Shulou(Shulou.com)11/24 Report--

On Tuesday, local time, biographer Ashlee Vance revealed that thousands of people were waiting in line for a brain implant to be implanted in Neuralink, Elon Musk's brain-computer interface company.

Neuralink is said to have performed 155 operations on sheep, pigs and monkeys in 2021 and 294 last year. But the company has not yet implanted its devices into the human body. Earlier this year, the US Food and Drug Administration (FDA) approved Neuralink to conduct human trials of its equipment, which Musk described as "Fitbit on the skull". In September, the company began recruiting volunteers for its first human trial.

Neuralink plans to perform 11 human surgeries next year, 27 in 2025 and 79 in 2026. After that, the number of surgeries will soar from 499 in 2027 to 22204 in 2030. The company expects annual revenue to reach $100 million within five years.

The following is the full text of translation:

Elon Musk (Elon Musk) is preparing for the most important launch event of his career. But this time it's not electric cars or rocket science, it's brain surgery. Neuralink, Musk's brain-computer interface company, is looking for volunteers for its first clinical trial, which means it is looking for people willing to have surgeons remove a skull so that large robots can insert many electrodes and ultra-thin wires into their brains. When the robot completes the operation, the skull fragments will be replaced by coin-sized computers. The machine's job will be to read and analyze human brain activity and then wirelessly transmit that information to a nearby laptop or tablet.

For this clinical trial, the ideal candidate is a tetraplegic adult under the age of 40. Such patients may implant Neuralink implants into the hand knob area of their premotor cortex, which controls the hands, wrists and forearms. The goal is to prove that the device can safely collect useful data from which parts of the patient's brain, a key step in Neuralink's translating human minds into a series of instructions that computers can understand.

SpaceX's first three rockets exploded, but "We can't screw up the first three human experiments."

Several companies and research teams have created implants that help patients perform basic tasks with their minds, such as clicking on objects on the screen with the cursor. But Neuralink's goals seem more ambitious. In the past four years, starting with the company's first public demonstration, it feels as if there will soon be a ubiquitous clinic where anyone can go in and undergo robotic surgery for about 15 minutes and then come out in a man-machine hybrid. These cyborgs will be able to download information like Keanu Reeves (Keanu Reeves) in the Matrix, or upload their ideas to memory or even to other people's brains. "it sounds strange, but eventually we will achieve symbiosis with artificial intelligence," Musk said in his first demonstration in 2019. " At the time, the company said human trials could begin in 2020.

Setting unrealistic timetables is one of Musk's favorite management skills. To his credit, he finally realized several impossible dreams. But while rockets and cars are important, nerve implants need to be perfect at completely different levels. Companies are in no hurry to bring brain implants to market and expect the best results.

Two other companies, Synchron and Onward, are more than a year ahead of Neuralink in human trials of brain implants and related technologies. However, Neuralink has received more attention, though not all praise, than incremental research (mainly academic research) in previous decades. Some neuroscientists say Neuralink is hyping the technology. Animal rights groups have accused the company of conducting brutal implantation trials on monkeys, pigs and other mammals. The key is Musk, whose increasingly manic and reactionary online image does not make him an ideal candidate for mass production of mind control devices.

All these concerns are justified. However, Neuralink's experiment is also exciting. The company seems to have made great strides in this slow and steady field, and now it has created the most powerful and anticipated brain implant in the world. If the product achieves the desired results, subsequent iterations could miraculously improve the lives of millions of patients with paralysis, stroke, Lugaray's disease, hearing and vision loss. At the same time, the high profile of Neuralink has made investors look for the next Neuralink.

Driven by Musk, the Neuralink team has made great achievements in terms of technology and ambition. As they prepare for human trials, the pressure to succeed is something Musk has never experienced before. After all, it took Tesla many years to mass-produce the car, and the first three SpaceX rockets exploded. Speaking of the brain, "We can't screw up the first three," said Shivon Zilis, director of special programs at Neuralink. "it's not our choice in this area."

The modern history of brain implantation began with technological advances in the 1990s. In general, thinking causes neurons to discharge in specific patterns that are consistent to a certain extent in the brain. In fact, when someone wants to move their arms and fingers, roughly the same neurons are activated whether they can move them or not. When people want to move the mouse cursor and click somewhere on the computer screen, the brain discharges in a similar way. It's the same with speaking: if you can think of saying a letter or word, it will activate the same neurons. Even if you can't speak, a well-trained computer should be able to identify your intentions. In theory, it can speak for you.

