Anyone with interest in wearable tech has been waiting with bated breath for Elon Musk’s Neuralink to come out of stealth mode. This week, a somewhat nervous sounding Elon and assorted team members delivered a presentation at the California Academy of Sciences in San Francisco to a select in-person audience and thousands live over the Internet. With a 100 plus team already at Neuralink, Elon revealed:
“The main reason for the presentation is recruiting.”
The Purpose of the Brain-Machine Interface
- Elon explained that the central thesis underpinning the work of Neuralink is to:
- Understand and treat brain disorders
- Preserve and enhance your brain
- Create a well-aligned future
How? Through the use of implanted and wearable technology and a slew of AI.
Why Brain Implants?
Neural implants extend from an already established field of implanted wearables including cochlear implants (around since the 1980s) and contraceptive implants (since the late 1990s). For some time, Implanted devices have either kept people alive or improved their quality of life from knee replacements to pacemakers.
The manipulation of the brain through the use of technological devices is accepted practice in serious conditions where medication and other forms of treatment have been unsuccessful. Research into brain-based disorders has led to treatment like Vagus nerve stimulation to treat severe epilepsy. Psychiatric neuro-technology researchers are focused on Transcranial direct current stimulation and Transcranial Magnet Stimulation to treat conditions like chronic pain, pharmaceutical resistant depression, fibromyalgia, OCD, Parkinson’s, and schizophrenia.
Taking this a step further, deep electrical stimulation of the brain (otherwise known as deep brain stimulation) can treat severe OCD, depression and a range of other disorders, enabling health professionals to target the specific parts of the brain which cause particular disorders. Electrodes are implanted into the brain and a device incorporating a battery and pulse generator is implanted into the chest or abdomen with wires to the skull connecting the electrodes. When turned on, the device emits an electrical current that stimulates the neural fibers carrying information from primitive brain areas associated with motivation to the frontal lobe.
Where Neuralink differs is that they claim to have advanced what’s made possible by current brain implants. In short, thinner electrode threads (4 to 6 μm in width (far thinner from human hair), a robot to insert them into the brain and wireless hardware to operate them. They’ve created an array of small and flexible electrode threads with as many as 3.072 electrodes per array distributed across 96 threads. Each thread can be individually inserted into the brain with micro precision to avoid vascular damage and target special brain regions using a neurosurgical robot created by Neuralink.
The robot is capable of inserting six threads (192 electrodes) per minute. The electrode array is packaged into a small implantable device that contains custom chips for low power onboard amplification and digitization. A single USB-C cable provides full-bandwidth data streaming from the device, recording from all channels simultaneously. The implants connect through the skull wirelessly to a device called “The Link”, a Bluetooth radio, and a battery. There is only one battery throughout the system making software and security updates easier.
Elon asserts their offerings thus far differ from the best FDA-approved system used to treat Parkinson’s disease through deep brain stimulation in that is capable of a thousand times more electrodes than is publicly approved and is read and writable.
Further, the implant offers the opportunity for highly customized treatment. Professor Philip Sabes, Senior Scientist noted:
“You can do better with deep brain stimulation if you can record the state of the brain and use that as a guide.” Neuralink’s efforts have the potential to individualize highly focused treatment beyond what is currently possible.
“Getting FDA approval for implants is quite difficult, this will be a slow process where we gradually increase the issues that we solve until ultimately we can achieve a (this is going to sound pretty weird) symbiosis with artificial intelligence”
Neuralink hopes to have the device implanted in a patient by the end of next year. Matthew MacDougall, a neurosurgeon at Neuralink, asserted that the guiding principle at Neuralink is safety: “Will this make me more likely to recommend to my family and friends?” He also said the procedure should in the future be safe enough that to be an elective procedure, akin to LASIK eye treatments.
It’s Stil a Slow Evolution of Incremental Process
It’s a slow incremental process. Those implanted have to learn how to use it, akin to never having a person without arms learn to pick up a glass using a prosthetic — its a long, slow, incremental process, they would first learn to control a cursor on a mobile device, then a keyboard or a mouse. It could in the future, potentially control a 3D avatar of themselves, operate gaming devices, or a range of assisted devices.
The presentation was concluded with a Q&A session. Responding to questions, Elon spoke of keenness to collaborate with the academic community. An audience member asked about the potential of about third-party software running on The Link. It’s an interesting concept considering it’s read and write abilities. Elon noted. “Conceivably there could be some kind of app store thing in the future,” while Max stressed, they not allow any advertising.
In response to a question about how to fund Neuralink if advertising wasn’t an option, Elon noted: “The cost of brain injuries is extremely high to society, so the economy of solving that make a ton of sense and if you enable someone to work and be productive and contribute to the economy the economics of that will easily pay for itself.” In terms of scale, Elon notes that robotics was what made Neuralink feasible — if it has to be done by a neurosurgeon it can’t t be scaled as there aren’t enough neurosurgeons. He also joked “if you want to be symbiotic with AI, I think it’s safe to say you could repay the loan with superhuman intelligence.”
Perhaps one of the biggest surprises was when Elon let slip that a monkey is already able to control a computer with its brain using the device. The team called animal use “the elephant in the room: we wish we didn’t have to work with animals.”
Neuralink Are Somewhat Trumped by DARPA
Neuralink isn’t the only company developing in this area. Over the past 18 years, DARPA has demonstrated increasingly sophisticated neurotechnologies that rely on surgically implanted electrodes to interface with the central or peripheral nervous systems. The agency has demonstrated achievements such as neural control of prosthetic limbs and restoration of the sense of touch to the users of those limbs, relief of otherwise intractable neuropsychiatric illnesses such as depression, and improvement of memory formation and recall. Earlier this year, DARPA awarded funding to six organizations to support their Next-Generation Nonsurgical Neurotechnology (N3) program. Their goal is to develop high-resolution, bidirectional brain-machine interfaces for use by able-bodied soldiers. These wearable interfaces could, in the future, include control of active cyber-defense systems and swarms of drones or teaming with computer systems to multitask during complex missions.
The N3 teams are pursuing a range of approaches that use optics, acoustics, and electromagnetics to record neural activity and/or send signals back to the brain at high speed and resolution. Teams are pursuing either completely noninvasive interfaces that are entirely external to the body or minutely invasive interface systems that include nano-transducers that can be temporarily and nonsurgically delivered to the brain to improve signal resolution.
But at any rate, if Neuralink sounds like something you might want to get involved in, the company is hiring for the following sectors: