Neuralink – innovative neuroscience or just bold claims?
Elon Musk insists Neuralink will allow AI melding with brains with little basis in research.
By Bārbala Ostrovska
Photo courtesy from Hal Gatewood from Unsplash.
Elon Musk is mostly known for his work in companies like Tesla and SpaceX, but the field of neuroscience has not escaped the billionaire’s ambitions either. Neuralink is one of his newer companies, founded only in 2016. Its main goal is developing two bits of equipment – a chip that could be implanted in a person’s skull, with electrodes fanning out into their brain and a robot that could automatically implant the chip. Both to “achieve symbiosis between the human brain and artificial intelligence.” Recently, the company has been under fire due to allegations of animal abuse. Over 700 pages of documents relating to monkeys used in research were filed by the Physicians Committee for Responsible Medicine. Neuralink has denied all accusations, but admitted to euthanizing eight animals during trials. With the magnifying glass pointed at the company, it is worth examining whether its science is as revolutionary as Musk likes to claim.
Neuralink has had two high-profile presentations so far. The first was advertised as being a ‘mind-blowing demonstration’ of neurons firing inside a living animal. It turned out to be a fenced enclosure with three pigs. One of the pigs had an implant in its brain and speakers were playing recordings of the animal’s neurons firing in real time. Far from ground-breaking science, it has been possible to hear electrical impulses from animal (and even human) brains for decades. The second demonstration came a couple of years later. It was a video of a macaque monkey playing the game /Pong./ At first it was using a joystick, receiving banana smoothies as a reward for playing correctly, but after a while the monkey could play by using only its brain. A monkey playing a game with its mind might seem like an exciting development. However, this is not anything new either. There are systems which already allow people with paralysis to use table computers at home and people with robotic arms to control their limbs. The first case of moving a cursor on a screen with a brain-computer interface was accomplished in humans in 2006.
Andrew Schwartz, a University of Pittsburgh neurobiologist, spoke on this saying: “Technically, from an engineering point of view, the recording equipment and the transmission of data seems to be very nice and that may be an advance, but in terms of the performance and the actual use of brain control, it is very rudimentary.” With this in mind, Musk’s claims about the possibilities of melding human consciousness with artificial intelligence, saving and replaying memories and even telepathically summoning your car seem severely far-fetched. "Not to say that that won't happen, but I think that the underlying neuroscience is much more shaky. We understand much less about how those processes work in the brain, and just because you can predict the position of the pig's leg when it's walking on a treadmill, that doesn't then automatically mean you'll be able to read thoughts," thinks Professor Andrew Jackson from Newcastle University. These presentations build hype, but don't do much in terms of advancing research. The video does not include any analysis of the implant or the monkey’s performance and there are no scientific publications to be found on Neuralink’s work with animals. As discussed by John Krakauer, a professor of neurology at Johns Hopkins University, “the device we saw was placed over a single sensorimotor area. If we want to read thoughts rather than movements (assuming we knew their neural basis) where do we put it? How many will we need? How does one avoid having one’s scalp studded with them? No mention of any of this of course.”
“To get any of these devices into your brain [...] is very, very high-risk surgery," said Dr Rylie Green from Imperial College London.
Amid the animal mistreatment allegations, Neuralink is looking to expand its research to human volunteers next year. Although the company has made claims about starting human testing in 2020 as well, they have not been able to receive approval by the US Food and Drug Administration to this day. Even if the science was fault proof, there are multiple ethical problems with performing brain surgery for anything other than essential treatment, which would need to be done if the company ever comes close to any of the goals claimed by Musk. “To get any of these devices into your brain [...] is very, very high-risk surgery," said Dr Rylie Green from Imperial College London. “People do it because they have severe limitations and there is a potential there to improve their life. Doing it for fun is not a great idea." What about medical applications of the technology? The ethics and implementations of those are not much clearer. Elon Musk has made dubious claims that the implant could “solve” autism and schizophrenia. But let’s imagine those are just his personal musings and focus on official company statements. The company website states that the main application explored at this time is helping people with quadriplegia (partial or full paralysis in all four limbs). This seems like a much more grounded goal. Despite that, it is hard to analyse how well Neuralink is achieving it so far. The only thing close to a research paper published by the company is a “white paper” (an informational document issued by a company to explain and promote their product) with the only authors listed being “Elon Musk; Neuralink.” No credit to any researchers working on the project or any mention of specific treatment plans. It is unclear how serious Neuralink is about treating disease at all since in many interviews Musk has strayed away from medicine to a more futuristic use of the device, which he calls the company’s overall aim.
It can be exciting to hear about possible brain-computer interface innovations proposed by Neuralink and it can even help to bring more attention to other ongoing research, which is great. However, neuroscience is still a very young field, and it must be understood that progress will involve rigorous research and collaboration of the scientific community, not one billionaire who’s only interested because he is afraid artificial intelligence will become “smarter than the smartest human on Earth."