Connect with us

Mobile

MIT’s AlterEgo Headset can hear your inner thoughts and answer back

Talking to yourself just got cool again.

Ever wonder if you said that potentially scandalous comment out loud? Well if this system developed by a graduate student at MIT Media Lab takes off, you’ll have to be careful of even your inner thoughts.

Subvocalization as a concept has been around since the 19th century, where it was said that internal verbalizations have a physical counterpart. It was extensively studied in the 1950’s, and there were efforts by researchers to eliminate it altogether as they thought it hindered reading speed. It’s long been a staple of SciFi, with notable authors such as William Gibson and Iain M Banks using it in their created worlds. NASA also has been developing it, coming up with a preliminary system in 2004 that could recognize a small set of words and numbers.

This system developed at MIT takes that further, partly due to the advances in the AI field over the last decade. A headset picks up neuromuscular signals from your jaw and feeds these into a neural net that is trained to correlate specific signal-word pairs. Responses are fed back to the user by a couple of bone-conduction headphones, which convey sounds without interfering with the normal functioning of the ear.

Now you have a virtually-silent feedback loop brimming with possibility. Don’t like asking “Siri” to do things? Do it silently, and receive the answer privately without disrupting anyone around you. Not just quiet areas benefit from this; it’s as useful in loud environments where traditional microphone technology has trouble. Air traffic controllers or machine-shop operators could subvocalize in areas that generally would drown out their voice. I see many military applications, with Special Ops probably being at the head of the queue.

Then there’s the world of assistive technology. Robert Ebert lost his jaw to cancer and can’t speak usually. It’s possible that this system could give him his voice back without surgery. Since it builds on an innate human feature, rehabilitation times would be low.

I look forward to the time when this technology is widespread, and I can talk in my head to other people.

Comments
Advertisement

More in Mobile