Do AI Chatbots Like ChatGPT Harm Our Brains?
Educational institutions worldwide are grappling with policies around AI tool usage as systems like ChatGPT become more sophisticated and accessible.
AI is being increasingly used across the education sector. ChatGPT in particular has been widely adopted across educational institutions since its public release.
Yet researchers at MIT have published findings suggesting that OpenAI’s ChatGPT may diminish critical thinking capabilities amongst users.
MIT’s study methodology
The study, conducted by MIT’s Media Lab, tracked 54 participants aged between 18 and 39 from the Boston metropolitan area over several months.
Participants were divided into three groups and asked to complete essay-writing tasks based on standardised academic test prompts.
One group used ChatGPT to assist with their writing, another relied on Google’s search engine and a third worked without technological assistance.
Electroencephalography equipment monitored participants’ neural activity across 32 brain regions during the writing exercises. The technology measures electrical activity in the brain through sensors placed on the scalp.
Results from the ChatGPT group: Cognitive differences
Results showed that ChatGPT users demonstrated the lowest levels of brain engagement across “neural, linguistic and behavioural measurements,” according to the researchers.
These participants exhibited reduced executive control and attentional engagement compared to the other groups.
The research found that participants using ChatGPT produced essays with similar content and structure, lacking original thought. Two English teachers who evaluated the submissions described them as “soulless.”
Additionally, participants who worked without technological assistance showed the highest neural connectivity, particularly in alpha, theta and delta brain wave frequencies.
These patterns are associated with creative thinking, memory processing and language comprehension.
“What really motivated me to put it out now before waiting for a full peer review is that I am afraid in 6-8 months, there will be some policymaker who decides, ‘let’s do GPT kindergarten,’” says Nataliya Kosmyna, the study’s Lead Author and Research Scientist at MIT Media Lab since 2021.
“I think that would be absolutely bad and detrimental. Developing brains are at the highest risk.”
Results from the Google search group: Higher engagement
The group using Google’s search engine demonstrated active brain function and reported satisfaction with their work – highlighting a distinction between traditional web search and AI-powered text generation tools.
Over the course of multiple essays, ChatGPT users became increasingly reliant on the system.
By their third assignment, many participants simply provided the essay prompt to ChatGPT and performed minimal editing on the generated content.
“It was more like, ‘just give me the essay, refine this sentence, edit it and I’m done,’” Nataliya says.
The MIT Media Lab also has allocated resources to studying Gen AI’s broader societal impacts – and found earlier research from the institution found correlations between extended ChatGPT usage and increased feelings of loneliness amongst users.
The current study included a role-reversal exercise where participants rewrote previous essays under different conditions.
ChatGPT users had to work without the tool, while the control group gained access to it.
Results from the brain-only group: Higher neural connectivity
The group that were left to write the SAT essays with nothing, showed the highest neural connectivity.
The participants showed this high neural connectivity especially in alpha, theta and delta bands – which are associated with creativity ideation, memory load and semantic processing.
The researchers observed that this group was more engaged and curious and claimed ownership and expressed higher satisfaction with their essays.
The memory integration concerns from the study
Participants who had previously relied on ChatGPT struggled to recall details from their own writing and showed weaker alpha and theta brain wave patterns. These frequencies are linked to deep memory formation processes.
“The task was executed, and you could say that it was efficient and convenient,” Nataliya says. “But as we show in the paper, you basically didn’t integrate any of it into your memory networks.”
Conversely, participants who initially worked without AI assistance performed well when given access to ChatGPT, showing increased brain connectivity across all measured frequency bands.
The research has not undergone peer review and involved a relatively small sample size of 54 individuals. However, the findings contribute to growing discussions about AI’s role in educational settings as adoption increases amongst students.
“I think that would be absolutely bad and detrimental,” Nataliya says regarding potential AI implementation in early childhood education.
Explore the latest edition of AI Magazine and be part of the conversation at our global conference series, Tech & AI LIVE.
Discover all our upcoming events and secure your tickets today.
Also sign up to our free weekly newsletter for the latest insights and stories straight into your inbox.
AI Magazine is a BizClik brand
