The technology is not "stunning", but it can't stop Musk's brain-machine plans from rushing to commercialization.

Musk's "Sebo pig" can't fly.
August 28th, at the launch of the Neuralink Brain-Computer Interaction Project, Musk demonstrated the latest brain-computer interface (BCI) devices - coin-sized hardware, The Link, a microscope electrode about a third of the diameter of hair, Neural Lace, and surgical robots.
on one side of the stage brought the scene to a climax, two of which were tested for brain-machine interfaces.
BCI, founded by Musk in 2016, aims to treat serious brain diseases such as hearing impairment, visual impairment, Alzheimer's, depression, Parkinson's and other conditions that are closely related to brain damage.
its long-term goal is to develop a new generation of brain-computer interfaces to strengthen human and even achieve "human-machine symbicability", such as uploading, preservation of consciousness and so on.
many are excited to predict that the real-life version of Hacking Empire will be coming soon, with cables inserted in the back of their head, computer downloads to learn kung fu, flying planes, and so on.
but at the current technical level, Neuralink is not too "amazing" to the academic community.
, of the University of Sussex, described the technique of recording nerve signals as "decades old."
, a professor of neuro-interface at the University of Newcastle, was less optimistic about the experiment.
" records there is a big difference between brain cells and 'reading thoughts', especially when it comes to high-level cognitive functions that we don't know much about.
" Musk responded to Jackson's statement: "Unfortunately, many people in academia place too much emphasis on the value of ideas and neglect to put them into practice."
, the results show the possibility of Neuralink's future commercialization and mass popularity.
the scale of the portable design and data collection that Neuralink presents is even more exciting for other players deep in the BCI field.
from Tesla to SpaceX, could Musk's commercial path extend to BCI? "We need to first tell what he's saying this time, what's realistic and what's just thinking about it," he said.
", an investor who has invested in SpaceX and founding partner of Fusion Fund, told Euro Health.
standing on the shoulders of giants, but technological breakthroughs have limited several major technological advances announced by Neuralink, with disputes focusing on The Link's chips, pig-related tests and electrode materials.
neuralink's custom-made chip, the N1 sensor, can read brain signals from 1,536 channels and take samples of 20,000 10-bit bytes of resolution on up to 1,024 electrodes, equivalent to about 200Mbps of neural data per channel, according to officials.
, its electrodes are thinner than the hair, and according to officials, its material is soft, minimizing damage.
WechatIMG576.png.pngThe Link in the co-founder of the strong co-founder, Neurosurgery Dr. Lu Qiang, compared to the previous brain interface can only collect dozens to 100 electrodes of signal, currently implanted in the skull of thousands of electrode positions.
is a provider of intelligent auxiliary diagnosis and treatment for neurological diseases.
, Neuralink has the ability to capture high-resolution neuron signals.
but some scholars believe it has not made any significant progress in neurodecoding.
"s speech from Musk, he can feel that he does not pay much attention to the principle of neural coding, his difficulty is not enough to understand, most of his interest in the implantation of the project implementation.
," Professor Hong Bo, an expert on brain-computer interfaces at Tsinghua University, posted on social media.
, a chip that can read nerve signals is not new.
, paradromics, another U.S. brain-computer interface company, is now able to collect and analyze signals as well as reverse feedback, according to Zhang.
, the company has developed a "neural input and output bus (NIOB)" brain-computer interface research terminal.
, NIOB is still in the preclinical development phase and is expected to conduct human trials in 2021 or 2022.
addition, in January this year, zhejiang University's College of Recance completed the exercises of eating, drinking water and shaking hands by implanting Utah electrode arrays into the cerebral cortical layer of high paraplegic patients.
Hodak, president of Neuralink, said at the same day's press conference:
Neuralink was not produced out of no means, in the past academic research has a long history, in a large sense, we are standing on the shoulders of giants."
also showed footage of pigs walking on treadmills and said Neuralink's device could be used to "predict the position of limbs with high precision."
" currently verifies the accuracy of signal acquisition. Huang Xiaoshan, CEO of
" Borikang, believes that compared with the achievements of brain-machine interfaces that have been achieved in previous academic circles, such as human-controlled robotic arms drinking cola and monkey-controlled robotic arms eating bananas, the system only achieves the actual movement intention of collecting and analyzing pigs, and the real control of prosthetic limbs is different.
is the first company in China to industrialization of brain-computer interfaces, focusing on non-invasive BCI technology.
wechatIMG581.png.png predicts the location of a pig's limbs " you can see it is healthy and happy.
" Musk highlighted the state of a piglet implanted in The Link for up to two months to 150,000 viewers who watched it live.
Huang Xiaoshan, an expert in the field of brain-computer interfaces, is expecting data from experiments on primates and humans.
safety and health are key to electrode materials for brain implants.
FDA hopes that medical devices that are not easily removed will remain in the body for at least 10 years, the safety of Neuralink electrodes (flexible polymers) remains in doubt.
if you want to test whether something lasts for 10 years, you really have to wait 10 years, " he said.
," said Matt Angle, CEO of Paradromics.
Musk said Neuralink's award in July of the U.S. Food and Drug Administration's Breakthrough Device Designation, which focuses on medical device development licenses for life-threatening conditions, does not mean the device is fully supported by the FDA.
"the FDA is still reviewing, does not approve Neuralink into human clinical, and needs to do more safety testing on it.
" Zhang said the FDA's Designation and Approval are two different things.
note that Neuralink has a number of technical "black boxes" that have not yet been published in peer-reviewed journals in relation to Friday's update.
