AI, for better or worse, has vastly improved its ability to decipher our thoughts.
Scientists at the University of Texas announced the following findings: Nature(opens in new tab) A non-invasive mind that can detect brain activity and capture the essence of what someone is thinking using functional magnetic resonance (fMRI) and an AI system called GPT-1, which predates ChatGPT I explained how I created the decoder.
To train the AI, the researchers put three people through an fMRI scan and played a fun podcast for them to listen to: The New York Times “Modern Love”(opens in new tab)and moss radio hour(opens in new tab). Scientists used podcast transcripts to track brain activity and figure out which parts of the brain are activated by different words.
How AI Tools Like ChatGPT Can Fight ADHD Paralysis
To see if AI could decipher the images, the scientists played silent clips of Pixar movies with subtitles to see if subjects could translate the associated stories they had in mind without speaking. tested. The results were surprisingly undetailed, but accurate enough for a decoder to understand the meaning behind the subject’s thoughts and translate them into text.
On the one hand, this is really exciting news. Imagine a future where people with neurological conditions and stroke survivors can communicate again with the help of this kind of technology.
However, the decoder is not yet fully developed. AI works only if it is trained using data of the brain activity of the person it is used in, which limits its potential for distribution. fMRI scans also have large and expensive barriers. Additionally, scientists have found that decoders can get confused if people think different things than they need and decide to “lie” to them.
These obstacles can be positive, as the possibility of creating machines that can decipher people’s thoughts raises serious privacy concerns. There is currently no way to limit the use of this technology to medicine, just imagine if decoders could be used as a means of surveillance or interrogation. Therefore, before further development of AI mind-reading techniques, scientists and policy makers should seriously consider the ethical implications and seek to ensure that this kind of technology is only used to benefit humanity. Laws to protect privacy need to be enforced.
