Asking chatGPT to make you smarter is actually making you dull, MIT study finds

New Delhi: Generative AI is rapidly becoming one of the biggest disruptions of the modern era. From large-scale systems to everyday tasks, it has found its place in nearly every part of human life. It can now perform tasks that were once dependent on basic human cognitive effort.

However, much like how the body weakens without exercise, the brain too begins to lose its sharpness when it is not challenged. Simply put, over-reliance on AI might be making us mentally lazy.

A recent study conducted by the Massachusetts Institute of Technology (MIT) has highlighted this concern. Titled ‘Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task,’ the study reveals that using generative AI could come at a “potential cognitive cost.”

According to the findings, people who relied on large language models (LLMs) such as ChatGPT, Grok, or Perplexity consistently underperformed across neural, linguistic, and behavioural levels.

The study involved 54 participants divided into three groups, each tasked with an essay writing assignment. The first group wrote essays using only their own cognitive ability. The second group used a search engine, while the third used generative AI tools.

Cognitive load during the task was measured using EEG (electroencephalography), and the essays were evaluated by both human teachers and AI models.

The study found that “Brain-only” and “Search Engine” users exhibited better memory recall and higher activation of the occipito-parietal and prefrontal regions of the brain. In contrast, those in the LLM group reported the lowest sense of ownership over their essays.

The report also highlighted a key insight. LLM users struggled to accurately quote from the essays they had written just minutes earlier. In fact, 83.3% of ChatGPT users were unable to recall exact details from their own work. Brain connectivity was found to have reduced by 47% among this group.

Even from a qualitative standpoint, essays generated with the help of AI were described as lacking creativity and emotional depth. 

Two English teachers who assessed the essays noted, “Some essays across all topics stood out because of a close to perfect use of language and structure while simultaneously failing to give personal insights or clear statements. These, often lengthy essays, included standard ideas, recurring typical formulations and statements. Which made the use of AI in the writing process rather obvious.”

The study raises important questions about the long-term educational implications of LLM dependence. It warns that in choosing short-term efficiency, humans may be giving up long-term cognitive control