Does ChatGPT make you stupid? MIT study suggests people who rely on AI tools are worse off.

by

With the growing interest in generative AI like ChatGPT, scientists are keen to explore its impacts. Past research indicates that excessive reliance on these tools can result in problematic behaviors akin to addiction, negatively impacting well-being, weakening critical thinking skills, and fostering feelings of isolation.

Brain Activity Decrease and Skill Deterioration,'” further proposes that individuals who heavily rely on such AI tools may see a decline in their essay-writing abilities when required to perform the task independently, without the aid of ChatGPT or similar tools.

In this research, we worked with a relatively small number of participants (54) who fell within the age range of 18 to 39 years old. These individuals were divided into three distinct categories: one group focused solely on brain activity, another group was able to utilize conventional search engines, and the third group had access to advanced language learning tools like ChatGPT.

Individuals who employed LLMs for their tasks found a significant reduction in mental effort, approximately 32%, compared to those using conventional software interfaces. Furthermore, they reported reduced levels of frustration while going through the process.

One significant insight emerged: Individuals employing AI tools like LLMs during their research found they exerted less cognitive energy to obtain information. Although this isn’t shocking, it clarifies why some might find themselves drawn to using AI tools for completing tasks.

Alongside detecting increased brain function, individuals in the brain-only category also expressed greater contentment concerning ethical aspects.

instead of fading away, it has evolved to influence user experiences by means of content tailored algorithmically. The content rated as top is essentially shaped by the interests of the LLM’s stakeholders.

Initially, the findings of the research seem self-evident. It’s hardly shocking that composing an essay demands more mental effort compared to using a tool for the same task. It’s reasonable to assume that delegating the task to another person would also require limited cognitive exertion as well.

Science isn’t just about guesswork; it involves gathering evidence and testing theories. Given that decisions regarding government policies and educational standards might be influenced by the impact of models like ChatGPT and others, it’s crucial to accumulate as much data as we can.

The MIT researchers said as much in the conclusion of the study.

It’s crucial to conduct long-term research on LLMs (Living Laboratory Models) to fully grasp their potential long-term effects on the human brain, so we can confirm that they offer more benefits than drawbacks for people.

The MIT research, which hasn’t undergone peer review yet, was based on a limited number of cases. Although it’s an initial move in accumulating valuable data, further investigation is needed to confirm its findings.

Is ChatGPT making people more stupid?

Additionally, the research investigated the impact on performance when individuals initially used a Language Learning Model (LLM) for essay writing, followed by attempting the identical task without any artificial intelligence support afterwards.

The researchers found that the observed trend might be due to people relying more on AI systems like ChatGPT, thereby reducing their own mental processing. This means that by using such tools, one may engage in less thinking, potentially leading to a decrease in cognitive ability over time.

As an enthusiast, I recently came across an intriguing exploration by Microsoft researchers, in conjunction with Carnegie Mellon University, delving into how generative AI influences critical thinking. In simpler terms, they drew parallels between the impacts of using generative AI and muscles weakening over time.

If data consistently supports the idea that employing generative AI models leads to reduced brain activity and critical thinking over time, it may result in broad implications.

According to a study conducted by The Guardian, as mentioned in the AI Commission report, it was discovered that AI-driven tools like ChatGPT are frequently employed by students to engage in academic dishonesty.

Using AI tools like ChatGPT extensively might lead students and workers to develop a reduced ability to perform specific tasks independently.

Read More

2025-06-19 15:39