Research indicates that artificial intelligence (AI) can distort people's memories, leading them to mistakenly believe events occurred that did not actually happen.
The Washington Post reported on October 15 that a study highlights these risks.
A team led by Patty Maes, a professor at the Massachusetts Institute of Technology (MIT) Media Lab, showed 24 ordinary photographs to 200 participants. Some participants were then presented with manipulated images altered by AI, and their memory recall was tested.
The results showed a significant increase in the rate of false memories among those who viewed the AI-modified images, with participants expressing greater confidence in their inaccurate recollections.
“I fundamentally see all of this as very dangerous,” Maes warned.
In another experiment, Maes' team measured brainwaves of 54 participants who wrote essays. Those who wrote without AI tools exhibited stronger connectivity between brain regions, while those who used ChatGPT showed weaker connectivity. Participants using ChatGPT also struggled significantly to reference their own recent writing.
Brain connectivity refers to the degree of interaction between different brain regions responsible for memory, language, and judgment.
However, the Washington Post noted that these findings do not prove that regular AI use inevitably leads to cognitive decline or permanent memory loss. The related research is still in its early stages.
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.