injury (a stroke, a cancerous lesion, congenital), in about four years.”
The second box was more difficult. Today it seems obvious that the use of the technology of the steam engine for the power of fire was to begin, that the industrial revolution happened. But if you spoke with someone in 1760 about it, clarity would be much less — what obstacles need to be overcome, what innovations to implement, how long it will take time. And here we are — trying to figure out how to be the match that will ignite nanorevolution, and how to create it.
The starting point for the discussion of the innovation will be discussion about the obstacles — why should there be any need for innovation. In the case of a Neuralink this list will be great. But even given the fact that the main limiting factor will be the engineering development, there are several major problems that is unlikely to be a major obstacle:
Public skepticism
Recently a survey was conducted, which found that Americans fear the future of biotechnology, in particular CQM, more than editing genes.
Flip Sabes do not share their fears.
When a scientist thinks about changing the fundamental nature of life — the creation of viruses, eugenics, etc. — creates a spectrum that is a biology find quite disturbing, but I know that when neuroscientists think about chips in the brain, it is not strange, because we already have chips in the brain. We have deep brain stimulation, which relieves the symptoms of Parkinson’s disease, we conduct the first test of the chips to restore vision, we have a cochlear implant — we don’t it seems strange to put the device in the brain to read and write information.
And learning all about chips in the brain, I agree — and when Americans find out all about them, they will change their minds.
History supports this prediction. People very quickly got used to eye surgery Lasik when it first appeared 20 years ago, only 20 000 people per year have resorted to surgery. Today, that number is already 2 000 000. The same with pacemakers. And defibrillators. And organ transplantation. But she gave frankensteining! Brain implants will be more of the same.
Our lack of understanding of the brain
Remember, “if you imagine the concept of the brain of one mile, we passed only three inches in her”? Flip thinks so too:
If we were to understand the brain to interact with it on the merits, we would have problems. But all of these things in the brain can be decoded without a complete understanding of the dynamics of computations in the brain. The ability to read it all — it’s the engineers ‘ problem. The ability to understand the origin and organization of neurons in the smallest details that would satisfy neuroscientists in full — it is a separate problem. And we don’t need to solve all these scientific issues to make progress.
If we can just using techniques to make neurons talk to computers, this will be enough, and machine learning will take care of the rest. That is, will teach us the science of the brain. As noted by Flip:
The reverse side is the phrase “we don’t need to understand the brain in order to progress” lies in the fact that progress in engineering will almost certainly increase our scientific knowledge — such as how Alpha Go will teach the best players in the world best strategies of the game of go. And this scientific advance will lead to more technological progress — engineering and science will push each other.
Evil giants
Tesla and SpaceX both come in a very large tails (e.g. automotive industry, oil and gas and the military-industrial complex). Large tails do not like when they come, so it is usually doing everything they can to impede the progress of attackers. Fortunately, Neuralink, there is no problem. There is not a large scope of activities, which can destroy Neuralink (at least in the foreseeable future — and there, it is possible to nanorevolution will disrupt almost every industry).
Obstacles Neuralink is the technological obstacles. A lot of them, but two of them are mansions, and if they are overcome, it can be enough to all the remaining walls fell and completely changed the trajectory of our future.
Big obstacle #1: throughput
At the same time in the human brain was never more than a couple of hundreds of electrodes. If you compare the vision, it is equivalent to extremely low resolution. If you compare with the engine, this is the simplest command with a low degree of control. When compared with thoughts, several hundred of electrodes would be sufficient only to pass just above the message.
We need more bandwidth. Much higher.
While thinking about the interface that could change the world, the team Neuralink has identified an approximate number “a million neurons simultaneously read”. They also say that 100,000 is the number will allow you to create many useful NKI with different applications.
Similar challenges faced the first computers. Primitive transistors were bulky and difficult to scale. But in 1959 the integrated circuit — a computer chip. Along with it there is a way to increase the number of transistors and Moore’s law — the notion that the number of transistors that can fit on a computer chip doubles every eighteen months.
Up to 90 electrodes for NRI doing hands. Then we started to figure out how to make these tiny 100-electrode multielectrode matrix, using modern semiconductor technology. Ben Rapoport from Neuralink believes that “the transition from manual production to the Utah electrode Array was the first hint that Moore’s law may have power in the region NKI”.
This is a huge potential. Today our high is several hundred electrodes, capable of measuring about 500 neurons simultaneously is not a million, not even close. If you add 500 neurons every year and a half, we will reach one million in 5017 year. If we double this number every year and a half, we get a million by 2034.
