How Mind-Reading AI Opens Up New Boundaries of Ethics

The idea that someone or something can read what you are thinking seemed unthinkable not long ago. However, new studies show convincing evidence that artificial intelligence (AI) may be able to do this with accuracy.

A study published in the journal Nature Neuroscience shows that AI can read Functional Magnetic Resonance Imaging (fMRI) scans of brainwaves and decipher the brain signals.

The study on whether AI models behind ChatGPT and Google’s Bard can read human thoughts and produce them in the form of text was led by Jerry Tang, a doctoral student in computer science, and Alex Huth, who is an assistant professor of neuroscience and computer science at The University of Texas at Austin.

The objective of the study was to find out if AI can read the thoughts of people, before repeating them in the form of text.

Three people listened to a story, with their brain scanned by non-invasive MRI machines.

The AI model could partially understand the thoughts based on the scan.

As seen in the above image, AI could translate “the gist” of brain waves — from

“I didn’t know whether to scream, cry or run away… instead I said ‘leave me alone'”,

to

“…started to scream and cry and then she just said ‘I told you to leave me alone”

The work relies in part on a transformer model, similar to the ones that power Open AI’s ChatGPT and Google Bard.

Unlike other language decoding systems in development, this system does not require subjects to have surgical implants, making the process noninvasive. Participants also do not need to use only words from a prescribed list.

Brain activity is measured using an fMRI scanner after extensive training of the decoder, in which the individual listens to hours of podcasts in the scanner.

Later, provided that the participant is open to having their thoughts decoded, their listening to a new story or imagining telling a story allows the machine to generate corresponding text from brain activity alone.

Assistant Professor Huth said: “For a noninvasive method, this is a real leap forward compared to what’s been done before, which is typically single words or short sentences

“We’re getting the model to decode continuous language for extended periods of time with complicated ideas.

“These kinds of systems could be especially helpful to people who are unable to physically speak, such as those who have had a stroke, and enable them to communicate more effectively.”

AI Mind-Reading: An Ethical Dilemma?

While AI potentially reading human thoughts is an exciting development, many questions are being asked from an ethical perspective. How will AI go about reading the minds of human beings? What will be done with the findings of the human mind? Who has access to the technology? How will you ensure that the technology is not used for privacy-breaking purposes?

“We take very seriously the concerns that it could be used for bad purposes and have worked to avoid that,” said one of the study’s leads, doctoral student Jerry Tang, UT Austin:
“We want to make sure people only use these types of technologies when they want to and that it helps them.”

Yu Takagi, a neuroscientist and an assistant professor at Osaka University acknowledged “For us, privacy issues are the most important thing…If a government or institution can read people’s minds, it’s a very sensitive issue. There needs to be high-level discussions to make sure this can’t happen.”

How Do We Regulate For AI Advances?

The main principle governing the development of this technology should be balanced and humane — but that is easier said than done.

Perhaps the world needs a powerful regulatory body to carefully review AI development — but getting the world to agree with that, and to then regulate in an extremely fast-moving industry seems nigh impossible.

Concerns that come to mind first are governments that have a less than impeccable approach to human rights, or this form of technology passing into the hands of malicious forces such as terrorists.

The Bottom Line

The question of ethics is pertinent and uncomfortable because it’s difficult, if not impossible to have all countries agree to a common minimum set of rules governing the development and implementation of AI.

The mind-reading capability seems to be one of the most potent capabilities that needs to be used extremely carefully, otherwise, it could spell disaster.? That said, mind-reading AI is definitely a fascinating development with multiple, legitimate use cases.