According to the latest research by researchers at the Massachusetts Institute of Technology (MIT), artificial intelligence chatbots like OpenAI's ChatGPT may weaken cognitive abilities.
In a study conducted at the MIT Media Lab, 54 participants completed three rounds of essay writing tasks using three designated methods and tools: ChatGPT, search engines, and relying solely on their own brains.
In the fourth round, the group using ChatGPT was asked not to use any tools, while the group relying solely on their brains was instructed to use a large language model (LLM).
The results were shocking, as AI users exhibited poorer memory recall abilities, with over 83% of ChatGPT users unable to cite content from essays they had written just minutes earlier.
Alex Vacca, co-founder of the sales technology agency ColdIQ, described this outcome as "frightening," adding that AI is not making us more efficient, "it is making us cognitively bankrupt."
"You write something down, click save, and your brain has already forgotten it because ChatGPT did the thinking for you."
Researchers noted that brain connectivity "systematically decreases" with increased external support.
"The group relying solely on their brains exhibited the strongest and most extensive networks, the search engine group showed moderate engagement, while the LLM-assisted group had the weakest overall coupling," they stated.
The researchers used electroencephalography (EEG) to record brain activity to assess participants' cognitive engagement and cognitive load during the tasks.
They found that repeated reliance on external systems like LLMs could lead to the accumulation of "cognitive debt," as these systems replace the cognitive processes required for independent thinking.
"Cognitive debt delays mental labor in the short term but incurs long-term costs, such as diminished critical inquiry abilities, increased susceptibility to manipulation, and decreased creativity."
This yet-to-be-peer-reviewed paper suggests that using AI LLMs may actually impair learning abilities, especially for younger users.
"In this study, we highlighted the urgent issue of potentially declining learning skills based on our findings," the researchers concluded.
They stated that "longitudinal studies" are needed to understand the long-term effects of AI chatbots on the human brain, "before LLMs are considered a net positive for humanity."
When Cointelegraph asked ChatGPT for its opinion on the study, the chatbot replied, "This study does not claim that ChatGPT is inherently harmful—rather, it warns against over-reliance without reflection or effort."
Related: Elon Musk's AI company xAI faces lawsuit over air pollution at data center
Original article: “Need Help from ChatGPT? It Might Make You Dumber”
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
