Dar es Salaam — As AI-powered tools like ChatGPT
and Deepseek become increasingly integrated into classrooms across the globe, a
groundbreaking study from the MIT Media Lab is raising fresh questions about
how these technologies are reshaping the way students learn—and think.
Led by Dr. Pattie Maes, a
renowned professor of Media Technology, and Dr. Cynthia Breazeal, a pioneer in
social robotics and human-computer interaction, the research team set out to
investigate how students engage cognitively when using AI compared to
traditional methods. Their findings are as thought-provoking as they are
urgent.
The study involved several
student groups tasked with writing essays under different conditions: some
worked independently, others had access to AI tools, and a third group blended
both approaches. Researchers measured brain activity through EEG sensors and
analyzed the linguistic complexity and originality of each essay. The goal? To
understand not just the quality of the writing, but the depth of thought behind
it.
The results were striking.
Students who leaned heavily on AI showed noticeably lower levels of cognitive
engagement. They struggled to recall specific quotes from their essays and
reported feeling less connected to their work. “It was like I wasn’t the one
who wrote it,” admitted one participant. Many described a sense of
detachment—an emotional distance that reflected in both their brains and their
words.
In a revealing twist, students
later switched roles. Those who had initially written without AI were allowed to
use it, and vice versa. The group returning to manual writing after using AI
performed worse and exhibited reduced brain engagement—a phenomenon the
researchers termed “cognitive debt.” Like muscles atrophying from lack of use,
the mind, too, seems to dull when over-reliant on external thinking tools.
This raises tough questions for
educators worldwide. If students are losing the ability to think deeply and
critically without AI, how should schools adapt? Many current curricula were
designed long before generative AI became mainstream. Yet today, with a few
prompts, students can generate full essays in seconds. Traditional assignments
no longer offer a reliable gauge of comprehension or original thinking.
Still, the MIT researchers stop
short of calling for a ban on AI in education. Instead, they argue for a
balanced approach. “AI, when used wisely, can free up time for deeper inquiry
and creative exploration,” said Dr. Maes. “But without guidance, it can become
a crutch that limits growth rather than enhances it.”
In Tanzania, where digital
learning is accelerating, the implications are especially relevant. Kusiluka
Aginiwe, innovation lead at Ekima Interactive, framed it bluntly: “Technology
is like water—you can’t stop it. You can only direct it.” He warns that
completely restricting AI use in schools would be shortsighted, especially when
industries are rapidly adopting AI across the board. “If we deny students the
chance to learn AI responsibly now, how will they thrive in a workforce that
increasingly demands it?”
Experts across the board agree:
the future of education lies not in resisting AI, but in redefining learning
goals. That means focusing on skills that machines can't replicate—like
critical thinking, ethical reasoning, creativity, and emotional intelligence.
Assignments should be redesigned to reward original thought and process, not
just polished output.
Ultimately, this isn’t just a conversation about technology. It’s about how we prepare the next generation to navigate complexity, ambiguity, and change. Used recklessly, AI might dilute intellectual rigor. But used wisely, it could empower students to think more boldly than ever before.
