FLASH
Shkruar nga Pamfleti
A recent study from the Massachusetts Institute of Technology (MIT) has raised serious concerns about the neurocognitive impacts that prolonged use of large language models (LLMs), such as ChatGPT, may have, particularly on academic and cognitive tasks.
The researchers followed 54 university students for a four-month period, divided into three groups: those who used ChatGPT, those who used traditional search tools (like Google), and those who worked without any technological assistance.
During and after completing the tasks, participants' brain activity was monitored via electroencephalography (EEG).
The results showed that regular use of ChatGPT for writing tasks was associated with significant decreases in neural activity in brain areas associated with memory consolidation and executive functions.
LLM users exhibited less mental engagement, poorer recall of content they had written themselves, and less originality.
This phenomenon was described by the authors as a form of "mental passivity".
The study, titled “The Cognitive Cost of Using LLMs,” also noted the risk of over-reliance on AI-generated answers, fostering the creation of echo chambers and hindering critical judgment.
What raises more concern is that even when participants who had previously used ChatGPT switched to tasks without the assistance of AI, they continued to show poor brain engagement, suggesting a lasting impact on cognitive flexibility.
On the other hand, those who had initially worked without the help of AI and were then exposed to it showed increased brain activity, suggesting that initial cognitive engagement may help in healthier use of such technologies.
This study highlights that, while artificial intelligence tools offer major efficiency gains, their integration into educational and clinical settings must be done with caution.
AI can be most effective when used as an aid, not a replacement for active human thinking.
Lini një Përgjigje