Semantic decoder can translate brainwaves into Text
Scientists have developed a helmet with sensors that can read the electrical brainwaves with the help of advanced AI, convert thoughts into written words. This method uses a cap to capture brain activity and an AI tool, DeWave, to turn these signals into text. Although it’s not perfect yet, achieving around 40% accuracy, newer studies show it might reach over 60% accuracy.
This technology is easy to use and affordable. In recent experiments, people read sentences in their heads, and the AI translated these thoughts into text. Previously, similar research used MRI scans, but this new method is more practical because it doesn’t require lying still in a scanner.
DeWave was trained to recognize specific brain signals associated with words or sentences. It was then combined with a powerful language model similar to ChatGPT‘s AI. This model acts like a sophisticated writer, using brain signals to form sentences.
Researchers believe this mindreading technology could be a breakthrough for people who can’t speak due to conditions like stroke. It might also have uses in robotics. Experts are excited about this progress, noting it’s a big step from previous attempts, which often produced meaningless text.
