Artificial intelligence has become an integral part of modern education, with tools like ChatGPT transforming how students approach writing, research, and problem-solving tasks.
Since its public release, ChatGPT has sparked debates on whether it truly enhances learning or promotes intellectual laziness. Educators and researchers are asking: does AI foster critical thinking, or does it reduce the need for cognitive effort?
These concerns gained new momentum after a recent study by MIT researchers suggested that relying on AI might lead to “cognitive debt,” implying a potential decline in thinking and learning abilities.
WHAT THE MIT STUDY EXPLORED: SETUP AND PARTICIPANTS
To understand the study’s implications, it’s important to look at its methodology. The MIT researchers selected 54 adult participants to write three essays under different conditions over four months.
Participants were divided into three groups: one used only their brains (brain-only), another used a traditional search engine, and the last group used ChatGPT to complete their essays. By comparing these groups, the researchers aimed to observe differences in brain activity, memory recall, and sense of ownership over their written work. Transitioning between these methods helped gauge the impact of each tool on learning.
MEASURING COGNITIVE ENGAGEMENT: BRAINWAVES AND ESSAY ANALYSIS
Cognitive engagement was assessed using both neurological measurements and linguistic analysis. The researchers tracked brain activity while participants wrote essays using the designated method.
They also evaluated the depth and structure of the essays, focusing on coherence, originality, and critical thinking. Additionally, participants were asked to recall quotes from their work to assess memory and engagement. The findings showed that those using ChatGPT exhibited significantly lower levels of brain activity and had poorer recall compared to the other two groups.
WHAT IS “COGNITIVE DEBT” AND WHY DOES IT MATTER?
The term “cognitive debt” refers to the reduced mental stimulation and skill development when individuals overly rely on external tools like AI. Over time, this leads to intellectual stagnation.
According to the MIT researchers, consistent use of ChatGPT diminished participants’ ability to think independently. When these users had to write essays unaided, they struggled with memory, engagement, and content structure. The implication is clear: if AI becomes a crutch too early in the learning process, it could hinder the development of critical academic skills.
BUT IS IT THAT SIMPLE? EXAMINING THE STUDY’S LIMITATIONS
While the findings are thought-provoking, they are not without limitations. Notably, only 18 participants completed the final round of essay writing, with just six per condition.
This small sample size makes it difficult to draw definitive conclusions about cognitive decline due to AI use. Furthermore, the differences in task repetition and experience may have skewed the results. The brain-only group had three chances to practice the task, while the ChatGPT group only experienced writing unaided once. This disparity introduces the “familiarisation effect,” which could explain the improved performance in the brain-only group.
FAMILIARISATION EFFECT VS. COGNITIVE DEBT: A NUANCED DISTINCTION
The familiarisation effect suggests that performance naturally improves with repetition. If the ChatGPT group had practiced essay writing without AI three times, their cognitive engagement might have increased too.
Therefore, the reduced engagement observed in the AI group might reflect task novelty rather than a genuine intellectual deficit. Without additional rounds of testing, it’s premature to attribute these differences solely to AI use.
Nonetheless, the study highlights the importance of understanding how and when AI tools are introduced in the learning process.
LESSONS FROM HISTORY: COMPARING AI TO THE CALCULATOR REVOLUTION
To better understand AI’s educational impact, consider the introduction of calculators in schools during the 1970s. Initially controversial, calculators eventually became essential tools.
However, their integration into education only worked because assessment standards were raised. Students were expected to use calculators for complex problem-solving, not basic arithmetic. Similarly, if educators adapt assessment methods to suit the presence of AI, students can be challenged to think critically with the help of technology—not in place of it.
THE REAL ISSUE: TASK DESIGN, NOT THE TOOL ITSELF
When students are asked to perform tasks that AI can easily complete, it’s unsurprising that they disengage. The problem lies not in using AI, but in what we ask students to do with it.
Requiring students to write standard five-paragraph essays may no longer serve as a true test of learning. Instead, students should be encouraged to apply AI creatively and then explain their decisions, fostering deeper understanding.
For instance, assignments could include using AI to create teaching material or generate multiple viewpoints on an issue, followed by a presentation or oral defense of the output.
CRITICAL THINKING IN THE AGE OF AI: A NEW DEFINITION
The definition of critical thinking must evolve in light of technological advancements. Memorising facts or composing essays unaided may no longer reflect true intellectual capacity.
Today, the real challenge lies in evaluating AI-generated content, verifying its accuracy, and integrating it into broader knowledge frameworks. These are the new hallmarks of cognitive engagement.
Knowing when and how to use AI responsibly will soon be as important as learning to read or write. Avoiding AI entirely may leave students ill-equipped for the future.
EDUCATORS MUST ADAPT: INTEGRATING AI INTO CURRICULA
To prepare students for a world where AI is ubiquitous, educators must rethink how they teach and assess learning. Integrating AI into lessons can promote engagement if done strategically.
For example, teachers might assign collaborative projects where students must guide ChatGPT to produce specific outcomes, then critique those outcomes using class concepts.
Such assignments require creativity, evaluation, and a deep understanding of both the subject matter and the AI tool being used.
MOVING FORWARD: SMARTER USE OF AI FOR SMARTER STUDENTS
Rather than fearing that AI will replace thinking, we should focus on how it can enhance it. Used correctly, AI can free up mental bandwidth for more complex, strategic thinking.
Students can offload routine tasks to AI and concentrate on making connections, asking better questions, and refining ideas—all essential skills in the digital age.
In this context, the MIT study serves more as a cautionary tale about uncritical use than a condemnation of AI itself.
RAISING THE BAR, NOT LOWERING STANDARDS
AI is here to stay. The question is not whether it will impact education, but how we will respond. Will we raise the bar, or let standards slide?
By redesigning tasks, reframing critical thinking, and integrating AI with purpose, we can ensure that students learn to think better—not less. The future of learning isn’t about banning tools like ChatGPT. It’s about teaching students to wield them wisely.