The challenge is to find every word that neurons use in English dictionaries, which requires collecting and studying a large amount of data about the signal patterns of many people's brains. To get the clearest signal, you need to place the sensor as close to the neuron as possible. Some researchers try to avoid surgery by placing their equipment outside the human skull, but the effect of increased distance and interference is not significant. The most accurate data usually come from electrodes next to brain cells.

"the long-term goal is to make this technology available to billions of people, unleashing human potential and beyond our biological capabilities."

For most of the past 20 years, the so-called Utah array (Utah Array) has been regarded as the most typical example of brain implants. It is a small square silicon chip that can be placed on a child's fingernails. The wire sticks out from the edge of the chip, and there are about 100 rigid spikes on the surface of the chip. In order to implant the Utah array, surgeons must perform a craniotomy, cut a large hole in the patient's skull, and then gently nail the spikes into the brain. These wires are connected to metal ports, which protrude significantly from the scalp after stitching. After the operation, to use the device, an ice-sized computer was connected to the patient's head.

Researchers have made significant progress in implanting Utah arrays. They use it to read and translate the brain activity of patients with paralysis and other diseases, and use this information to create software that allows patients to communicate with caregivers or their relatives, or to manipulate robotic arms to pick up objects. The problem is that the design of the hardware is so clumsy that it hasn't changed for more than 20 years. The Utah array also requires a large number of trained personnel to operate equipment, as well as a large amount of medical care, which makes them mainly limited to the deployment of research laboratories.

Musk co-founded Neuralink with a $100 million grant from seven scientists in 2016. His investment proved compelling, and his commitment to the technology was irresistible to venture capitalists. Since then, Neuralink has raised more than $500m, including $280 million this year. This attention has also attracted investors' attention to other brain-computer interface projects, including long-standing university programs and newer startups. Last year, 37 such companies raised more than $560 million, according to research firm PitchBook.

Most of these companies share the same primary goal: to create a brain scanning device that can leave the lab. The ideal implant would have sufficient computing power to record and input large amounts of data and transmit data through strong wireless signals. All of this must be done with as little battery power as possible, and the device must not be overheated, otherwise it will irritate or hurt the patient. In addition to hardware, brain-computer interface companies also need machines to learn software skills and conduct thousands of tests.

The implant of Neuralink is hidden under the scalp and flush with the skull. It is also equipped with enough computing power to handle tasks that go far beyond thinking and clicking. In the near future, this idea will make it possible to type at high speed and use the cursor seamlessly. Neuralink is also working on an auxiliary spinal implant designed to restore movement and sensation in paralyzed patients. DJ Seo, co-founder and vice president of engineering at Neuralink, said: "the company's short-term goal is to establish a universal brain interface to restore autonomy to people who suffer from neurasthenia and unmet medical needs. The long-term goal is to enable billions of people to use this technology, unleashing human potential and beyond our biological capabilities."

There have been more than 400 animal implantation experiments in the past two years, preparing for human trials.

Although some competitors are already ahead of Neuralink in human trials, the company's original technology is closest to becoming a general-purpose computer in the brain. The implant has more than 1000 electrodes used to collect brain data, while competitors have only about 16 devices. Neuralink hardware integrates processing, communication and charging system, including battery and signal amplification system. At the same time, competitors still have to wire their implants to bulky battery and amplifier units the size of pacemakers, which are usually surgically implanted into a patient's chest. The Neuralink battery lasts for several hours and can be wirelessly charged in a few hours with a custom baseball cap.

Another of Musk's favorite innovations is the internalization of key manufacturing operations, which increases financial risk but saves time. Neuralink even makes its own semiconductor chips, which is extremely rare in the medical device industry. The company tailored chips for its low-power, low-calorie needs. In Austin, Neuralink turned a bar into a huge implant production line and testing center. In addition to the usual milling machines, lathes and laser cutting machines, the factory is equipped with more bizarre equipment, such as refrigerator-sized cabinets filled with synthetic brain fluid that can heat, cool and squeeze implants to simulate years of wear.

The first task during the operation is to avoid any bleeding or scar tissue in the patient's brain. To that end, Neuralink also built his own surgical robot. The robot is white, about 2.1 meters high, has a stable, cube-shaped base and supports an electronic tower.

Once a human surgeon cuts a hole in a patient's skull, the robot performs an extremely delicate task of inserting electrode wires into the brain. The robot is equipped with a camera, a sensor and a tiny laser grinding needle, which hooks into a ring at the end of each thread. Needles push 64 wires, each with 16 electrodes, into the brain one by one, while carefully avoiding blood vessels. Considering that each thread is only 5 microns thick, or about 1 beat 14 of the diameter of human hair, no one is allowed to try this method. To further avoid tissue damage, the threads are designed to be slender, flexible and strong, and coated with a special polymer to prevent them from deteriorating many years later.

