A man with paralysis being connected to the brain-computer interface system
Lisa E Howard/Maitreyee Wairagkar et al. 2025
A Breakthrough in Communication: Man Regains Speech Through Brain-Computer Interface
A remarkable breakthrough in communication technology has enabled a man who lost the ability to speak to engage in real-time conversations and even sing using a brain-controlled synthetic voice.
The innovative brain-computer interface system reads the man’s neural activity through implanted electrodes in his brain, allowing for the instantaneous generation of speech sounds that accurately reflect his intended pitch, intonation, and emphasis.
According to Sergey Stavisky from the University of California, Davis, this technology represents a significant advancement in instantaneous voice synthesis, achieving this feat within an impressive 25-millisecond timeframe.
While the system still requires improvements to enhance speech clarity, Maitreyee Wairagkar, also from UC Davis, notes that the man, who lost his ability to speak due to amyotrophic lateral sclerosis, finds the technology uplifting, describing it as resembling his real voice.
Unlike existing speech neuroprostheses, which have a delay in converting brain activity into sounds, this new system allows for natural conversations by providing real-time feedback, enabling interruptions, clarifications, and responses on the fly, as mentioned by Stavisky.
To achieve more realistic speech synthesis, Wairagkar, Stavisky, and their team implanted 256 electrodes in the regions of the man’s brain responsible for controlling facial muscles used in speech. Through multiple sessions, the researchers exposed him to various sentences on a screen, prompting him to vocalize them while recording his brain activity.
Stavisky explains the significance of this approach, stating, “The idea is that, for example, you could say, ‘How are you doing today?’ or ‘How are you doing today?’, and that changes the semantics of the sentence. This leads to a much richer and more natural exchange, marking a substantial improvement over previous systems.”
By feeding this data into an artificial intelligence model trained to correlate specific neural patterns with the man’s intended words and intonations, the system can generate speech based on his brain signals, creating a voice that mirrors both his intended message and desired delivery.
The team even utilized voice-cloning technology to train the AI on voice recordings from before the man’s condition worsened, ensuring that the synthetic voice closely resembles his original one.
In a fascinating aspect of the experiment, the man was able to sing simple melodies with different pitches, with the AI decoding his intended pitch in real time and adjusting the synthesized singing voice accordingly.
Furthermore, he could spontaneously produce sounds like “hmm,” “eww,” or even invent new words using the system, showcasing the versatility and potential of this groundbreaking technology.
David Brandman, another team member from UC Davis, praises the man’s resilience and adaptability, noting his transition from paralysis and speechlessness to maintaining a full-time job and engaging in meaningful conversations through this transformative technology.