As Professor He Bin, head of the Department of Biomedical Engineering at Carnegie Mellon University, says, eventually people need to demonstrate what the system can do to decode brain signals and correctly interpret the state and intent of the brain.
press conference, Musk stressed that his main purpose was to recruit talent for Neuralink.
also tells us that almost all of the technology we're seeing at Neuralink today is a very early example.
BCI treatment is not new, and commercial ambitions for brain-machine interfaces have been used in medicine for 20-30 years, such as the use of deep brain stimulation or the treatment of neurological disorders such as Parkinson's disease through nerve implants.
there are also signs that such techniques may be helpful in treating autism.
neurological disorders, Neuralink says she can also treat addiction and depression.
other scientists are also doing research, such as implanting electrodes in the brain's inner sac location and improving cognitive function by stimulating connections to the pre-cortectal corties.
, a psychiatrist at the University of Minnesota, says about 200 patients worldwide have tried the technique to treat depression.
, opioid addicts implant electrodes into areas of the brain that control addiction.
is not surprising that Neuralink's technology applications are not surprising in the field of disease, neuralink's latest products have demonstrated their commercialization in terms of user-friendliness and engineering.
first of all, device miniaturization is more user-friendly.
Compared to last year's ear-mounted device, this year's device replaces a coin-sized skull on the top of the skull, which is smaller and more portable, and can even be charged in a way similar to a mobile phone's wireless charging (only its official description is yet to be verified).
between last year's and this year's Neuralink devices, "in many ways, it's like a Fitbit on a skull with tiny wire on it."
" Musk made the analogy at the press conference.
brain-machine interface surgery, the head looks almost ind no different from the average person, further optimizing the user experience, thereby helping to lower the data collection threshold and build a data base for the commercial market.
, Neuralink's surgical robots have the potential for large-scale production.
Musk said that in the future, the expected implantation time is reduced to one hour, no hospitalization, or even general anesthesia, like laser myopia surgery, for product development, it is conducive to clear product size, interface type and other development standards, improve the speed of future product optimization iterations. Musk,
Neuralink's surgical robot, also admits that the cost of the device is high, but as the technology matures and scales up in the future, he expects the average person to be able to install a brain-machine interface device for a few thousand dollars in the future, including the cost of surgery.
, Neuralink plans to bring such devices to market around 2023.
" should be a big part of industrialization and standardization than other companies.
, co-founder of Brainland Technology, believes that this is not something that research institutions or universities can do quickly, but also requires resource scheduling in production and supply chains.
" because equipment costs are relatively high, large-scale production can control costs.
" brain land technology is also a domestic focus on non-intrusive BCI technology start-up.
the difference between academics and entrepreneurs is that medical experts say BCI companies are not aware of the pathogenesis of some brain diseases, and that weak basic research is the "layman."
but in Matt Angle's view, building useful BCI techniques doesn't require a perfect understanding of the brain.
know how to record brain signals to control prosthetics and computer cursors, and how to enter tactile data into the brain for prosthesis feedback," he said.
" he said, if you wait until you have a complete understanding of the brain, then establish BCI-based therapy, as if the human gene sequencing is complete before polio vaccine.
In clearly, Neuralink's schedule is moving forward: by the end of the year, brain-machine interfaces will be implanted into the brains of paralysed patients to begin human trials; in the next 10 years, people who have two brain-machine interfaces will be able to communicate directly with brain signals; and within 25 years, "whole-brain interfaces" will be achieved.
after SpaceX and Tesla, can Musk see Neuralink's technology iteration and commercialization in BCI on schedule? Zhang believes that his schedule is somewhat idealized, and that the basic theoretical research of brain-computer interface technology, including neuroscience, still requires a lot of work to make a breakthrough.
"Before he entered the field of commercial aerospace and electric vehicles, the basic theory of rockets and batteries in both fields had matured decades earlier, and his work was more focused on the maturation and low cost of commercial applications, especially in commercial spaceflight.
non-intrusive BCI companies such as CTRL-Labs and Neuroable explore commercialization in areas such as consumer entertainment, but the commercialization of intrusive BCI, represented by Neuralink, will depend on treatment.
BCI's current use in medicine still faces many challenges that cannot be achieved overnight.
first BCI technology has its natural difficult "genes".
have hundreds of billions of neurons in the human brain, and humans now know only 5% of the brain's structure.
even now Neuralink has broken through a thousand levels of neuron signals, compared with the huge total.
, the pathogenesis of many brain diseases, such as depression and Alzheimer's disease, is complex.
" for these diseases can be three-dimensionally directed electrode implantation, in which accurate targeting is the key issue, but not all diseases have a positive, single, accurate target.
" Lu exuberant explanation, to Parkinson's disease, for example, its deep brain electrostational stimulation treatment directly against the hypothyrea, sending out high-frequency electrical impulse signals, stimulate the brain neuronuclear cluster or nerve conduction beam regulation abnormal neural loop.
, the more neuron signals are collected, the more brain-computer interfaces are expected to evolve into whole-brain interfaces.
this is also the ultimate form of brain-computer interfaces that simultaneously monitor the activity of tens of billions of neurons and have the ability to control and modify any of these neuron signals.
Neuralink reduced that number to 1 million neurons, or 1 million neurons in the bandwidth of the interface.
, according to Ian H Stevenson, a professor at the University of Connecticut, moore's law on brain-computer interfaces is that the average number of neurons recorded at the same time doubles in 7.4 years.
at this rate, humans will be able to record 1 million neurons at the same time by 2100 and all neurons in the human brain will be recorded by 2225.
also face problems with biosyndability in terms of the material of the device, and the immune system is automatic.