
MRI scans and AI technology can literally read what we’re thinking. The implications are scary
(Nanowerk News) For the first time, researchers have successfully used GPT1, the precursor to the AI chatbot ChatGPT, to translate MRI imagery into text in an effort to understand what someone is thinking.
A recent breakthrough allows researchers at the University of Texas at Austin to “read” a person’s mind as a continuous stream of text, based on what they hear, imagine, or watch.
This raises significant concerns for privacy, freedom of thought, and even freedom to dream undisturbed. Our laws are not equipped to deal with the widespread commercial use of mind reading technology – free speech laws do not cover protecting our minds.
Participants in the Texas study were asked to listen to an audiobook for 16 hours while inside an MRI scanner. At the same time, the computers “learn” how to associate their brain activity from the MRI with what they are listening to. Once trained, the decoder can generate text from a person’s mind as they listen to a new story, or imagine their own story.
According to the researchers, the process is labor-intensive and the computer only manages to capture the gist of what a person is thinking. However, the findings are still a significant breakthrough in the field of brain-machine interfaces which, until recently, relied on invasive medical implants. Previous non-invasive devices could only decipher a few words or pictures.
Here’s an example of what one subject heard (from the audiobook):
And this is what the computer “reads” from the subject’s brain activity:
Study participants must work together to train and apply the decoder, so that the privacy of their thoughts is maintained. However, the researchers warn that “future developments might allow decoders to bypass this requirement”. In other words, mind-reading technology could one day be applied to people against their will.
Future research may also speed up the training and decoding process. While it took 16 hours to train a machine to read what someone was thinking in the current version, this will be significantly reduced in a future update. And as we’ve seen with other AI applications, decoders also tend to become more accurate over time.
There’s another reason why it represents a step change. Researchers have been working for decades on brain-machine interfaces in the race to create mind-reading technology that can understand a person’s thoughts and turn them into text or images. Typically, however, this research has focused on medical implants, with a focus on helping people with disabilities speak their mind.
Neuralink, a neurotechnology company founded by Elon Musk developing medical implants which can “let you control your computer or mobile device wherever you go”. But the need to have brain surgery to implant the device inside you will likely remain a barrier to using such technology.
However, the increased accuracy of this new non-invasive technology could make it a game changer. For the first time, mind reading technology looks feasible by combining two already available technologies – albeit at a hefty price tag. Current MRI machines cost between US$150,000 and US$1 million (£120,000 and £800,000).
Legal and ethical consequences
Current data privacy laws do not consider thoughts to be a form of data. We need new laws that prevent the emergence of thought crimes, mind data breaches, and even one day, perhaps, the cultivation or manipulation of thoughts. Going from reading thoughts to instilling them may take some time, but both require some planning and supervision.
Researchers from the University of Oxford debate the legal right to mental integritywhich they describe as: “The right to significant and non-consensual interference with one’s mind.”
Others began to defend the new human right to freedom of thought. It would go beyond traditional definitions of free speech, to protect our capacity to reflect, wonder and dream.
A world without regulation can become a dystopia very quickly. Imagine a boss, teacher, or state official being able to invade your private thoughts – or even worse, being able to change and manipulate them.
We’ve seen eye scanning technology deployed in classrooms to track students’ eye movements during lessons, to find out if they are paying attention. What happens when technology reads minds next?
Similarly, what happens at work when employees are no longer allowed to think about dinner, or anything else outside of work? The level of abusive control of workers can exceed anything previously imagined.
George Orwell wrote convincingly about the dangers of “Thoughtcrime”, in which the state makes it a crime to think rebellious thoughts about authoritarian regimes. The Nineteen Eighty-Four plot, however, is based on state officials reading body language, diaries, or other external indications of what a person is thinking.
With new mind-reading technology, Orwell’s novel will be incredibly short – maybe even as short as a sentence: Winston Smith thought to himself: “Down with Big Brother” – after that, he was caught and executed.