Meta has developed a new AI model called Brain2Qwerty. This model can read typed sentences directly from brainwaves. No operation is needed for this. This technique is a major advance for connecting humans and computers.
This development is a real breakthrough. It changes the boundary between thought and digital input. It is a glimpse into a future where direct communication without a keyboard or speech becomes possible. This can change lives. But it also raises important questions about the protection of our privacy.
Researchers at Meta have introduced their AI model Brain2Qwerty. It deciphers typed sentences directly from brainwaves. For this, they use EEG (electroencephalography) and MEG (magnetoencephalography) data. These are measurement methods that do not require implants in the head. The model achieves an average word accuracy of 61 percent.
For individuals, this technology could significantly change communication. People who are paralyzed or have lost their voice could 'speak' again this way. Imagine being able to write an email or message just by thinking. Your phone then becomes a direct extension of your thoughts. But there is a downside: What if not only what is typed, but also other brain activities could be deciphered?
Companies face new challenges and opportunities. In healthcare, new avenues for assistive technologies and rehabilitation methods are emerging. These are aids for people with disabilities. Tech companies could work on entirely new input devices and user interfaces. At the same time, the business world must address the ethical and legal questions that arise when data is processed directly from the brain. This concerns data protection, data security, and compliance.
The greatest opportunity lies in restoring communication abilities. This applies to people with severe physical disabilities. This could greatly improve their quality of life. New, simple control options could also emerge for healthy people. For example, in VR/AR (virtual or augmented reality) or smart home systems. An AI model that converts thoughts into text is like an interpreter for the subconscious.
The most obvious risk is data privacy. If a system can translate brainwaves into text, how secure are our most private thoughts? There is a risk that such technologies could be misused. For example, for surveillance or unauthorized access to mental data. The accuracy of 61 percent is also not yet perfect. This can lead to misinterpretations and frustration.
