All you need to know about Musk’s craziest venture so far
Elon Musk is a master of moonshots in tech. His public persona today is built on his incredible ability to turn something outrageously difficult into imminently possible while building a super successful/profitable company around it. But even by his lofty standards, a neurotechnology company that makes implantable brain-machine interfaces (BMI) sounded too crazy. However, it has just been four years since the company was founded and Elon has already done a couple of demos that simply blew the internet away. And with that he has once again forced us to ask how close is this seemingly crazy idea to seeing the light of day?
While Musk is someone who loves the spectacle of a product demo, the ones he has presided over haven’t really told us a great deal so far. Take, for instance, the demo of the monkey playing a videogame with a BMI implanted in its skull. The technology to do has existed for the better part of two decades now and was first demoed in 2002. Musk’s team however has made the connection wireless and possibly established a stronger connection with the use of more electrodes on the brain.
Last year, leading up to an event that was widely covered, Musk promised a live demonstration of neurons firing inside the brain. We were presented with Gertrude, a pig in an enclosure with a brain implant. As the pig moved around and interacted with its surroundings, the activity of neurons in her brain was represented on a graph. Musk had called it “a Fitbit in your skull” then. It was soon met with criticism from neuroscientists who have been recording electrical impulses in brains for years. Musk made claims about curing blindness, paralysis, mental illnesses, see radar with superhuman vision. However, he has not specified any timelines apart from his plans to start human trials by this year. The nature of these human trials is also not yet known. Several neuroscientists believe that solutions to these problems are decades away.
The two demos in themselves, as you must have gathered, didn’t exactly showcase something path breaking. However, the progress on the engineering side of things is definitely commendable.
The implant consists of probes (flexible ones at that so that tissue damage can be avoided) that are composed almost entirely of polyimide, a biocompatible material, with a thin gold or platinum conductor. The probes are inserted into the brain using an automated process performed by a surgical robot designed by Neuralink to ensure that the implant can be attached to the brain as safely and efficiently as possible. Each probe consists of a section of wires with electrodes that capture the electrical signals in the brain. There is a sensory zone where the wires interact with electronics that amplify the brain signal. Each probe contains 48 or 96 wires, which in turn have 32 electrodes each, adding up to 3072 electrodes.
Neuralink claims that it has developed an Application-Specific Integrated Circuit (ASIC) to create a 1,536-channel recording system. There are 256 amplifiers capable of being individually programmed along with analog-to-digital converters within the chip and a peripheral circuit control to serialize the information received. The information is then converted into binary code. As of now, Neuralink has not been able to record the firing of individual neurons as a consequence of larger sized probes, so they can record only the firing of a group of neurons.
Unlike SpaceX or Tesla, Neuralink is taking on the challenge of developing not just a new technology but also the science behind it—that’s partly the reason why Neuralink is looked at with scepticism. Musk has enjoyed all of his success working with science that already existed. Yes, he went on to develop path-breaking engineering solutions for challenges that seemed unsurmountable. But with the brain scientists know precious little about the inner workings or pathways that are used to perform sections. Research in such a discipline could take years or even decades to uncover new truths that would be essential to Neuralink as a company. Musk’s commitment to the project (he has invested $100 million in Neuralink) as well as his star persona could mean that a lot more interest and subsequently money could flow into it. That, Musk would expect, could fast track the project.
But all considered, Neuralink, by the very nature of the challenge it’s tackling, could be something that takes years more before even being close to introducing a viable product. Musk’s tall claims notwithstanding, Neuralink has made very little progress on the science that will drive the technology. A map of electrons in the brain is after all just a map without an understanding of the brain itself.