It is now possible to transcribe a person’s thoughts with the help of mind-reading technology. This technology utilizes an artificial intelligence (AI) system that reads the brain’s activity via an fMRI scan and translates it into texts in real-time. A study, which was conducted by nature neuroscience and published in the publication of the same name, tested this concept with three participants and succeeded in translating their thoughts from brain images into full texts.
This breakthrough technology does not require implants and it can convert imagined speech into actual speech. Tests were conducted involving three participants where they listened to podcasts while under an fMRI scan to check their blood oxygenation levels. The AI was then employed to match patterns of the brain activity to the familiar words heard by the participants. The results show that it is possible to interpret meaning from brain activity on top of recognizing words.
AI was made possible due to the OpenAI GPT-4 and Google’s Bard large language models. These models generate maps that suggest connections between different words. The team at the university was able to use these maps to detect when brain activity associated with a certain thought is produced. To ensure validity, the participants were presented with new material to measure how accurately the AI could interpret their thoughts.
Though this technology is exciting and definitely has great potential, it also raises questions about mental privacy. The concern is that this technology can now potentially be used to track thoughts of people without their knowledge and be misused by a powerful—government or employer. However, the research team assured that only trained systems could interpret a person’s thoughts. They posit that if one curtails thinking about a subject, the AI decoding process will be disrupted.
The study led by Dr. Alexander Huth and also includes scientist Greta Tuckute from the Massachusetts Institute of Technology (MIT). Dr. Huth commended the technology for being able to identify meaning from brain activities. On the other hand, Tuckute noted how the research showed how the brain houses similar representations when forming a thought or responding to external stimuli.
If used ethically, this technology can benefit people with disabilities or even those looking to unlock hidden potentials. Neuralink, a company led by Elon Musk, has already started searching for new ways to help people with paralysis express themselves more easily with a brain-computer interface.
The technology discussed in this article holds great potential for the medical and tech world. Scientists and ethical practices should be employed to ensure that it is utilized in a way that protects the interests of people and their thoughts.
