Neuroscientists from UC San Francisco have created a brain implant that uses artificial intelligence to read and translate brainwaves in people who can't speak. The system can allow those who've lost vocal function to control a synthesized version of speech through thought only. The study was published in the journal: Nature.
The articulation isn't perfect but its a huge improvement over the current technology which uses facial or eye movements to translate one letter at a time producing 10 words a minute at best. Attached to this story is a bit of sound of the technology at work. The first sentence you'll hear is a research participant reading a sentence then you'll hear the same sentence formed by that person's brain waves only and synthesized.