I will show you that it is actually working.
Neuralink is developing interface technology that connects the brain and a machine.Founder, Eron Musk, has finally announced on August 28 a moving device of Neuralink.
Live WebCast of WORKING @NEURALINK DEVICEFRIDAY 3PM Pacific Https: // T.Co/Poulbrgzfu
— Elon Musk (@elonmusk) August 26, 2020Moving Neuralink device, live web cast
Friday 3:00 pm, US Pacific Time (Translation: Japan Time August 29 7:00 am) https: // T.Co/Poulbrgzfu
It is a science fiction itself to connect the brain and a computer directly, but research in this field has been progressing rapidly in recent years.However, it is still a long way off that you can send an email just by thinking in the brain.Unlike startups such as SPACEX and TESLA, which Musk has created, there are various restrictions on the development and implementation of neuralink, and it should not be so exploded.However, I understand that this is theoretical, how much it can be realized (it doesn't seem), and that there are other people who do the same thing.I will summarize it in the article.
Neuralink, which was announced by Musk in 2017, is said to be "connected to humankind with humanity" using the "Superhiro Binding Brain Machine Interface".Connect the human brain and computer with a chip that embeds the brain.
Neuralink's brain machine interface aims to help you treat brain abnormalities such as Parkinson's disease, epilepsy, and depression.It is also said that by combining with advanced support devices, it can operate prosthesis etc. by thinking of the person who wears it.However, if Mr. Musk's ultimate vision can be realized, for example, Esper, for example, to operate equipment that is not directly connected to the body with his mind, send his thoughts directly to another person's brain, expand his cognitive skills, and enhance his intelligence and memory.You will be able to do it ...If it could be really realized.
More concepts, Neuralink for Musk is a means to avoid the end of the world due to AI (artificial intelligence), and in 2017, this technology would "symbiotic with AI".。The idea is that by attaching AI to enhance the power of AI to the small brain of humans, human beings will catch up with evolved technology.It's already a bad idea to win AI, so it's the idea of incorporating that power into yourself, but there is a problem that it will be so convenient even if it is technically incorporated.
Also, the idea of embedding something embedded in the brain and connecting with an external machine is not a new one.In science fiction, this kind of story has been made for decades, and it is drawn in various forms, such as William Gibsun's skull jack, Ian M. Banks neural lace, and brain plugs in the movie "Matrix".I have come.
It is really exciting that such a novel or in a movie is actually made, but the problem is that unlike electric vehicles and rockets, the brain computer interface is positioned as a medical device.In other words, Neuralink also needs to be approved by the United States Food and Drug Administration (FDA) for clinical experiments and commercialization of equipment used for humans.
Like a general pharmaceutical company or a medical device manufacturer, Neuralink has to prove the safety and effectiveness of the product, and it is common for a detailed and time -consuming clinical experiment.In the case of Neuralink, there is a unique difficulty in the sense that chips are being embedded in the brain of a healthy person, and it will take ten years.There are only a few therapeutic parts of the disease among what Neuralink is aiming for, and the rest is human functional expansion, so it will be troublesome.
Despite these hurdles, researchers in this field have already advanced significantly to embody the science fiction world.Most of the media's attention is gathered by Mr. Musk, but more sober researchers have created new technologies and have revealed their evolution.
For example, in 2019, the Neuroscient Team of the University of Colombia has succeeded in converting brain waves into utterances, and the team at the University of California San Francisco is a physical, such as a chin or lip when humans speak.Focusing on the brain waves related to movement, we have a simulation that moves a virtual voice (organs that speak out) based on the brain waves.In 2016, the person who cut his hands could also move his righteous finger via a brain implant.There are also research results such as the external skeleton controlled by thinking, the restoration of the tactile sensation, and the recovery of the lost athletic function.In addition, experiments directly communicating with human brain vs. brain have also begun in the early stages.
Good results are also available in animal experiments.In a large place, monkeys can operate wheelchairs using a wireless brain machine interface, and monkeys with brain implants can type 12 words per minute.
So Mr. Musk's Neuralink is not particularly groundbreaking in that direction.However, if there is a different point, it means that the size, funds, what Neuralink is aiming for, and of course the charisma called Elon Musk is doing it.However, this field has a rival outside the academic world, and Facebook recently acquired the startup Ctrl-Labs of the Nerve Interface for $ 1 billion (approximately 530-106 billion yen), and BRAINTREE's founder Bryan Johnson.He launched a project called Kernel by throwing $ 100 million (about 10.5 billion yen).And the US military DARPA has also invested $ 65 million in this field.The competition in the brain machine interface field is gradually becoming intense, and it is still unknown whether Neuralink can be successful.
