MIT researchers have developed a headset that can identify words you think of but don't actually say, by reading signals the brain sends to the face and jaw during internal speech.
The AlterEgo headset captures the neuromuscular signals that occur when people intend to speak. It then uses a neural network to reconstruct the word.
The system can read those facial signals with 92 percent accuracy, according to the MIT researchers, and may allow users to control a computer by thought.
That could mean the end of saying, "OK Google" to a Google Home speaker, which the researchers argue has drawbacks due to the dependence on voice. For example, others can hear what you're telling Google or Siri, AI assistants are always listening, and they require attention to use.
The headset is worn on one ear with a mouthpiece that extends around the jaw line to just below the lip. The earpiece contains a pair of bone-conduction headphones that transmit vibrations through the bone to the inner ear.
The idea is to make it easier for the wearer to talk with others while receiving information from the device.
A video demo shows an AlterEgo wearer controlling a smart TV, asking Google the time, and calculating shopping bills on the go.
MIT says one experiment allowed the wearer to silently report a chess opponent's moves and get computer-generated strategic tips.
MIT Media Lab graduate student and lead developer of AlterEgo, Arnav Kapur, calls the wearable an "intelligence-augmentation device".
"Our idea was: could we have a computing platform that's more internal, that melds human and machine in some ways and that feels like an internal extension of our own cognition?"
The wearable could be quicker and less disruptive to use than a smartphone for searching on the web. Users for example wouldn't need to unlock a phone, open an app, and then type or speak a keyword.
The researchers believe AlterEgo could be used for controlling IoT devices, virtual- and augmented-reality apps, ordering an Uber ride home, or a tool for setting a calendar meeting.
Thad Starner, a professor at Georgia Tech's College of Computing, believes the technology also has potential as an industrial communications tool to assist workers in noisy environment such as airports or military.
"You've got jet noise all around you, you're wearing these big ear-protection things -- wouldn't it be great to communicate with voice in an environment where you normally wouldn't be able to?" said Starner.
"You can imagine all these situations where you have a high-noise environment, like the flight deck of an aircraft carrier, or even places with a lot of machinery, like a power plant or a printing press."