“According to a news release, the trained system produces text that closely or precisely matches the intended meaning of the participant’s original words around half of the time.”
The University of Texas decoder needs several hours of each person’s podcast listening inside an fMRI scanner before it works. Then it gets the gist right about half the time. “I don’t have my driver’s license yet” came out as “She has not even started to learn to drive yet.” It is a coin flip that costs a hospital’s worth of magnet, and the press release calls it a leap forward.