Overview
- Neuralink released a demo showing its N1 brain implant letting ALS patient Kenneth Shock produce spoken words without moving his mouth in the VOICE clinical trial.
- The system reads activity in speech-related brain areas, converts it to text, then speaks it through a computer in Shock’s voice recorded before ALS changed how he sounded.
- Engineers trained the AI by having Shock speak aloud, then silently mouth sentences, then only imagine them, which taught the model to map brain patterns to phonemes and build words.
- The company says the prototype still has a delay of a few seconds and makes errors, and it is working toward real-time brain-to-voice output.
- Elon Musk described the update as restoring speech, while the company stresses it remains experimental with regulatory, long-term safety, and neural data privacy hurdles ahead.