AI Chatbots May Be Weakening Human Thinking Skills, Researchers Warn

AI Chatbots May Be Weakening Human Thinking Skills, Researchers Warn

 

The rapid rise of artificial intelligence tools like ChatGPT, Google Gemini and Claude is raising new concerns among scientists, who say heavy reliance on such systems could be quietly harming human intelligence and cognitive abilities.

Researchers studying how people interact with AI are increasingly warning that “cognitive offloading” — relying on machines to think for us — may come at a cost. While these tools can boost productivity and simplify complex tasks, they may also reduce memory, creativity and critical thinking over time.

At the Massachusetts Institute of Technology (MIT), research scientist Nataliya Kosmyna first noticed something unusual when reviewing job applications. Many cover letters appeared strikingly similar — polished, structured and oddly generic — suggesting widespread use of AI tools.

At the same time, she observed a worrying trend among students: they seemed to struggle more with recalling information compared to previous years. This led her to investigate whether increasing dependence on AI tools was affecting how people think.

Reduced Brain Activity Linked to AI Use

Kosmyna and her colleagues conducted an experiment involving 54 students, dividing them into three groups. One group used ChatGPT, another relied on traditional search engines, and a third completed tasks without any technology.

Participants were asked to write short essays on open-ended topics while researchers monitored their brain activity.

The findings were striking. Students who worked without AI showed strong and widespread brain activity, particularly in areas linked to creativity and reasoning. Those using search engines also demonstrated solid engagement, especially in visual processing areas.

However, the group using ChatGPT showed significantly lower brain activity — reduced by as much as 55%.

According to Kosmyna, while the brain was still functioning, there was noticeably less activation in regions responsible for deep thinking and information processing. This suggests that AI tools may reduce the mental effort required to complete tasks.

Memory and Ownership Also Affected

The study also found that AI users struggled to recall what they had written. Many participants could not quote from their own essays and reported feeling less ownership over their work.

This aligns with broader research showing that reliance on AI tools can weaken memory retention. Similar effects have been observed with earlier technologies — often referred to as the “Google effect” — where people remember less information because they know it can be easily accessed online.

But experts warn that AI could amplify this trend, as it doesn’t just provide information — it can generate complete answers, reducing the need for independent thinking altogether.

Risk of “Cognitive Surrender”

Other studies suggest that users may also become overly trusting of AI outputs. Researchers describe this as “cognitive surrender,” where individuals accept AI-generated responses without questioning them.

This phenomenon has been observed across different fields. In one case, medical professionals using AI tools to detect colon cancer became less accurate when diagnosing cases without assistance after relying on the technology for several months.

Such findings raise concerns that overdependence on AI could weaken core decision-making skills, even in high-stakes environments.

Creativity and Original Thinking at Risk

Another key concern is the potential loss of creativity. Essays written with AI in Kosmyna’s study were described as uniform, lacking originality and depth. Teachers reviewing the work even questioned whether students had collaborated due to the similarity in style.

This suggests that while AI can produce technically strong outputs, it may discourage unique thinking and personal expression.

Long-Term Cognitive Impact Still Unclear

Although many of these studies focus on short-term effects, researchers are increasingly worried about long-term consequences.

Computational neuroscientist Vivienne Ming has warned that reduced mental effort — particularly lower gamma brainwave activity — could be linked to cognitive decline over time. Some studies associate weaker brain activity with a higher risk of conditions like dementia.

Ming argues that deep thinking is essential for maintaining brain health. If people consistently rely on AI for answers instead of engaging mentally, it could have lasting consequences.

There are parallels with other technologies. For example, research has shown that heavy use of GPS navigation can weaken spatial memory, a skill linked to early signs of Alzheimer's disease.

Finding a Balance with AI

Despite these concerns, experts do not suggest abandoning AI tools altogether. Instead, they emphasise the importance of using them wisely.

One proposed approach is “hybrid intelligence,” where humans actively think and analyse while using AI as a support tool rather than a replacement. This means engaging with problems first, then using AI to refine or challenge ideas.

Ming also recommends techniques such as asking AI to critique your thinking instead of providing direct answers. This creates what she calls “productive friction,” encouraging deeper engagement rather than passive acceptance.

Ultimately, researchers agree that maintaining cognitive health requires regular mental effort. While AI offers powerful capabilities, over-reliance could undermine the very skills that make human thinking unique.

As AI becomes more integrated into everyday life, the challenge will be finding the right balance — using technology to enhance intelligence without replacing it.

Post a Comment

0 Comments