Neuralink's dozen robots performed 155 surgeries on sheep, pigs and monkeys in 2021 and 294 in 2022. For human subjects, surgical preparation and craniotomy are expected to take several hours, followed by about 25 minutes of actual implantation. DJ Seo, co-founder and vice president of engineering at Neuralink, said: "for the past two years, we have been focused on building a product that is suitable for human use, and now it is time to help real humans."

During my first visit, Musk urged the engineers to speed up the progress. He hopes the robot can complete the operation in a shorter time, preferably without the help of a human surgeon. He also wants semiconductor experts to forget what they learned in school and try simpler manufacturing techniques. He wants implants to look smoother and last longer, and maybe everyone needs to rethink everything. Musk considered a series of designs in advance so as not to affect mass production. His attitude seems to give employees confidence that he is right.

Musk's management style has its advantages, which has helped to build the world's most successful rocket company and the most valuable electric car company. Of course, anyone who has searched for "Cybertruck windows" on Google knows that his initial plans can be a mess, and that engineers' efforts don't always lead to rational product choices. To this, Musk's explanation is: "We need to achieve the goal before artificial intelligence takes over, time is very tight!"

22204 human implants will be performed in 2030

Seo and Musk are the only two founders of Neuralink. The other six left voluntarily, or because they disagreed on the direction of the company. Many of them later started similar companies. Seo is in charge of the development of implants and surgical robots, and Jeremy Barenholtz, a computer scientist who graduated from Stanford University two years ago, has become one of the company's chief executives and successfully passed the stringent approval process of the U.S. Food and Drug Administration (FDA). Musk continues to serve as co-chief executive.

In July 2022, the author visited the company's headquarters in Fremont, California, and demonstrated its operating mechanism. Musk gathered executives and engineers to keep them abreast of the latest developments in human trials. Team members gathered at a rectangular desk in the main work area of the office. This is a large open space with people huddled in front of computers, robot prototypes and testing equipment. Musk, dressed in a black suit and Red Bull in hand, stood at the front of the table and began to ask about his competitors. When the employees reported the latest situation, he threw a series of technical problems at them. He is particularly concerned about Synchron, which has been approved by regulators to begin human trials.

The selling point of Synchron is that it doesn't need to cut the skull. The company makes a small product called stents that slide into the blood vessels of the brain through the patient's vascular system. To get the best signal, the hardware cannot be directly connected to neurons, but paralyzed patients already use it to browse the web and send WhatsApp messages. "We should go far beyond that," Musk said. "they are leaving us behind at the moment and I want to catch up next year."

His deputies made faces at the same time. Barenhoz's task is complex, and he wants to explain to Musk that FDA hopes to wait at least a year after the company's first surgery before trying more human implant trials. Musk replied: "unacceptable. If things go well, you will write to FDA and I promise they will take action. It's like SpaceX getting federal approval for more rocket tests!"

At another "update" meeting a few months later, Musk set his sights on Onward, which makes implants close to the spine. Its equipment sends electrical impulses to help rejuvenate muscles and make paralyzed people walk again. Although their gait is often clumsy, it is a miracle for paralyzed people and their loved ones to stand and walk again.

Musk was also preparing for an important demonstration. Neuralink plans to announce that it has begun studying its own spinal implants to match brain implants. This time, Musk seems to be more excited and work harder to push forward with the project.

Of course, none of this is a joke. Engineer Joseph O'Dougherty (Joseph O'doherty) leisurely delivered a long lecture on the early advances in spinal technology in Neuralink, showing a video of an implant used to stimulate pigs' legs to walk on a treadmill. This project involves different parts of the brain, parts of the spine, spinous grids, joint angles, machine learning models, etc. Musk has no formal medical training, but he keeps an eye on progress. He suggested how to adjust the implant, which may reduce the jitter in the animal's gait.

His ideas are sometimes a far cry from reality, but they are often to the point. After an inspection, semiconductor engineers, based on Musk's suggestion, readjusted the process of connecting threads to the company's chips, increasing their manufacturing speed by 50% and reducing defects, according to Zack Tedoff, head of brain interfaces in the chip department. The team studying spinal implants went back to the drawing board, trying to get their pigs to walk in a more realistic way, while Barrenhoz began living in an office to meet Musk's every request.

