Scientists in Texas develop mind-reading AI system (Image credit- NFT News Pro)
A group of academics at the University of Texas at Austin has recently added another exciting advancement to the list of AI’s exciting developments!
An artificial intelligence system that can read people’s thoughts and translate them into a continuous stream of text has reportedly been built, according to a recent study.
Using the Semantic Decoder
The AI system, known as the semantic decoder, can convert a person’s brain activity into text without the use of implants or word lists. Instead, after intensively training the decoder, the system makes use of an fMRI scanner to record brain activity.
The participants listen to many hours of podcasts while being scanned, allowing the machine to generate text solely based on their brain activity.
The study’s principal investigators were UT Austin assistant professor of neuroscience and computer science Alex Huth and computer science Ph.D. student Jerry Tang.
A transformer model, like the ones used by ChatGPT by Open AI and Bard by Google, was used by the researchers to build the AI system.
Also read: Scientists develop A.I. system that can transform peoples’ thoughts into text
The semantic decoder’s ability to decode continuous discourse for extended periods of time with complex ideas distinguishes it from other language decoding systems.
Huth observed, “For a noninvasive technology, this is a huge leap forward compared to what has been done before, which is usually single words or brief sentences.