Currently, we are somewhere in between. Ian Stevenson and Konrad Kording published a paper in which he considered the maximum number of neurons that were being read simultaneously at different points over the past 50 years (from any animal) and brought the result on this chart:
This study, which is also called the law of Stevenson, suggests that the number of neurons we can record at the same time, apparently, doubles every 7.4 years. If this figure will hold until the end of this century we will be able to reach million and 2225 year — record every neuron in the brain and get ready for our wizard hat.
In General, equivalent integrated circuits for NRI yet, because 7.4 years — too many for the beginning of the revolution. A breakthrough will be made a device that can record a million neurons, and with the paradigm shift, due to which this schedule is more like Moore’s law and less of Stevenson. Once that happens, will be followed by millions of neurons.
Big obstacle #2: implantation
NRI will not be able to take over the world, if every time to implement them will have to open up the skull.
This is an important topic in Neuralink. I think the word “non-invasive” or “noninvasive” was uttered forty times during my conversations with the team.
Besides the fact that this is a serious barrier to entry and a serious problem for security, invasive brain surgery is expensive and demanding. Elon said the final process of implantation, NCI needs to be automated. “Machine which will be capable of it, should be something like Lasik, automated process — because otherwise you will be limited number of neurosurgeons, and the cost will be too high. You need a car by type of Lasik is to scale the process.”
The creation of the NCI high throughput would be a breakthrough in itself, not to mention the development of non-invasive implants. But the implementation of both items will start a revolution.
Other obstacles
Today’s patients from NKI go with the wire sticking out of the head. In the future, of course, do not fly. Neuralink plans to work on devices that are wireless. But it is also fraught with problems. You need a device that can wirelessly transmit and receive a bunch of data. So it’s got to take care of such things as signal amplification, conversion of analogue to digital, as well as data compression. And it also should work on an induction current.
Another big problem is biocompatibility. Sensitive electronics are usually not very well get along in the jelly ball. And the human body poorly takes foreign objects in themselves. But the brain interfaces of the future will have to work forever and without interruption. Consequently, the device will be tightly Packed and solid enough to survive decades of hum and bias of the neurons around. And the brain — which sees modern devices as intruders and covers them with scar tissue — will have to somehow fooled into thinking that this device is a normal part of the brain.
There is still the problem with space. Where exactly you will place your unit, which can communicate with a million neurons in the skull, which divides the space into a 100 billion neurons? A million electrodes, using modern multi-electrode arrays will be the size of a Golf ball. Therefore, further miniaturization is one ongoing innovation that can be added to the list.
There is also the fact that the current electrodes is mainly optimized for simple electrical simple electrical recording or stimulation. If we really need an effective interface, you’ll need something other than hard single-function electrodes — something with the mechanical complexity of the neural circuits that can record and stimulate, and can also interact with the neurons chemically, mechanically and electrically.
And let’s just agree that it all fit perfectly — broadband, long-term, biocompatible, bi-directional, communicative, non-invasive-implantable device. Now we can have a dialogue with millions of neurons simultaneously. Except… we don’t know how to talk to neurons. Not so easy to decipher static flash hundreds of neurons, but we are, in fact, trying to explore a set of flares that meet certain simple commands. With millions of signals it won’t work. Regular translator, in fact, uses two dictionary, replacing words from one words in the other — but this does not mean to understand language. We need to exercise a powerful leap forward in machine learning before the computer learns the language, and an even greater leap will need to be done to understand the brain’s language — because people definitely will learn to decipher the code a million simultaneously activated neurons.
How simple now it seems the colonization of Mars.
But I’m willing to bet that the phone, the car and landing on the moon would have seemed insurmountable technological problems for people several decades earlier. And I’m willing to bet that it is —
— seem absolutely insoluble for times people do this:
And Yes, it’s in your pocket. If the past is we learned something, it’s that there will always be future technologies that were unthinkable for people of the past. We do not know what technologies that seem to us absolutely impossible, in the future will become ubiquitous, but there will be those. People always underestimate the Human Colossus.
If everyone you know, 40 years there will be electronics in the skull, this will happen through the shift of paradigm which has caused a fundamental shift in this whole industry. This shift is just trying to organize the team Neuralink. Other teams are working on this too, and cool ideas have already begun to appear:
Current innovations in the field NKI
A group at the University of Illinois is developing the interface of silk:
Silk can be rolled into a thin bundle and relatively non-invasively enter the brain. There he theoretically straighten out and settle in the contours, such as shrink wrap. On the silk are flexible silicon transistor arrays.
