Researchers warn that early use of AI will harm fact-retention and critical thinking.
A new preprint study from MIT reveals that using AI chatbots to create tasks significantly reduces brain activity and impairs memory retention.
The study, led by Dr. Nataliya Kosmyna of MIT Media Lab, was involved in students from the Boston area writing essays under three conditions: Participants wore EEG headsets to monitor brain activity through brain activity.
Results showed that people who rely on AI showed the weakest neural connectivity with up to 55% lower cognitive involvement than the non-adjuvant group. Those using search engines showed a moderate drop of up to 48%.
Researchers used dynamically oriented forwarding functions (DDTFs) to assess cognitive load and information flow across brain regions. They found that while the auxiliary groups have activated a wide range of neural networks, AI users are primarily engaged in procedural tasks with shallow encoding of information.
Participants using the GPT-4o showed their worst in recall and recognized ownership of the written work. In the follow-up session, students who previously rely on AI struggle more when the tool is removed, suggesting a decline in internal processing skills.
Meanwhile, those who previously used their own cognitive skills had improved performance when they were later given AI support.
Findings suggest that early use of AI in education may hinder deeper learning and critical thinking. Researchers recommend that students engage in self-directed learning first and incorporate AI tools to enhance understanding.
Dr. Kosmyna emphasized that although the results are preliminary and have not yet been peer-reviewed, the study highlighted the need to carefully examine the cognitive impacts of AI.
The MIT team is currently investigating similar effects in task coding and studying how AI tools like code generators affect brain function and learning outcomes.
Want to learn more about AI, technology, and digital diplomacy? If so, ask our Diplo chatbot!
