Researchers at the University of Texas at Austin have developed an artificial intelligence (AI) model like ChatGPT which is capable of reading, interpreting and reconstructing human thoughts.
According to a study published in the Nature Neuroscience journal, AI models powering OpenAI’s ChatGPT and Google’s Bard AI can read individuals’ minds and reproduce brain activity in words.
Jerry Tang, a doctoral student in computer science, and Alex Huth, a professor at UT Austin, have led the study.
The findings demonstrated the viability of non-invasive language brain–computer interfaces. “These kinds of systems could be especially helpful to people who are unable to physically speak, such as those who have had a stroke, and enable them to communicate more effectively,” the researchers noted.
According to the researchers, current methods of decoding thoughts into words are invasive or limited as they can only identify stimuli from among a small set of words or phrases.
They went on to introduce a non-invasive decoder that reconstructs continuous language from cortical semantic representations recorded using functional magnetic resonance imaging (fMRI).
The University observed that AI technology was able to get things right approximately 50 per cent of the time.
“Our privacy analysis suggests that subject cooperation is currently required both to train and use the decoder. However, future developments might enable decoders to bypass these requirements. Moreover, even if decoder predictions are inaccurate without subject cooperation, they could be intentionally misinterpreted for malicious purposes,” the researchers said.