Stanford University develops brain chip that turns thoughts into written text

A team of researchers at Stanford University has announced the development of an experimental brain-implanted chip that can directly translate human thoughts into written text, without the need for speech or lip movement. The aim of this technology is to help people with speech impairments, such as the deaf-mute, communicate more smoothly.

Initial trials revealed that the chip’s accuracy reaches around 74%, marking a significant scientific breakthrough in this field.

The innovation relies on brain-computer interfaces (BCIs), technologies that connect the human nervous system to devices capable of analyzing and interpreting neural signals. This allows users to carry out commands such as operating a computer or controlling prosthetic limbs simply by thinking.

Among the most similar technologies is Neuralink, developed by Elon Musk, which is currently undergoing clinical trials to assess its effectiveness and safety, especially for people with severe motor disabilities.

In a recent study published in Cell, researchers noted that this technology could represent a promising future for improving communication methods for individuals with speech impairments.

Stanford researcher Erin Koontz stated: “For the first time, we are able to understand brain signals linked solely to thinking about speech, without the need to actually say it.”

During the experiments, microscopic electrodes were implanted in the motor cortex responsible for speech in four participants. They were asked either to attempt speaking certain words or to imagine saying them. Results showed that both actions activated similar brain regions and generated comparable neural patterns.

The researchers relied on advanced artificial intelligence models to interpret the signals produced by imagined words, and the chip successfully converted those thoughts into text with 74% accuracy.

Despite this progress, scientists emphasize that the current ability to “read thoughts” remains limited. However, with further model development, it may one day be possible to overcome these constraints and even distinguish between intentional thoughts and unwanted inner speech.

Professor Frank Willett stated: “This development brings hope that brain-computer interface technologies will enable people to regain their ability to communicate naturally and seamlessly, as if they were actually speaking.”