A team of scientists has made a groundbreaking discovery by employing a Generative Pre-trained Transformer (GPT) AI model similar to ChatGPT to reconstruct human thoughts with up to 82% accuracy from functional MRI (fMRI) recordings. This unprecedented level of accuracy in decoding human thoughts from non-invasive signals paves the way for a myriad of scientific opportunities and potential future applications, the researchers say.
I read the study abstract because I couldn’t tell the impact from the headline.
It says that they can recover perceived speech pretty well, but other forms of communication or broad interpretation are less reliable (silent movies were like 20% accuracy).
It also notes that subject cooperation is needed for this to work. That was for training and decoding. However, it’s the first step towards a more concerning outcome.