It already seems like big tech companies can read our minds—it's a bit eerie to see an ad for something that we were thinking about. You can chalk those instances up to a combination of fancy predicative algorithms
and coincidence. Nevertheless, Facebook
is actually developing a brain-computer interface (BCI) that can actually transcribe your thoughts into written text.
Sounds like the stuff of science fiction, right? Not that long ago, it would have been. But it is an actual project at Facebook Reality Labs, and it has been a point of focus
since at least 2017. Facebook is also not alone in this kind of research—a start-up called CRTL-labs has been dabbling in BCI
technologies for at least as long, and more recently, Elon Musk unveiled a BCI implant
that is being tested on animals.
Facebook is testing this sort of thing as well, in humans. It has been collaborating with researchers from the University of California, San Francisco (UCSF) on special electrodes that have been implanted into a trio of patients with epilepsy. What Facebook is ultimately trying to develop is a non-invasive, wearable device for people who can't speak. It could also end up "a very powerful way for people to interact with their digital devices."
"The past decade has seen tremendous strides in neuroscience—we know a lot more about how the brain understands and produces speech. At the same time, new AI research has improved our ability to translate speech to text. Taken together, these technologies could one day help people communicate by imagining what they want to say—a possibility that could dramatically improve the lives of people living with paralysis," Facebook says.
As for the human testing that is currently underway, it involves volunteer participants with normal speech who were already undergoing brain surgery to treat epilepsy. The tests are the final phase of what Facebook is calling Project Steno, which involves a year-long study to see if it is possible to use brain activity to restore a participant's ability to communicate in the face of disability.
Researchers involved in the project hope to achieve a real-time decoding speed of 100 words per minute with a 1,000 word vocabulary and word error rate of less than 17 percent.
"By demonstrating a proof of concept using implanted electrodes as part of their effort to help patients with speech loss, we hope UCSF’s work will inform our development of the decoding algorithms and technical specifications needed for a fully non-invasive, wearable device," Facebook says.
That said, Facebook admits there is a sizable gap between the results that might be possible with implanted electrodes, versus non-invasive BCI methods. In time, though, this could all become a reality.