The Implications of AI on Learning

In the talk about the role of gestures in learning, Dr. Delmar spoke about how gesture can be seen as both an index of change as well as a part of the process of change. This includes how the relationship between gestures and speech aids in learning by minimizing cognitive load and creating cognitive instability. Specifically, in this post, I would like to focus on how the minimization of cognitive load and creation of cognitive instability through gestures facilitate learning. One of the mechanisms of learning through cognitive instability was through gesture speech mismatches. In the review shared by Dr. Delmar, it is mentioned that children who express "two ideas – one in speech and another in gesture," a mismatch, are shown to be more receptive to instruction.  The possible mechanism stated in the text is that a mismatch causes cognitive instability, which essentially works out the brain, leading to learning. Furthermore, when referring to the minimization of cognitive load, gesturing helps save cognitive resources by shifting part of a task away from the verbal store to the visuospatial store so that more resources can be allocated toward another aspect of that task.

In the wake of AI becoming a prevalent tool in schools across the world, these ideas about cognitive load and cognitive instability as they pertain to learning become even more relevant. However, the question is what the consequence may be as students are increasingly relying on this tool to learn. In a recent study done at MIT, researchers analyzed the cognitive cost of using AI in educational contexts. In the study, students completed an essay using search engines, their brain only or AI. Using electroencephalography (EEG), researchers recorded participants’ brain activity to assess their cognitive engagement and cognitive load. Based on the data, researchers found reduced brain connectivity in the students who had external support. The students using AI had the weakest level of coordinated activity between different brain regions, indicating a lower cognitive load. However, unlike gestures, this decreased cognitive load did not have the same effect in facilitating learning. Instead, participants had decreased memory recall about the content after the task.

Unlike AI, gesture promotes learning by creating cognitive instability by the simultaneous activation of two ideas. Although gestures also lessen cognitive load, they still keep the learner engaged in the content and challenge them to actively retrieve ideas and synthesize their own thoughts. Although lessening cognitive load may be beneficial in learning, the MIT study clearly demonstrates its harm when performed incorrectly. It should be noted that the way in which gestures minimize cognitive load varies a lot from AI. In gesturing, part of the task is shifted to a different brain region, not offloaded to a different cognitive system entirely. The common phrase “if you don’t use it, you lose it” is incredibly applicable here. When students stop engaging the neural circuits responsible for critical thinking by outsourcing it to AI, they might severely hinder their ability to do so in the future. This is especially concerning at the university level, where, in my opinion, the goal is not to simply regurgitate information you have been told, but to cultivate the capacity for deep critical thinking and independent thought.

References:

GOLDINMEADOW, S, and S WAGNER. “How our hands help us learn.” Trends in Cognitive Sciences, vol. 9, no. 5, May 2005, pp. 234–241, https://doi.org/10.1016/j.tics.2005.03.006.

Kosmyna, Nataliya, et al. “Project Publications ’ Your Brain on Chatgpt.” MIT Media Lab, www.media.mit.edu/projects/your-brain-on-chatgpt/publications/. Accessed 10 Dec. 2025.