According to Wall Street Journal in 2017, the Neuralink system uses the technology called "Neural Race" (a tribute to Ian M. Banks).This is probably to connect the brain to an external computer via a brain implant (mesh -shaped).
Neuralink has been invested $ 158 million (about 16 billion yen) as of 2019, and New York Times reports that $ 100 million (about 10.5 billion yen) is a private property of Musk.There are now about 90 employees, and the Neurosurgeon at Stanford University may be cooperating, and other research institutions may enter.
Neuralink says that it takes step -level approaches, so you should start with brain disease and gradually scale up to the extension.It seems that expanding the band of information that can be read from the brain will be super important, so it is almost certainly necessary to have a brain implant instead of a non -hit method like brain wave tests, and for that purpose, surgery, durable and flexible.You also need a biological compatible component.
In 2019, Musk announced further details of Neuralink at the California Science Academy.Their method, which was specified in Neuralink's white paper, was to embed an extra -fine thread or probe in the brain using a robot like a dedicated sewing machine.This thread is only 4 to 6 micrometers thick and is much thinner than the thickness of 50-100 micrometers.
This extra fine thread is embedded, which leads to the chips in the head.As shown in the white paper, this robot allows you to embed six threads per minute or 192 electrodes.Neuralink also has a "96 polymer thread with 32 electrodes, a total of 3,072 high -speed embedding of 3,072 electrodes".Even if this robot has this robot, you will need a brain surgeon, but according to New York Times, Neuralink's president Max Hodak also aims to make the same task possible with a laser and do not make a hole in the head.increase.
Neuralink has already announced a system that can read information from 1,500 electrodes embedded in the brain, which was a rat brain.Nevertheless, it was 15 times the performance than a similar system for humans.
"It's wonderful that they have realized so quickly, and it's interesting how far they can go," commented Andrew Jackson, a professor at Newcastle University and specializing in neural interfaces."What Neuralink is doing is one of the many initiatives trying to read the electrical activity of a large amount of brain cells. The Neuralink approach uses a certain sewing machine to insert a lot of flexible polymer threads into the brain., The thread is connected to an electric component embedded under the skin. "
Jackson also introduced other approaches using small silicon needles.Tim Harris and others from Naelia Research Campus Naelia Research Campus (about 5.5 million dollars (approximately 5).In a joint project that spent 800 million yen, we have created a probe called Neuropixels, which can record more than 700 neurons at the same time.There is also a concept called "neural dust", which allows small wireless implants to the entire brain.
"You can only understand whether the" horse "on which Mr. Elon rides was right," said Jackson."Surely, it is possible that the entire neural interface field can evolve by providing commercial funds. Neuroscientian has recently used a fairly old -fashioned facility for brain records, so this interest from Silicon Valley.I am very grateful to be able to fund. "
Initially, Neuralink should have already started experiments for humans, but not yet.They may have been overwhelmed when making a schedule, and may not have been approved by the FDA.Neuralink was interested in creating his own animal experimental facilities in San Francisco, emphasizing the need for animal models.
Kevin Warwick, an emeritus professor at the Coventory University and Ladding University, a brain machine interface expert, evaluates the use of polymer probes.It's not just the polymer is flexible.
"It is very good to be able to make various patterns of electrodes. It will be useful in structural aspects in the sense that it is hard to break," he explained by email."The problem is how to insert it into the brain, but they designed an embedded robot."
Warwick says that the explanation of the embedded robot is very fluffy in the white paper, but this is the key to this robot."If Neuralink says, if you can embed a robot for embedding, you should be able to bury a lot of electrodes in multiple brains. But from my point of view, it really requires proof. They really do such a polymer.Can you prove that you can use a robot to be used in your brain, safely and safely and accurately? "
From now on, Neuralink's team and anyone who develops the neural interface need to overcome many major issues.For example, the invasiveness of this technology (holes in the head), a universal method that maps the signal of the brain (each system needs to learn the individual specificity), the size of the necessary experiments is safe and safe.How to spread it in an ethical and effective way (for both animals and humans).
For example, there will be unexpected problems, such as unusual fever of implants, or embedded devices are rapidly outdated.is.After that, it seems difficult for researchers to judge whether the data taken out of the brain can be used for something useful or whether it will be money.There are still many things that I don't know about the human brain and its mechanism, so it may be impossible to think that the way we are thinking about can go smoothly as it is.
I'm looking forward to what Musk will make on August 28, but I can't get on it yet.Considering the complexity of the brain machine interface, progress will be slowly and gradually.