The future of our ability to think assertively

A growing body of early research suggests we may be outsourcing more of our cognitive effort than we realize — and the consequences could be subtle but significant.

A recent research, last year, of a group led by MIT scientist Nataliya Kosmyna set out to observe what actually happens inside the brain when students rely on AI tools while writing. Instead of simply comparing the quality of essays, the team used electroencephalogram (EEG) sensors to track neural activity in real time. Their 54 participants were divided into three groups:

one group wrote essays with the help of ChatGPT

another used Google search (without AI summaries)

the last group relied solely on their own knowledge

Each student wrote one essay per month for three months. A smaller subset switched conditions in a fourth month so the researchers could observe what changed when someone started — or stopped — using ChatGPT.

According to the unpublished paper, the students using ChatGPT showed lower neural activation in regions associated with creativity, information processing, and sustained cognitive effort. Natalya Kosmyna described it bluntly in an interview [“Nataliya Kosmyna: Cognitive Debt Accumulates, You Can’t Pay It Back”]: the brain didn’t “shut off,” but it was clearly doing less work.

Even more concerning, the decline wasn’t static. With each new essay, the ChatGPT group showed progressively weaker engagement, suggesting a creeping habit of mental offloading.

This wasn’t just visible in brain data. Behaviorally, these students: produced less original writing ; struggled to recall what they had written ; and, in some cases, generated essays so similar that a teacher involved in the study wondered whether students were copying each other.

That last point echoes a broader concern: when many people use the same AI tool, the outputs can converge, flattening individual voice and originality.

The research findings aren’t isolated. Another recent study — one of the first to claim causal evidence — argues that relying on AI tools can gradually weaken our reasoning abilities. The researchers compared the effect to a boiling frog: the decline is slow, subtle, and easy to miss until it’s well underway.

Similarly, a University of Pennsylvania team found that when participants were given the option to use ChatGPT to answer reasoning questions, most people chose the AI even when they could have solved the problems themselves. The researchers called this “cognitive surrender” — a willingness to hand over thinking simply because the option exists.

Outside the lab, many AI users report something similar. People say they feel less creative, less articulate, or more dependent on AI to express complex ideas. Teenagers, in particular, describe AI tools as addictive and worry that constant use is dulling their ability to think independently.

Of course, it’s worth noting — as I’ve written in a previous article — that stories about “AI destroying our brains” tend to resurface every few months, often wrapped in dramatic headlines designed to provoke fear. Not every study is conclusive, and many are still in early stages.

But even with that caveat, the emerging pattern is hard to ignore. We may not fully understand how AI reshapes cognition, but the early signals suggest caution. As many researches put it, our brains love shortcuts — and that’s exactly why we need to be mindful of how often we take them.

If we rely too heavily on AI for everyday thinking, we will weaken our ability to think assertively, creatively, and independently. Not because AI is harmful by nature, but because humans tend to choose the path of least resistance — and AI makes that path extremely smooth.

AI doesn’t automatically erode thinking. What erodes thinking is uncritical dependence — what researchers call cognitive offloading or cognitive surrender. If someone uses AI as a partner — to challenge ideas, expand perspectives, or test reasoning — it can actually sharpen thinking. If someone uses AI as a crutch — to avoid effort — it dulls the process of thinking.

The issue isn’t whether AI will make us less assertive thinkers. The issue is whether we will choose to stay mentally assertive in an era where thinking is optional.

©Ca De Luce> MINDFUL MIND MEDIUM Blog 2025. Unauthorized use of text or media is not allowed. All images and photo are fulfilling the copyrights regulations. Much obliged to you all!