It looks like actually reading the minds and thoughts of people has just come one step closer to reality since scientists have recently developed an artificial intelligence that is capable of turning your brain activity into actual readable text.
At the moment, the system currently works on neutral patterns that can be detected whenever a person is speaking aloud. However, the experts behind it are saying that a future where the AI system can be used to aid communication in patients who are unable to both speak and type, like people who are suffering from locked-in syndrome, is not far from now.
“We are not there yet but we think this could be the basis of a speech prosthesis,” Dr. Joseph Makin, co-author of the research from the University of California, San Francisco, said.
With their research published in the scientific journal Nature Neuroscience, Makin and his colleagues were able to reveal how they were able to develop the system, as well as the four participants where they had tested it on. Per Makin, all four recruits had electrode arrays that were implanted in their brain in order to monitor any epileptic seizures. The participants were then made to read aloud from 50 set sentences while the team tracked their neural activity.
The data taken was then fed into a machine-learning algorithm, which is a type of artificial intelligence system that converts a spoken word into a string of numbers, to be translated later on.
Of course, after the tests, the system still wasn’t perfect, with some sentences recorded to be completely different ones. Nevertheless, its accuracy was far higher than previous researchers, which bodes well for the team behind it.
Furthermore, other researchers also called it as exciting since the system was still able to work despite only using a limited set of data, such as the 40 minutes of training for each participant.
“By doing so they achieve levels of accuracy that haven’t been achieved so far,” Dr. Christian Herff an expert in the field from Maastricht University, who was not involved in the study, said.