Elon Musk unveils plan to connect human brains with machines

Elon Musk and his start-up Neuralink unveil plans for futuristic technology linking human brains to computers.

Elon Musk is bankrolling much of Neuralink’s work himself. Picture: AFP

Tech billionaire Elon Musk and top-level scientists from his neuroscience start-up Neuralink have unveiled plans for brain-reading “threads” and a robot to insert them, with the secretive company taking steps towards building a next-generation brain-computer interface.

Mr Musk has given the public a first glimpse of his futuristic technology, which he says would connect human brains and machines with more precision than other available devices.

Neuralink has been developing the technology for roughly two years, and is busy putting together a submission to the US Food and Drug Administration to start testing the technology in humans. To date, the devices have only been tested on rats, and monkeys.

Mr Musk, who is also the CEO of SpaceX and Tesla, is bankrolling the majority of the endeavour himself, pouring in about $US100 million ($142.63m) of the $US158 million raised so far.

Neuralink’s goal is to use the platform to treat neurological conditions like movement disorders, spinal-cord injury and blindness, and it’s looking to have its first human patient with its technology before the end of 2020. At first, the company wants humans to be able to control mobile phones and computer keyboards via the brain.

“We hope we’re less than a year from the first safety study on the order of five patients,” said Neuralink President Max Hodak, in an interview. He emphasised that it could take years before the device could help a range of patients. “The road is long,” he said.

Neuralink is one of several companies, including Facebook, Kernel, CTRL-Labs and Paradromics, trying to build neural interfaces for clinical and nonclinical applications. In recent years, neurotechnology development has been spurred by public and private investment, including the US Brain Initiative, which was started by the Obama administration in 2013.

The goal of many of these projects is to access as many neurons as possible because that would give scientists more precise reads on activity that underpins walking, speech and mood, among other brain functions. They can then turn neural recordings into electrical signals that can be fed into a robotic device or back into the nervous system to produce movement or vision to help patients, according to experts.

At an event in San Francisco, Neuralink described a tiny probe with nearly 3100 electrodes laid out across about 100 flexible wires, or threads, each individually inserted into rat brains by a custom-made surgical robot. The device can monitor the activity of upward of 1000 neurons at a time, according to the company.

The sewing-machine-like robot can target very specific brain areas, helping surgeons avoid major blood vessels—an important consideration for minimising inflammation and long-term damage, according to a paper from the company. Data were processed and analysed by proprietary chips and software.

Neuroscientists and neurotechnologists said that a platform that can insert tiny electrodes robotically throughout the brain and then analyse activity with custom software is exciting, but cautioned it is too early to tell how quickly Neuralink’s device could safely be used in patients.

“If you’re trying to walk yourself toward human prosthetics, this is a more promising direction than currently available technology,” said Tim Harris, a senior fellow at the Howard Hughes Medical Institute’s Janelia Research Campus and a developer of research-grade neural interfaces. Among the questions that are left unanswered by the company’s paper, he said, is how long it lasts in the brain.

“If you’re going to do this for people, you should be aiming for at least five years, minimum,” he said. “To do an implantation surgery of this level of intricacy, a year or two is not enough.”

The paper, which wasn’t peer-reviewed, didn’t include data on the long-term stability of recorded neural signals nor the brain’s inflammatory response.

“That is utterly critical” before any device can advance to human trials, said Loren Frank, a University of California, San Francisco neuroscientist developing brain-computer interfaces.

Neuralink has said it is doing those experiments but isn’t ready to make the data public.

The device, in theory, was designed to also stimulate brain cells, but “we have not demonstrated these capabilities here,” according to the paper. Direct brain stimulation with implanted electrodes is a longstanding approach to treating movement disorders and epilepsy. Most brain-computer interfaces are so-called open-looped systems that don’t adapt to a patient’s needs and experience. Neurosurgeons and technologists have pointed to that drawback as reason why brain stimulation hasn’t worked for treating mood disorders.

The advantage of a system like Neuralink envisions would be its ability to analyse recordings using machine learning and to adapt the type of stimulation it delivers to a patient’s brain, according to the company and other experts.

Enke Bashllari, a neuroscientist by training who now heads venture capital firm Arkitekt Ventures, said there are considerable safety measures that must be ensured with implantable technology and said she believes that non-invasive devices also have a valuable role to play in augmenting human movement or cognitive performance. But she said the highest unmet medical needs will require technology that goes inside.

“It has to allow for two main things—high spatial resolution and high bandwidth. It has to interface with millions of neurons at the same time and you need to know exactly which neuron is firing,” she said. “That can currently only be done invasively.”

Additional reporting: The Wall Street Journal