The word-recognition system was then paired with wearable speakers that played an ultrasonic recording of the words that a person whispered or mouthed. The speakers emit ultrasound in a narrow beam, at a six-degree angle, so that only people directly in …
( read original story …)