When AI Makes Us Smarter—But Also Weaker

Artificial intelligence has quickly become a trusted companion in daily life. From writing essays to solving complex coding problems, AI tools promise speed, efficiency, and convenience. But a recent study suggests there may be a hidden downside: the more we rely on AI, the less we rely on our own minds.

A joint study conducted by researchers in the United States and the United Kingdom—reported by Engadget—raises concerns about how AI affects human cognition. Titled “AI Assistance Reduces Persistence and Hurts Independent Performance,” the research highlights a troubling pattern: while AI boosts short-term performance, it may weaken long-term mental resilience and independence.

The concept at the center of the study is something researchers call a “cognitive cost.” In simple terms, this means that although AI helps us complete tasks faster, it may reduce our ability to think critically and persist when challenges arise.

The findings are striking. In experiments involving hundreds of participants, researchers observed that just ten minutes of AI assistance was enough to make users more dependent. When that assistance was suddenly removed, performance dropped sharply. Even more concerning, many participants simply gave up rather than attempting to solve problems on their own.

The study focused on cognitively demanding tasks such as writing, coding, and brainstorming. In one experiment, 350 participants were asked to solve fraction-based math problems. They were divided into two groups: one with access to a GPT-5-based chatbot, and another working independently.

Halfway through the test, researchers removed AI access from the assisted group. The result? A dramatic decline in correct answers—and a noticeable drop in motivation. Many participants stopped trying altogether.

This pattern was repeated in a second experiment involving 670 participants working on reading comprehension tasks. According to Rachit Dubey, an assistant professor at University of California and one of the study’s authors, the issue goes beyond incorrect answers.

“When AI is taken away, people don’t just struggle—they often refuse to try,” he explained. “Their persistence drops significantly.”

A Silent Threat to Creativity and Learning

The implications of these findings extend far beyond individual productivity. Dubey warns that rapid AI adoption in education could create a generation that underestimates its own potential. If students rely too heavily on AI, they may never fully develop the creativity and problem-solving skills needed for innovation.

The study compares this effect to the “boiling frog” phenomenon: gradual changes that go unnoticed until it’s too late. Over time, constant reliance on AI may quietly erode motivation, discipline, and intellectual curiosity—qualities essential for lifelong learning.

What Happens in the Workplace?

Interestingly, the research also revealed a more nuanced picture in professional settings. Participants who used AI as a support tool—asking for hints or clarification rather than full answers—adapted much better when AI was removed. This suggests that how we use AI matters just as much as how often we use it.

However, a new concern is emerging among workers: “AI brain fry,” a term used to describe mental fatigue caused by overreliance on AI tools. Ironically, instead of making work easier, AI can sometimes lead to longer working hours and increased pressure to perform.

Education at a Crossroads

Perhaps the most visible impact is in education. Previous studies have already shown that students who depend heavily on AI tools tend to perform worse in exams and exhibit weaker social and intellectual development. Learning, after all, is not just about getting the right answer—it’s about the process of thinking, struggling, and understanding.

A Tool, Not a Replacement

Although the study has not yet undergone full peer review, its message is clear: AI should be used carefully. It is a powerful tool, but not a substitute for human effort.

The key lies in balance. When used wisely, AI can enhance learning and productivity. But when it replaces thinking altogether, it risks diminishing the very abilities that make us human.

In the end, the question is not whether we should use AI—but how we can use it without losing ourselves in the process.