Meta's new AI system can turn your thoughts into text
What's the story
Meta has unveiled a groundbreaking artificial intelligence (AI) system, Brain2Qwerty v2, that can convert brain activity into text without any surgical implants. The technology is said to be approaching accuracy levels previously only achievable through invasive methods like brain surgery. The development marks a major milestone in the field of brain-computer interfaces, and could potentially help millions of people with communication disorders caused by brain lesions or other conditions.
Innovative technique
How Brain2Qwerty v2 works
Unlike invasive methods such as stereotactic electroencephalography and electrocorticography, which require surgical implants to capture brain signals, Brain2Qwerty employs non-invasive recordings for text decoding. The system was trained on about 22,000 sentences from nine volunteers who wore a magnetoencephalography (MEG) device while typing. Brain2Qwerty v2 uses end-to-end deep learning to decode language directly from raw brain signals instead of relying on manually designed pipelines for identifying neural events.
Performance metrics
Accuracy rates of Brain2Qwerty v2
Meta's new system has achieved a word accuracy rate of 61%, a huge leap from the 8% word accuracy of other non-invasive methods. For the best-performing participant in the study, this number jumped to an impressive 78%, with over half of all decoded sentences having one word error or less. The company also found that decoding accuracy improved with more training data, hinting that larger datasets could further narrow the performance gap between non-invasive and surgical methods.
Research transparency
Meta is making Brain2Qwerty v1 and v2 training code public
To further advance this field of study, Meta is making the complete training code for Brain2Qwerty v1 and v2 publicly available. Its research partner, the Basque Center on Cognition, Brain and Language (BCBL), will also make the Brain2Qwerty v1 dataset public. These efforts are part of Meta's larger strategy to create open foundational models of the brain and push neuroscience research forward for better diagnosis, treatment, and understanding of neurological disorders.