Fork the consciousness, or download the project and create your own.

Maitreyee Wairagkar

UC Davis Neuroprosthetics Lab

Builder The claim is about what the hardware does, and the philosophy is left open

Maitreyee Wairagkar led the step from decoded text to decoded sound. Her team mapped activity from 256 electrodes in the ventral precentral gyrus straight to voice features, decoded causally within 10 ms, and played the result back to the participant in closed loop.

The decoder recovered paralinguistic features as well as speech sounds. The participant could raise his pitch to ask a question, stress a word, and sing short melodies, and listeners understood almost 60% of his words with the device against about 4% without it.

Her work shows that the tone of a voice, and not only its words, leaves a motor pattern that a machine can learn. That makes the readout of intention richer, and it remains a readout of output, not of experience.

The speech neuroprosthesis article puts her 2025 voice study beside the other two. A short film shows the work on a glass brain. Watch Can a machine hear the words your brain tries to say? on YouTube.

Known for. First author of the 2025 Nature study on an instantaneous voice-synthesis neuroprosthesis

Coverage on this site

Related researchers