DT Next Explains | Too much ChatGPT may reshape your brain's learning loop, kill cognitive ability

CHENNAI: From finding things in a 200-page document, pinpointing the exact portion of my need, getting context for something or just simply brainstorming ideas, people see and use ChatGPT as ‘Google on steroids.’ Do we completely rely on it? No. But what if we do? A recent study by the MIT Media Lab tried to answer the question.

DT Next tries to break it down with study insights from an AI expert and a neurologist.

"In younger individuals, the brain is still wiring itself. It's called neuroplasticity, and it peaks in your 20s," says Dr Sruthi Degapudi, consultant neurologist at Prashanth Hospitals. "So when someone repeatedly uses AI to write without a first attempt, it becomes a habit that can reshape the brain's learning loop."

Researchers at the MIT Media Lab, along with collaborators set out to measure the cognitive cost of using AI tools like ChatGPT in education. They focused on one task - Essay writing. A total of 54 students (aged 18–39) were split into three groups. Brain-only group: wrote essays with no external help. Search group: used Google, but no AI. LLM group: used ChatGPT and nothing else.

Participants using no tools (brain only) showed the highest brain connectivity; search group participants showed moderate activity. Those using AI had the weakest neural effort. In Session 4, those who had previously used ChatGPT were then asked to write without it. They displayed lower brain activation than their counterparts. By contrast, those who started without AI and then used it showed higher neural connectivity.

This shift isn't unfamiliar to technologists. "When people begin relying on AI for everything, from summarising documents to forming opinions, it becomes easy to mentally disengage," says Ramprakash Ramamoorthy, ManageEngine, Director of AI Research at Zoho Corp. "But that's not a flaw in the technology. It's about how we use it."

He draws parallels with earlier tech shifts – from GPS dulling our sense of direction to phones eroding memory for numbers. "It should leave you sharper, not passive, "he says.

The MIT study itself did not argue that AI harms the brain. "We are interested in how cognitive load shifts when it's used, and what happens when it becomes habitual." The word habitual is the pivot.

Because what matters isn't just how we use AI but what we stop doing when we use it. Is there a tipping point where such cognitive offloading becomes counterproductive?

Dr Sruthi doesn't see such cognitive offloading as inherently bad. "We do it all the time, like using calculators. But it depends on what's being offloaded. If a literature student turns to AI to write essays, they're outsourcing the very skill they're supposed to build. That's very different from using AI to look up data or reorganise notes."

Ramamoorthy echoes this. "LLMs can boost productivity and creativity when used intentionally," he says. "I've seen designers and professionals use them to challenge assumptions or test early ideas. That's augmentation, not replacement."

The MIT study doesn't draw dramatic conclusions. It merely offers a lens to see how our cognitive patterns shift as we delegate more and more mental effort to tools like ChatGPT. The effects may not show up immediately. But over time, what we repeatedly outsource might slowly become what we stop practicing.

What is AI costing humans?

*Cognitive skill: Too much interaction with technology pushes individuals to think like algorithms without understanding basics and concepts. Employees turn to AI to complete routine tasks, losing out on opportunities to refine their cognitive abilities, leading to mental atrophy that can limit independent thinking. Students accept AI-generated answers without fully understanding basic concepts, leaving them with poor analytical skills.

Critical thinking: AI Improves efficiency but reduces critical engagement in routine tasks, raising concerns of diminished problem-solving skills. AI affects the human ability to think, reason and solve problems

Adaptability: Human brain can process new, unfamiliar situations and improvise. When we integrate AI into decision-making processes, we stifle our own critical thinking capabilities. AI can limit the human brain's thinking capacity, starve the human brain of thoughtfulness, and mental efforts as it penetrates deeper into activities like planning and organisation.

Imagination: Over dependence on AI to solve problems hampers the human ability to think outside the box. AI can mimic art but cannot invent a new art form

Human interaction: Dependence on AI for human connection, like AI customer service tools and chatbots, can result in diminished human interaction