Elon Musk’s brain-computer interface company Neuralink has finally broken its silence. Since the company was formed in 2016, it has kept its plans secret, but in a presentation on Tuesday night it showed off its vision and explained what the firm has done so far.
What is Neuralink building?
At the event, the company unveiled a brain-computer interface – a technology that allows machines to read brain activity. Neuralink says its device will have around 3000 surgically implanted electrodes, each of which will be able monitor around 1000 neurons at a time. The electrodes will be embedded in around 100 extremely thin threads, between 4 and 6 micrometres wide, which is much less than the width of a hair.
The threads collect the measurements from the electrodes and will then be gathered through a small incision behind the ear, where a chip will sit to analyse the results. The information will then be sent via bluetooth to a smartphone app.
Why are they doing it?
Neuralink says the interface could be used for everything from helping people with paralysis to control prosthetics to allowing people to directly interact with artificial intelligence. “This is going to sound pretty weird, but achieve a sort of symbiosis with artificial intelligence,” said Musk at the event.
At the moment, we rely on an interface with technology such as our laptops that is slowed by our fingers or our eyes, and inserting a chip into our brains to speed things up will be key to overcoming that, said Musk.
Will it work?
There is still a long way to go. Many research groups are working on brain-computer interfaces and there has been some progress made in recent years.
One system, called BrainGate, is being trialled for people with ALS or spinal cord injury who have lost control of their limbs. The implant converts brain activity into digital commands that can control prosthetics. People have used similar devices to move a cursor on a screen, play Pong, or even control a robotic hand.
However, controlling the devices normally takes some time to master and involves laboriously thinking about performing a specific action. These thoughts can then be translated into an action, such as selecting something on a screen.
What has Neuralink done so far?
The company has tested the concept in mice and a monkey, though the details are still sparse. With the mice, Neuralink says they used a bespoke robot, which they’ve likened to a sewing machine, to individually insert 1500 electrodes into the brain.
During the Q&A session, Musk also revealed that a monkey has used the device to control a computer. This was later confirmed by a spokesperson, but not elaborated on further.
Musk said that the brain-machine interfaces could be ready for medical trial in humans by 2020. Earlier this year, the US Food and Drug Administration released guidance on what companies will need to prove to get devices such as this approved, and the team says they are working through that with the intention of having these implanted in people with spinal cord damage by the end of next year. Musk said the primary aim of the event was to recruit people to work for the company.
What has the reaction been?
The sophistication of the new technology is “very exciting”, and could progress more rapidly than other smaller competitors with the funding and entrepreneurialism brought by Musk, says Anthony Hannan of the Florey Institute in Australia. However, he is concerned about the idea that this could be used by healthy individuals.
Not only is it dangerous to perform an invasive surgery unnecessarily, but serious caution should be taken with any technology that has the potential to enable somebody else to read or control one’s thoughts or actions, he says.
“My main concern is how the device will prevent infection from getting into the brain along the threads, though I am sure that is something they are working hard on,” says David Grayden at the University of Melbourne.
More on these topics: