Brain Decoder

Credit: Nolan Zunk

Artificial intelligence made many remarkable advances this year, and a type of mind reading was among them. University of Texas at Austin neuroscience and computer science researchers Alex Huth and Jerry Tang developed a system called a semantic decoder that can translate a person’s brain activity — while listening to a story, silently imagining telling a story or watching a video — into a continuous stream of text. The system relies on extensive training in an fMRI machine like this one with willing participants and a transformer model similar to ChatGPT. It could one day help give a voice to people who are unable to speak.