In his speech, TEDx Talk, hon Yeo demonstrated an array of electrodes applied to the skin like a temporary tattoo, and scientists believe that this method can potentially use in the brain:
Another group is working on a sort of nanoscale electrode neural mesh so tiny that it can enter the brain with a syringe:
For comparison: this red tube on the right is the tip of the syringe.
Other noninvasive methods include the occurrence in veins and arteries. Elon mentioned the following: “the Least invasive method would be something like a durable stent, which enters through the femoral artery and deployed in the blood system to interact with the neurons. Neurons use a lot of energy, so it is essentially a road grid to each neuron”.
DARPA, a division of technological innovation of the armed forces of the United States, thanks to a newly funded program of the BRAIN, is developing a tiny “closed” neural implants, which can replace medication.
The second DARPA project aims to install a million electrodes in a device the size of a coin.
Another idea, which is, what is transcranial magnetic stimulation (TMS) in which a magnetic coil outside the head can create electrical impulses inside the brain.
These impulses can stimulate the target region of the neurons, providing a completely non-invasive type of deep brain stimulation.
One of the founders Neuralink, D. J. SEO has made efforts to develop even more cool interface called “neural dust”. Neural dust is a tiny silicon sensors with a size of 100 µm (approximately equal to the width of a hair), which must be injected directly into the bark. Very close, over the dural sheath, will be placed 3-mm device that can communicate with sensors in the dust by ultrasound.
This is another example of innovative benefits derived from the interdisciplinary team. D. J. explained to me that “there are technologies which do not think in this area, but we can introduce some principles of their work.” He says that neural dust was inspired by the principles of the technology of microchips and RFID. You can easily see the cross influence of the different fields:
Others are working on even more incredible ideas, such as optogenetics (when you insert the virus, which attached to the brain cell, causing it to subsequently stimulated by light) or by using carbon nanotubes, a million of which can be linked together and sent to the brain via the bloodstream.
These people are working on innovation in the company.
Now it is a relatively small group, but when a breakthrough will really start to make itself felt, this will quickly change. Events will begin to grow quickly. The throughput capacity of the brain will become better and better, as the procedure will become easier and cheaper. Arise public interest. And when the public interest will gain momentum, you will notice the opportunity and the Colossus of Human — and then the development speed will jump to the skies. Just as advances in computer hardware led to the development of software and big industry connected to the development of smart applications and advanced machines that will work together with brain-computer interfaces. Sometime in 2052 you’ll be telling some kid about how it all began, and it would be boring.
I tried to get the team Neuralink to discuss with me 2052. I wanted to know what will happen, when will it all come to life. I wanted to know what they would like to put in place of the dash. But it was difficult because this team was created specifically to focus on concrete results and not empty words.
But I continued to ask, gritting his teeth until they outlined their thoughts about the future. I also spent most of his discussions on the topic of the distant future with Elon and Moran CERF, a neuroscientist who works on NKI and thinks a lot about long-term consequences of their work. Finally, one of the team members Neuralink told me that, of course, he and his colleagues have a lot of dreams — otherwise they would not do what you do — and that many things in their region was created under the influence of science fiction. He recommended that I talk with Ramez Naaman, author of the famous trilogy “the nexus”. So I asked the 435 questions Ramez to get the full picture.
On the basis of this conversation I left absolutely killed. Once I wrote what will be, if we go back to 1750 — a time when there was no electricity, fuel or telecommunications — get George Washington and show him our modern world. He’ll be so shocked that they will die. Then I thought about the concept, how many years into the future you need to go to experience a fatal shock from progress. I called it the Deadly Progress (TSP).
Since then, what was the Colossus of man, our world has been strange property over time, it becomes all the more magical. Works TSP. And because the development generates more rapid development, the trend is that over time, TSP is getting closer, shorter. For George Washington, the TSP was a few hundred years, not so much in the overall scheme of human history. But now we live in a time when everything flies by so fast that we can experience a few TSP for your life. The volume of everything that happened from 1750 to 2017, can be repeated already in your life, and repeatedly. It’s a magical time to live and is hard to understand, hard to see, because the life that we live, we live inside.
Anyway, I think a lot about TSP I always wonder what would happen if we went forward in a time machine and have experienced what George would have experienced here. What should be the future that I died of shock? You can talk about such things as artificial intelligence and editing of genes, and I have no doubt that progress in these areas could lead to my death from the shock. But the phrase “who knows how it will be” was never descriptive.
I think I might have finally found a descriptive picture of a shocking future. Let me describe it to you.
To be continued.
Neuralink Elon Musk. Part five: the challenge Neuralink
Ilya Hel