Billionaire entrepreneur Elon Musk has unveiled another high-concept project, one that he says puts humans on a path to “symbiosis with artificial intelligence.” In other words, allow them to control computers with their minds.
His start-up Neuralink said it has tested its elaborate interface of ultra-thin “threads” — which connects with a chip embedded in the skull — on rats and wants to begin human trials as soon as next year. The company said the device could eventually be used to treat such neurological disorders as paralysis and blindness.
Musk and the company outlined their plans, which require Food and Drug Administration approval, during a presentation Tuesday at the California Academy of Sciences in San Francisco. He said the trials would be done in partnership with neurosurgeons at Stanford University and other academic institutions. The procedure involves drilling four tiny holes in the skull to insert the threads. Eventually, Musk said, the goal is to implant the device with less-invasive laser surgery.
Musk and other Neuralink officials said the technology is in very early stages and that the road to reality will be full of rigorous research and regulatory hurdles.
“It’s not like suddenly we will have this incredible neural lace and will take over people’s brains,” Musk told the audience. “It will take a long time.”
Tuesday’s event, the company’s first public presentation since its 2017 debut, was primarily geared toward recruiting top-tier talent for the company, which has about 100 employees. The start-up has raised about $158 million since its inception, with Musk himself contributing $100 million.
Max Hodak, Neuralink’s president, said the company knows it will need to collaborate with academic and scientific communities.
“We are under no illusion that we can do all the scientific research ourselves,” Hodak said.
An advanced, “sewing machine”-esque robot would embed the device’s tiny, flexible threads — each a fraction of the width of a human hair and collectively outfitted with thousands of electrodes — deep in the brain without doing major damage to the surrounding tissue. Once in place, the threads would relay information to a chip, which in turn would transmit it to an external Bluetooth-like device worn behind the ear, Musk said.
If successful, the system would mark a significant advancement from the stiff needles used in current technology, which allow for far fewer electrode channels, degrade over time and are disrupted when the brain shifts in the skull.
“Neuralink picked the best of existing lab technology and pushed it forward in a number of important dimensions, and most impressively has an integrated implantable product that goes beyond the current state of the art,” tweeted Andrew Hires, an assistant professor of neurobiology at the University of Southern California.
Neuralink released an unpublished research paper outlining the company’s progress, but it wasn’t peer-reviewed, as is standard.
In Neuralink’s first demonstration in front of a reporter, Bloomberg reported, it showed how the system had been implemented in a lab rat, allowing it to receive information from 1,500 electrodes — 15 times better than current systems employed in humans, according to the New York Times.
In the research paper, Neuralink said it completed 19 surgeries on animals, and that robots placed the threads accurately 87 percent of the time. During the question-and-answer portion of the presentation, Musk hinted at successful testing on animals much closer to humans at the University of California at Davis.
“A monkey has been able to control a computer with its brain” using Neuralink’s interface, Musk said. He did not provide further detail, and the company has not put forth any evidence to this effect.
Musk’s vision of “superhuman cognition” borne from humans merging with artificial intelligence rivals that of his other projects — colonizing Mars, hyperloops, life-size Lego buildings — in ambition, but Neuralink is one of a handful of start-ups working on brain-computer interface technology. Other companies, like Kernel, CTRL-labs and Paradromics have been puzzling over ways to help human brains commune with machines, but not all are focused on technology implanted in the brain.
One agency in the Defense Department that specializes in emerging technology for the military is funding a slate of brain-machine interface projects at institutions like Carnegie Mellon University and the Johns Hopkins University Applied Physics Laboratory.
The FDA released regulatory guidelines for brain-computer interfaces in February, with the explicit aim of fueling development of the technology.
“These devices have the potential to benefit people with severe disabilities by increasing their ability to interact with their environment,” FDA Commissioner Scott Gottlieb said in a statement accompanying the guidelines designed to “help potential developers of these novel devices address this area of unmet medical need and promote innovation while maintaining appropriate patient protections.”
Though Neuralink’s interface was framed as a possible cure for patients with serious, unmet medical needs, the company’s chief neurosurgeon, Matt MacDougall, said Neuralink’s long-term goal is for the system to be available through an elective procedure.
Musk, meanwhile, told the audience that he sees Neuralink not only as a medical solution but also as a means for humans to keep from being “left behind” by the artificial intelligence he believes will eventually overtake humanity.
“With a high-bandwidth brain-machine interface, we can go along for the ride,” Musk said, “and effectively have the option of merging with AI.”