In a study in Nature, researchers reported being able to identify words and phrases in volunteers undergoing fMRI imaging reasonable accuracy. The process is non-invasive, unlike implanted electrodes, but requires hours of preparation and scanning.
This technology would be a significant breakthrough for people suffering debilitating conditions that prevent them from speaking or otherwise communicating. Previously, decoding language required the use of extensive electrode implants.
The participants, two male and one female, listened to recordings of radio shows. This was used to train a language model which was based on an early version of ChatGPT. By looking at the brain’s responses, the language model was able to capture the gist of what the participants were thinking, sometimes replicating exact words or entire phrases.
Marked safe from ‘Big Brother’… for now
At this stage, the technology used requires the subject to cooperate, the researchers wrote, allaying concerns over any malicious use of this technology to tap into people’s private thoughts. Testing the decoding model on people who it hadn’t been trained on produced unintelligible results, as was the case when the trained participants put up resistance.
While the technology cannot be used for nefarious mind-reading, the march of progress means that one day such concerns will become real.
Nita Farahany, JD, PhD, of Duke University in Durham, North Carolina, told MedPage Today that the technology could one day be used against people. “This research illustrates the rapid advances being made toward an age of much greater brain transparency, where even continuous language and semantic meaning can be decoded from the brain.
“While people can employ effective countermeasures to prevent decoding their brains using fMRI, as brain wearables become widespread that may not be an effective way to protect us from interception, manipulation, or even punishment for our thoughts.”
While lugging around a massive MRI machine would be a challenge for future thought police, smaller, more portable means of measuring brain activity remotely. Senior author Alexander Huth, PhD, of the University of Texas at Austin, says that one such technology could be functional near-infrared spectroscopy (fNIRS).
“fNIRS measures where there’s more or less blood flow in the brain at different points in time, which, it turns out, is exactly the same kind of signal that fMRI is measuring,” Huth said. “So, our exact kind of approach should translate to fNIRS,” but the resolution with fNIRS would be lower.
