Meta's AI System Can Do Wonders! It Will Read Thoughts Without You Saying a Word

Meta Brain2Qwerty v2: Meta has developed a new AI system capable of converting thoughts into text. It requires neither surgery nor implants.

Meta Brain2Qwerty v2: Meta has created an AI system that can read the thoughts running through your mind. Named "Brain2Qwerty v2," this system converts brain activity into text. In other words, it can display your thoughts on a screen in text format without you having to speak. A key feature is that it requires neither surgery nor implantation to use. This AI system will be incredibly useful for people who are paralyzed or unable to speak due to other conditions. It is the second version of Meta's Brain2Qwerty AI system, which was introduced last year.

How will this system work?

Meta explained that the system utilizes magnetoencephalography (MEG). Such systems do not require surgery; instead, they use a helmet equipped with sensors to measure magnetic fields generated by brain activity (neural activity). To put it simply, the system operates using a special helmet that resembles a hair dryer. This helmet contains multiple sensors that capture signals from brain activity. Its standout feature is that it eliminates the need for surgery or implants, making it both safe and cost-effective to use.

Many more features in the new version

According to Meta, Brain2Qwerty v2 was trained using over 22,000 sentences provided by nine volunteers. Each volunteer spent 10 hours wearing the system and typing out their thoughts. Compared to the older version, the new version utilizes end-to-end deep learning to decode language directly from brain signals. The system can autonomously identify patterns in brain signals with the help of AI.

How accurate is the new system?

In terms of accuracy, Meta claims a per-word accuracy rate of 61 percent, which reached up to 78 percent for the best-performing participant. Other similar systems typically achieve an accuracy of only 8 percent. To enhance the system's performance and efficiency, Meta has also fine-tuned a large language model, enabling it to account for grammatical context during conversation.