TLDR: A recent MIT Media Lab study indicates that using AI tools like ChatGPT for tasks such as essay writing significantly reduces brain engagement, memory recall, and a sense of ownership over work, potentially leading to ‘cognitive debt’ and decreased learning skills. While preliminary, the findings raise concerns about the long-term impact of AI reliance, particularly on developing brains.
A groundbreaking study from the Massachusetts Institute of Technology (MIT) Media Lab has shed new light on the neurological effects of using generative artificial intelligence, specifically ChatGPT, on the human brain. The research, titled ‘Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task,’ suggests that relying on AI for cognitive tasks can lead to reduced brain activity and hinder learning processes.
The study involved 54 adult participants, aged 18 to 39, from the Boston area. They were divided into three groups: one using ChatGPT, another using Google’s search engine, and a ‘brain-only’ group that used no external tools. Over a period of four months, participants were tasked with writing several SAT-style essays. Researchers meticulously measured brain activity using electroencephalography (EEG) across 32 regions of the brain and conducted linguistic analyses of the essays, followed by interviews to assess recall and ownership.
The findings revealed a stark contrast in brain engagement. The group that utilized ChatGPT consistently exhibited the lowest levels of neural connectivity and overall brain activity. According to the study, these participants ‘consistently underperformed at neural, linguistic, and behavioral levels.’ As the study progressed, the ChatGPT users demonstrated a tendency to become increasingly ‘lazier,’ often resorting to direct copy-pasting of AI-generated text, sometimes from a single prompt.
Furthermore, the AI-reliant group struggled significantly with memory recall, finding it difficult to quote from their own essays and expressing a diminished sense of ownership over their written work. EEG results indicated weaker alpha and theta brain waves in this group, suggesting a bypass of deep memory processes. When later asked to rewrite an essay without the aid of ChatGPT, these participants remembered very little of their original content and showed continued low brain activity.
Conversely, the ‘brain-only’ group displayed higher neural connectivity and robust brain activity. They demonstrated superior recall and were more adept at quoting from their essays. Interestingly, when members of this group were later allowed to use ChatGPT, their brains maintained high levels of activity, and they continued to produce quality essays. This suggests that for individuals who have already developed strong critical thinking and writing skills, AI might serve as a supplementary tool rather than a replacement for cognitive effort.
Participants who used Google Search also reported high satisfaction and showed active brain function, though their neural connectivity was not as high as the ‘brain-only’ group.
Nataliya Kosmyna, the lead author of the study and a research scientist at MIT, emphasized the urgency of these preliminary findings, despite the preprint not yet undergoing full peer review. She expressed concern that policymakers might prematurely integrate AI into early education, stating, ‘I am afraid in 6-8 months, there will be some policymaker who decides, ‘let’s do GPT kindergarten.’ I think that would be absolutely bad and detrimental.’ Kosmyna highlighted that ‘developing brains are at the highest risk’ from over-reliance on AI.
Psychiatrist Zishan Khan echoed these concerns, noting that excessive dependence on large language models (LLMs) can lead to ‘unintended psychological and cognitive consequences,’ particularly for young people whose brains are still developing. He warned that ‘these neural connections that help you in accessing information, the memory of facts, and the ability to be resilient: all that is going to weaken.’
Also Read:
- The Psychological Toll of AI: Kindness and Connection Masking Risks of Dependency and Delusion
- OpenAI CEO Sam Altman Urges Caution: ChatGPT’s Hallucinations Demand Rethink of AI Dependence
While the study’s sample size is relatively small, its implications are significant, pointing towards a potential ‘cognitive debt’ and a likely decrease in learning skills associated with the habitual use of AI for complex tasks. The research underscores the critical need for further investigation into how AI tools are reshaping human cognition and learning in the long term.
