Outsourcing the mind: When convenience replaces cognitive effort 

Walk into almost any community in Bangladesh today — rich or poor, urban or rural — and you are likely to find a young person with a smartphone. Mobile internet has become remarkably widespread, even if access and quality still vary.

Even in remote areas, where digital access remains uneven and quality education is often limited, a teenager can open an AI chatbot and get an instant answer to any question — a math problem, an English essay, a science project. 

This is, by any measure, a remarkable democratisation of knowledge. It is also a quiet and largely unmonitored experiment in what happens to a human mind that is never asked to wrestle with a difficult question. 

I do not make this argument lightly, nor do I make it a technophobe. I use Gen AI frequently in my own teaching and learning. But my discipline — information systems — has spent decades studying what happens when humans hand decisions to automated systems without understanding how those systems work. 

We have a term for the failure mode: "automation complacency". We are now running that experiment on the young generation, on a global scale, with almost no mentors, and the youth, least equipped to notice the damage, are the youth most exposed to it. 

But I want to raise a quiet worry. Not about the technology itself — about what happens to a young mind that never has to struggle.

Why struggle matters

Think about how anyone actually learns to think. You have a problem. You feel confused. You try something. It doesn't work. You try again. Slowly, an idea forms — and it is yours. That process of struggling, failing, and trying again is not a side effect of learning. It is learning.

Now imagine a student who never goes through that process because the AI hands them a finished answer in 5-10 seconds. They copied it. They never wrestled with the problem themselves. The assignment is complete, but nothing was learned.

The mechanism is not mysterious

Scientists call this "cognitive offloading" — handing your thinking over to a tool instead of doing it yourself. A major 2025 study led by Michael Gerlich found that the more often young people use Gen AI, the weaker their critical thinking scores become — and the effect was strongest among the youngest users, those still developing this skill. 

Other researchers have described students who stop questioning AI answers altogether, simply accepting whatever the machine says, as many of us used to do with the first Google result. This matters everywhere. But it matters more in a country like ours.

Why Bangladesh is a special case

Bangladesh has done something remarkable: it has put a smartphone in almost every youth's hand, in cities and villages alike. But access to a smart device is not the same as knowing how to use it wisely. 

Surveys show that while smartphone ownership has spread rapidly, digital literacy — the ability to question, verify, and think critically about what a screen shows you — has not kept pace. A student in Dhaka (and/or in an urban region), with an expert mentor, is far more likely to be taught how to debate an AI's answer than a student in a rural area, with an overcrowded classroom and an underpaid, overworked teacher.

In doing so, we risk developing a dualistic nation marked by digital literacy disparity. One where AI is a tool that sharpens young minds, because someone taught them how to use it critically. And one where AI quietly replaces thinking altogether, because no one taught them the difference.

That second group — youth in hard-to-reach areas, first-generation learners, and youth whose parents never attended a higher-education institution — is exactly the learner this country most needs to lift up.  

If AI hollows out their ability to think independently rather than strengthening it, we will have widened the gap between the rich, middle-income, and poor, between city and village, not closed it, no matter how many smartphones we hand out.

Youths in hard-to-reach areas, first-generation learners, and whose parents never attended a higher-education institution are exactly the learners this country most needs to lift up. If AI hollows out their ability to think independently rather than strengthening it, we will have widened the gap between the rich, middle-income, and poor, between city and village, not closed it — no matter how many smartphones we hand out.

An impactful, imperfect analogy

In 2024, Nature published landmark research on "model collapse", showing that when AI models are repeatedly trained on AI-generated content rather than original human-created data, their performance gradually deteriorates. It showed that when an AI model is repeatedly trained on content generated by earlier AI models — rather than on novel, original human thinking — it slowly degrades. 

Details get flattened, rare and original patterns disappear first, and the system converges toward a narrower, blander version of itself. It is not a perfect metaphor for human learning, and I want to be careful not to overstate it — a youth's mind is not a training corpus. 

But the underlying warning is this: a system, artificial or human, that stops learning from original, effortful, first-hand engagement with a problem and instead relies on recycled processed outputs will tend toward homogeneity, not creativity. 

If an entire generation's first draft of thinking is outsourced before it is ever attempted independently, we should not be surprised if what comes back is narrower than what a mind wrestling with the problem alone would have produced.

What 'human in the loop' requires

The phrase "human in the loop" gets used loosely in AI ethics circles, usually to describe a system design choice. 

I want to argue it should also be a pedagogical obligation. A human genuinely in the loop is not a person who reviews an AI's output after the fact; it is a person who has first attempted the problem, formed an opinion, and only then compared it against the machine's answer. That ordering — struggle first, verify second — is what preserves the skill. Reverse it, and the "loop" becomes decorative.

What should be done

Research shows that when students are taught to use AI critically, they are asked to check its answers, argue with it, and revise its output rather than copy it — their thinking actually improves. The danger is not AI. The danger is AI without a mentor to guide its use.

For Bangladesh, I love to recommend three simple, affordable steps:

Teach AI literacy in every educational institute, not just urban schools

Before youths use AI to answer questions, teach them, in one or two lessons, how these tools actually work and why they can be wrong. AI literacy can turn Gen AI from a shortcut into a learning scaffold. 

In low-resource schools, students do not need advanced technical training; they need practical habits: try first, question AI answers, check sources, identify hallucinations, and explain their reasoning. Even simple classroom routines — comparing AI responses with textbooks, requiring drafts, and asking students to reflect on what they changed — can preserve independent thought. 

Proper AI literacy helps vulnerable learners gain the benefits of access without becoming dependent on machine-generated answers.

Grade the thinking, not just the final answer

Ask students to show their work, their drafts, their reasoning — not just a polished essay that could have been generated in ten seconds.

Put the human first, the machine second

Encourage students to try a problem themselves before checking the AI's answer — not the other way around. That order is what protects the skill.

Bangladesh's greatest resource has always been its people — young, hardworking, and hungry to learn. Gen AI can be an extraordinary tool to lift that generation up. But only if we teach them to think with it, not to let it think for them. The smartphone is already in their hands. What we put in their mind is still up to their usage behaviours.

Najmul Hasan is an associate professor of Business Analytics and Information Systems at BRAC Business School, BRAC University, Dhaka. 

Disclaimer: The views and opinions expressed in this article are those of the author and do not necessarily reflect the opinions and views of The Business Standard.