In yet another fascinating piece of news about artificial intelligence: Researchers at the University of Texas at Austin have developed an A.I. that can “read minds” by analyzing fMRI brain scans and translating the information into words reflecting the private thoughts of the human subject.

The A.I. was trained with measurements taken while the subjects listened to narrative podcasts, and was able to turn imagined speech into actual speech and generate descriptions of silent films. However, the language-decoding method has limitations, as fMRI scanners are expensive, and training the model is long and tedious, and must be done on individuals.