This implant allowed a man to type by imagining himself writing

Mind control is very real — and faster than ever.

A new brain-computer interface could allow people with physical impairments to type on a computer us...

Neural implants have long been a thing of science fiction, but maybe not for much longer. Putting aside Elon Musk teaching monkeys to play video games, getting potential use cases to actually meet the current possibilities implants afford has been difficult.

YouTube / CNET

One use academics have investigated is helping people with impaired motor function regain some basic skills, like typing, by translating their thoughts into computer commands.


A recently published report in the journal Nature documents a novel “brain-to-text” communication system that has enabled a paralyzed man to think of typing and make it actually happen on a computer screen.

Westend61/Westend61/Getty Images

In tests, the man, paralyzed from the neck down, was able to type 90 characters per minute just by thinking about them. That isn’t going to win any world typing records, but it does win the record for implant-based typing.


A previous system created by the same researchers topped out at about 25 characters per minute. In that system, the user would look at a virtual keyboard and implants would monitor muscles in the eyes to guess where they were going. But it was slow and required a lot of focus on the part of the user.

YouTube / NPG Press

The new system targets an earlier step in the writing process: the intention phase. That’s found in the premotor cortex, where a person decides what letter they want to write.

Two implants monitor electrode activity for patterns associated with the thought of writing different letters. They asked the man to imagine himself writing out the alphabet and recorded the neural activity to identify patterns associated with each character.

YouTube / NPG Press

The researchers caution it’s still early days for this type of system — it was only tested on one individual, and their neural network was trained on only a basic alphabet. But the speed improvements over previous systems show a lot of promise for the future.


And who knows, this work could someday use in mass-market applications, like allowing people to communicate with their AR glasses seamlessly, without any clunky controllers or menus. As you’re willing to open up your skull for it, of course.

picture alliance/picture alliance/Getty Images

Thanks for reading,
head home for more!