When news hit yesterday that serial entrepreneur and futurist Elon Musk was investing in a brain chip venture called Neuralink, the billionaire’s fan club went wild. Theories ranged from this being Musk’s bold plan to forestall an AI apocalypse to more measured responses about this being a promising contribution toward curing neurodegenerative diseases. Some also, only half-jokingly it seems, accused Musk of having played a little too much Mass Effect: Andromeda over the weekend.
But to unpack Neuralink’s purported goal a little further, you have to understand what is and is not currently possible in the realm of neuroscience, and why Silicon Valley is putting more time, money, and energy into exploring cognitive enhancement. Neuralink isn’t the first company to look into what are called brain-computer interfaces, however it is perhaps the most high profile now that Musk’s name is attached. Other companies, like Braintree co-founder Bryan Johnson’s new company Kernel, are also looking into ways to improve human cognition.
The ultimate goal here is not necessarily to prevent some robot uprising or keep AI “shackled” to humans as a form of control. As neat and sci-fi as those concepts appear, they are far more outlandish and irrelevant to the conversation Musk, Johnson, and other futurists are actually participating in right now. What Neuralink and Kernel are trying to do is take the first steps toward hacking the brain, so to speak, so that human beings can in the future stay healthier for longer and potentially enjoy the benefits of treating the human brain like a computing platform.
This means using a chip inside the skull or some other electronic device that could improve our memory and our ability to perform complex mental tasks, as well as increase speed at which we could communicate with one another. It could even allow us to directly link with the cloud and other forms of internet infrastructure. This all sounds impossible, and right now it is.
The only concrete conception of these ideas comes from science fiction, like cyberpunk classic Ghost in the Shell or the Deus Ex video game series, in which it’s taken for granted that society has figured out most of the hard problems and ironed out all the kinks. These fictions often treat the arduous cycle of scientific progress as a footnote or a flashback, glazing over out of necessity the decades of medical research required to merge humans with machines.
As it stands today, brain implants containing multi-electrode arrays are used to try and ameliorate the effects of Parkinson’s disease and epilepsy, while less complex, stimulating devices are used for people suffering from Tourette's, eating disorders, and depression. But these techniques are still experimental compared with more mainstream medical procedures. Because operating on the brain is difficult and dangerous, even the most cutting-edge researchers still openly admit to not having the necessary tools or data to fully comprehend how the organ operates or how to repair it when it starts to malfunction.
So Neuralink and Kernel are trying to accelerate progress in this field using a mix of financial resources and a kind of brain trust approach to innovation. The idea is that if you put enough talented people with enough money in one place, you can achieve breakthroughs that otherwise would take years for traditional research organizations. It appears both companies understand that these advancements must happen first in the medical field; The Wall Street Journal reported yesterday that Neuralink was registered in California as a medical research company.
So that means we’re a long way from even discussing the merging of the human brain with AI. Down the road, perhaps decades from now, that will become a possibility. (Although it’s way less clear how that would prevent some sort of robot uprising.) But it’s necessary to first understand how the brain works and what makes it fail.
If Musk and Johnson's companies can do that, they just might gain some insight into how to improve the brains of healthy people. Of course, then you’d have to pioneer surgery methods that would allow people to get brain surgery without it being incredibly invasive and dangerous. Right now, people only willingly undergo this type of surgery as a last resort to give movement to a prosthetic limb or prevent severe seizures. It’s not uncommon to see people online volunteering for this type of implant, but there is absolutely no realistic scenario in which someone without a severe disease could receive a brain chip for the foreseeable future.
That doesn’t mean that Musk and Johnson aren’t concerned about AI or haven’t taken its rise into account when thinking about the goals of Neuralink and Kernel. Both entrepreneurs think that improving human cognition is the surest path forward for humanity. “I think if humanity were to identify a singular thing to work on, the thing that would demand the greatest minds of our generation, it’s human intelligence,” Johnson told me in an interview earlier this year about Kernel, “specifically, the ability to co-evolve with artificial intelligence.”
Johnson sees the advancement of AI not necessarily as an existential threat so much as it’s an opportunity to start thinking about the ways humans could be improved by software just as algorithms and machines are constantly modified and upgraded. In my conversations with him, he told me the human brain can and should be treated like the genome. “What I wanted to do was work with the brain the same way we work with other complex biological systems like biology and genetics,” he says.
Musk is a little more apocalyptic in his statements about AI. It doesn’t help that he often to likes to joke onstage about his pontifications, like whether humanity is in fact living inside a giant simulation. In the last couple of years, Musk has made headlines for comparing AI to “summoning the demon” and calling it the greatest threat to mankind — “potentially more dangerous than nukes,” he once said. He’s insinuated that it’s partly why he wants mankind to become a multi-planetary species, in preparation for an AI-related doomsday event. He’s also pointed a finger at Google as a company he fears might develop AI that runs rampant and turns on humanity one day.
This sounds like the oddball fears of a futurist billionaire, and in part they are. But when you look at Musk’s actual investments, he is very much concerned with allowing AI to develop while ensuring it does so transparently and without malicious intent. The entire purpose of OpenAI, a nonprofit Musk funds alongside Y Combinator president Sam Altman, is to promote open-source research in the field and ensure that AI be developed while taking into account social, philosophical, and moral issues. More recently, Musk gave a very measured preamble of sorts about Neuralink, not mentioning the company by name but telling a crowd in Dubai, “Over time I think we will probably see a closer merger of biological intelligence and digital intelligence.” This is not so humans can go to battle against the singularity, but more so that organic life doesn’t get aggressively outpaced by synthetic life.
So it’s important to keep all of these considerations in mind when discussing Neuralink, Kernel, and the scores of brain-computer interface companies that will undoubtedly sprout up in their wake. It’s fun to talk about fighting Skynet or preventing The Matrix, but Musk and other entrepreneurs in Silicon Valley do have more realistic concerns and are, at least right now, more open to admitting that the more near-term goal is combating diseases and funding neuroscience research. It doesn’t grab headlines quite like combatting the extinction of the human race, but we should all rest easy knowing that we’re not quite there yet.