Musk's view of FDA proved to be correct. Neuralink, which has received applications from thousands of potential patients, recently gave it the green light to conduct more implant trials in 2024 without having to wait for a year-long evaluation period. Neuralink estimates that each implant will cost about $10500, including tests, components and labor, and will charge insurers about $40, 000. The company expects annual revenue to reach $100 million within five years. Neuralink says 11 surgeries are planned in 2024, 27 in 2025 and 79 in 2026. According to documents provided to investors, the number of surgeries will increase from 499 in 2027 to 22204 in 2030.

The treatment of experimental animals has aroused concern.

However, testing medical equipment on animals is not a pleasant thing. To some extent, this practice is regarded as a kind of animal sacrifice on the altar of science, increasing their suffering in order to reduce human suffering. However, Neuralink has come under special scrutiny in its treatment of animal subjects, especially this year. A number of media reports describe in detail the complications, side effects and long-term pain of animal surgery, especially in primates. It is reported that some monkeys who receive implants scratch and tug at their heads until they bleed or show frustration or pain until they are euthanized.

Neuralink admitted that it made a mistake in exploratory surgery, although the company blamed it on human error rather than equipment problems. It stressed that the most disturbing reports came from its early years, when it had not yet established its own testing facility in Fremont and that it had made every effort to provide better living conditions there. "I can always find a way to protect the animals in front of me," says Autumn Sorrells, who manages Neuralink's non-human test subjects and used to be in charge of laboratory animal welfare at the University of California, San Francisco. "We are called 'killers' and 'animal abusers' and then have to come to work and snuggle up to a sheep to make sure they have a good day. It's too hard." Compared with other laboratories, Neuralink animals have larger cages, more choices of food and entertainment, and more social interaction, she said.

I saw the same group of rhesus monkeys living in Fremont for three years, and they all had implants in different parts of the brain. These devices can be removed, some of which have been upgraded to newer models. Seventeen of the monkeys were still active and healthy, and provided brain data for Neuralink at the scene. Three monkeys retired into the shelter, and one of them was euthanized during the planned deathbed procedure.

You must wear gloves, robes, boots, masks and protective glasses before entering the animal facility, and you will get a briefing on how to approach the animal. Inside, the spacious fence is full of toys, artificial trees and playground equipment. Music is often played throughout the facility, the TV is at hand, and most of the programs are nature shows.

The main role of these monkeys is to prove that both equipment and surgical robots work as expected. When they are happy, they also contribute to the company's mind translator by playing computer games.

Neuralink has a room full of monkeys who sit in front of a computer and are reading their minds, which looks weirder than it sounds. For several hours a day, they stared at the laptop screen pushed out of the cage. They can choose games that use joysticks and touch screens (such as tracking letters and spelling words), or games that rely on the brain to control clicks. In one example, a grid of 35 by 35 boxes appears on the screen, and then a box suddenly lights up. The monkey's goal is to move the cursor over the lit box through the mind. With the passage of time, monkeys will complete the task faster and faster, and humans are expected to do the same.

Neuralink is starting to move its animal testing and most of its operations from California to Texas, with a newer and larger park on a ranch outside Austin. The area has surgical facilities, multiple operating rooms, barns, pathology buildings and a sci-fi-themed staff bar. Neuralink says they plan to build an indoor and outdoor space for primates. Today, there are dozens of sheep and pigs there. On a recent visit, the pigs carried small backpacks with batteries to power the patches on their heads and recharge their implants. These animals also have buttons in their enclosures, which they can press with their noses to ask for food or walk outside the barn.

There is still a long way to go from the nose button to completing 22000 human implants a year. Just like a rocket failure, a surgical error or an implant leaking chemicals into a person's brain can set the company back several years. In addition to ensuring basic security, the device must also deliver on Musk's promise. Humans will be able to tell the world about implants, while monkeys can't, including where its limits are. So far, downloading kung fu to your brain and fighting evil super artificial intelligence are still science fiction scenes. Future implants may have 128 or more threads, and the next version of Neuralink's custom chip will extend battery life to 11 hours, Seo said. "our goal is to recharge throughout the day," he said. " At that time, the patient will be able to charge the implant at night through the charging board in the pillow.

For Musk, all of this is hard to achieve. However, if his efforts are rewarded, he will play a huge role in greatly improving the lives of many people. Even if Synchron, Onward or other companies become industry standards, Musk deserves some credit.

Related readings:

"Musk's brain computer company Neuralink is preparing to conduct its first human trial."

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.