Stavisky’s system uses 256 microelectrodes implanted in the brain’s speech-coordination region to capture neural firing patterns. These signals are processed by two AI models: one translates brain activity into phonemes, while the second uses large-language-modeling techniques to reconstruct fluent speech. This approach marks a significant shift from earlier brain-computer interfaces that prioritized cursor movement, focusing instead on the more urgent human need for communication. In clinical trials, the technology has achieved up to 97.5% word accuracy, with a 30-millisecond latency that mimics natural conversation.
Sergey Stavisky Wins AI Prize for Restoring Speech Through Brain Signals
Sergey Stavisky, an associate professor at UC Davis, has been awarded the Chen Institute and Science Prize for AI Accelerated Research for developing a neuroprosthesis that allows paralyzed individuals to communicate in their own voices, enabling patients to hold full-time jobs and even sing using direct brain-to-speech synthesis.
The prize also recognized finalists Dr. Nicholas C. Jacobson and Dr. Zach (Zhiling) Zheng. Jacobson developed Therabot, an AI-driven psychotherapy system that has demonstrated clinical efficacy in treating anxiety and depression, potentially addressing global shortages in mental health providers. Meanwhile, Zheng created an autonomous laboratory system that integrates literature mining with robotic experimentation to accelerate the discovery of new materials, such as metal-organic frameworks for water harvesting. Together, these researchers exemplify the use of AI to solve long-standing problems in medicine and materials science, moving beyond theoretical models to provide practical, life-altering tools.
Comments (0)
No comments yet. Be the first!