Musk’s Neuralink venture promises to open a ‘working’ brain computer device


By Tina Bellon

Aug 28 (Reuters) – Billionaire entrepreneur Elon Musk’s neurological startup expects Friday to detail his latest innovations for implanting human computer chips in human brains, raising expectations among scientists who are closely monitoring the company.

Neuralink was co-founded by Musk in 2016, and aims to implant wireless interfaces for brain computers that include thousands of electrodes in the most complex human organ to help cure neurological conditions such as Alzheimer’s, dementia and spinal cord injuries and eventually it fusing humanity with artificial intelligence.

The company said it will provide an update on its work during a live webcast late Friday afternoon, with Musk tweeting that the presentation will include a “working Neuralink device”.

Musk, who often warns of the risks of artificial intelligence, is no stranger to industry revolution as CEO of electric car company Tesla Inc and aerospace manufacturer SpaceX.

During a Neuralink presentation in July 2019, Musk said the company intended to receive regulatory approval to implant its device in human trials by the end of this year.

“This has a very good purpose, which is to cure major diseases – and ultimately to secure the future of mankind as a civilization relative to AI,” Musk said at the time.

The company promises to implant a sensor about eight millimeters in diameter, or smaller than a fingertip, potentially under local anesthesia alone. With the help of a sophisticated robot, flexible threads or threads smaller than a human hair are implanted in brain areas responsible for motor and sensory functions.

Neuroscientists say that although Neuralink’s mission to read and stimulate brain activity in humans is possible, the company’s timeline appeared too ambitious.

“Everyone in the field would be very impressed if they actually showed data from a device implanted in a human,” said Graeme Moffat, a research fellow in neuroscience at the University of Toronto.

Small devices that electronically stimulate nerves and brain areas to treat hearing loss and Parkinson’s disease have been implanted in humans for decades.

Neuroscientists have also conducted experiments with brain implants with a small number of people who have lost control of body functions due to injuries to spiral cord or neurological conditions such as strokes. People in those trials were able to control robot members or small objects, such as a computer keyboard or mouse pointer, but still performed sophisticated tasks.

Most of today’s leading research in brain machine interface is conducted on animals, scientists note, with safety challenges and lengthy regulatory approval procedures that prevent larger human trials.

Brain-machine interface science has seen a rise in investment and business activity over the past five years, thanks in large part to advances in materials, wireless and signal technology.

But scientists still have a variety of problems, including the occurrence of tissue scars around the implant, the quality of measurements and the development of machine learning algorithms to interpret brain signals, said Amy Orsborn, an assistant professor at the University of Washington which investigates neural interfaces,

“I do not think we know what the magic bullet is, we only know the problem,” Orsborn told Reuters.

(Reporting by Tina Bellon in New York; Edited by Joe White and Dan Grebler)