The AI robot that can translate thoughts into written words. It is non-invasive / video
Translating one's brain activity into written words may sound like a science fiction dream, but a new artificial intelligence (AI) model developed at the University of Texas at Austin has managed to do just that. Using only non-invasive scanning methods, the model can be trained to decode complex language from someone's thoughts over long periods of time.
"For a non-invasive method, this is a real leap forward compared to what has been done before," said Alex Huth, assistant professor of neuroscience and computer science, according to IFL Science . Similar systems are being developed elsewhere, but what sets this one apart is that participants don't have to undergo surgery to have the implants installed, nor are they restricted to a list of words they can use.
Using technology like ChatGPT and Google's Bard chatbots, the model – called a semantic decoder – is trained on data obtained from a person while they listen to podcasts, and the brain is scanned by functional magnetic resonance imaging ( fMRI). Later, with the participant's consent, it can decode the thoughts while the person listens to something else. The model will generate a text stream.
The decoder cannot synthesize the person's thoughts word for word, but can often capture the essence of what they are thinking. After extensive training, he is able to produce text that is a good and sometimes accurate representation of the person's thoughts. Read More…