Scientist Translates Thoughts Into Words Using MRI technology
Groundbreaking research utilizes brain scans and artificial intelligence to convert mental imagery into descriptive language, opening doors for communication assistance and raising critical privacy concerns.
A new technique developed by a scientist in japan is transforming the landscape of brain-computer interfaces, allowing for the translation of thoght into text, to interpreting the complex imagery within our minds.
While previous advancements have focused on decoding words we think, converting complex mental images into language has remained a formidable challenge. According to the study’s author, Tomoyasu Horikawa, this new method leverages the power of artificial intelligence to generate text that reflects the intricate details of brain activity related to visual facts – objects, places, actions, events, and their relationships.
Horikawa, a researcher at NTT’s Communication Science Laboratories outside Tokyo, conducted the study by analyzing the brain activity of six participants – four men and two women, all native Japanese speakers between the ages of 22 and 37. Participants underwent brain scans while watching 2,180 silent music videos, each lasting only seconds and featuring diverse content.
The process involved utilizing large language models – generative AI systems trained on extensive datasets – to convert subtitles from the music videos into numerical sequences. Horikawa then trained separate AI “decoders” to correlate scanned brain activity with these numerical sequences.These decoders were subsequently used to interpret brain activity as participants watched or recalled videos not previously encountered during training. A final algorithm then generated word sequences that best aligned with the decoded brain activity.
As the AI learned,the descriptive text tool became increasingly adept at translating brain scans into descriptions of the viewed videos. “It’s just another step forward towards what, in my view, we can legitimately call brain reading or mind reading,” stated a leading expert in AI ethics and neuroscience at the Technical University of Munich in Germany, who was not involved in the study.
Remarkably, the AI model generated text in English, even when participants were not native English speakers, demonstrating the technology’s potential for cross-linguistic application. The method can even function without relying on activity in language-related brain regions, suggesting it’s utility for individuals with damage to those areas.
The potential applications of this technology are far-reaching.Researchers suggest it could assist individuals with aphasia, a communication disorder resulting from language network damage, and those suffering from amyotrophic lateral sclerosis (ALS), a progressive neurodegenerative disease affecting speech.”I believe this study paves the way for some profound interventions for people with communication difficulties, including nonverbal autistic people,” noted a psychologist at Barnard college in New York, who was not involved in the research.
However, experts caution that careful consideration must be given to the ethical implications. “We need to use it carefully and make sure we’re not being invasive and that everyone is consenting to it,” one researcher emphasized. The ability to decode thoughts – even extending to the potential of interpreting the thoughts of babies,animals,or the content of dreams – raises significant privacy concerns,with the possibility of revealing private thoughts before they are consciously expressed.
The study highlights the potential for revealing an individual’s private thoughts before they have verbalized them, a concern echoed by experts in the field. If this technology expands beyond biomedical applications and becomes available to consumers,”I believe this is the ultimate privacy challenge,” warned a professor of AI ethics. He pointed to companies like Neuralink, which are actively developing brain implant technology for the general population, emphasizing the need for “very, very strict rules” regarding access to individuals’ minds and brains, given the “sensitive information” they contain, including indicators of mental health conditions.
Recent research published in the journal Cell in August suggests a potential safeguard: a mechanism requiring users to think of a specific keyword to unlock the decoding tool, preventing unintentional thought leakage.
“Neuroscience is advancing rapidly and the assistive potential is enormous – but mental privacy and freedom of thought protections cannot wait,” stated a social scientist and director of European affairs at the Neurorights Foundation in New York.This groundbreaking research underscores the urgent need for robust ethical frameworks to govern the development and deployment of these powerful new technologies.
