AI TRENDS | Meta Introduces Brain2Qwerty v2 With 61% Average Word Accuracy for Brain-to-Text

Meta has introduced Brain2Qwerty v2, a non-invasive AI system designed to decode brain activity into text, reporting an average word accuracy of 61%. According to NS3.AI, Meta said earlier non-invasive approaches achieved about 8% accuracy. Meta described Brain2Qwerty v2 as a brain-to-text decoding system that converts measured brain activity into written words, positioning the update as a significant accuracy improvement over prior non-invasive methods.