Meta Unveils New Tech That Uses AI to Translate Brain Activity Into Text—Without Surgery
Meta introduced Brain2Qwerty v2, a non-invasive brain-to-text system that converts brain activity into typed sentences using MEG brain scans and an end-to-end AI model. The goal is to help people who have lost speech or typing ability after brain injury or lesions. The system was trained on about 22,000 sentences from nine volunteers who each spent 10 hours typing while wearing a helmet-like MEG scanner. Meta says it improves decoding by fine-tuning language models on neural data so the system can use semantic context to interpret noisy signals. The company reported 61% average word accuracy, far above prior non-invasive methods at about 8%, and said performance improved with more training data. Meta is releasing the code and dataset through its Digital Brain Project and a $5 million open neuroscience fund.
