Have you ever fantasized about being able to read another person’s mind? Or with being able to automatically transcribe your own thoughts? Humans have dreamed of power for centuries delve into the intricacies of the brain to discover what ideas, feelings and memories They hide in our gray matter. A study published this Monday in the journal ‘Nature’ brings us closer to this dream, as a group of researchers has achieved create an experimental technique able to ‘read minds’ and transcribe thoughts with amazing precision.
This is a novel method developed by scientists at the University of Texas at Austin in which, on the one hand, brain imaging analysis and, on the other, artificial intelligence tools similar to those used, for example, by ChatGPT. As explained by Jerry Tang and Alex Huth, the researchers in charge of this work, the combination of these techniques has made it possible to accurately “decode” thought from a small group of volunteers.
The scientists used artificial intelligence tools similar to those of ChatGPT
It is not the first time that it has been possible to create a connection between a person’s brain and a computer. In fact, more and more studies are trying to create a ‘translator’ of thoughts of this type. Until now, the most promising work has focused on patients with mobility problems and the use of brain implants (microchips) to capture simple ideas (or sentences). The technique presented this Monday in ‘Nature’ would be the first that achieves reproduce complex thoughts without the need for surgical interventions.
Technique to ‘read the mind’
But how did these scientists manage to decode other people’s thoughts? Pay attention because the process is more complex than you might think (and no, It’s not something that works instantly.). The study began by putting the volunteers in MRI machines. From there, they put them to listen to a podcast for several hours to, on the one hand, study their brain activity in the face of this type of stimuli and, on the other, to train an artificial intelligence tool to learn to ‘read’ their ideas.
The volunteers had to spend more than 15 hours under a scanner to be ‘studied’ by the interface
Then came the moment of truth. They took these same volunteers, put them back into the scanners and asked them to think of a story. It was then that the machine, which had previously learned to interpret how each participant’s brain works, began to ‘translate’ neural signals into written texts. “We didn’t get an exact word-for-word transcription, but we did capture the essence of thoughts“, explain the scientists who have led this work.
As explained in the article published this Monday, the technique worked in at least half of the cases. In some volunteers an almost exact transcription of their thoughts was achieved. In others, however, practically nothing was captured. “Some managed to actively resist to the decoding attempts brain with tactics such as, for example, thinking about animals or imagining themselves in silence”, explain the authors of this work.
Could they read your thoughts?
this new technique for ‘decoding thoughts’ appears, for now, as a technical achievement. In fact, according to its own creators, this tool has only worked on people who had past more than 15 hours under a scanner being ‘studied’ by this artificial intelligence interface. The system was also tested with people who had not gone through this process and the result, as expected, was something unintelligible. In other words, the machine was not able to read his mind.
Related news
There is still a long way to go before these types of tools can be used in real life, but even so, scientists urge caution in future experiments of this type. “Right now, even though this technology is in such an early stage, it’s important to be very proactive and begin to ask ourselves about the possible uses of this technology“, says Tang. “We must ensure that people only use these types of technologies when they want and for their own benefit,” adds the expert.
“We must ensure that people only use these types of technologies when they want and for their own benefit”
This “warning to responsibility“He also supports her. David Rodriguez-Arias Vailhen, Deputy Director of FiloLab and Professor of Bioethics at the University of Granada. “That a machine can end up reading your mind once trained opens the door to what, involuntarily and without your consent (for example, while you sleep), it can go translating pieces of your thought”, reflects this expert in statements to the Science Media Center Spain. “This finding could be the first step so that in the future that freedom would be compromised. It would be better for us then to dream only of beautiful melodies,” adds the scientist.