Learning Energy for Sustainable Learning with Artificial Intelligence in Higher Education
This perspective article introduces the "Learning Energy" framework, a multidimensional model integrating five psychological theories to explain how generative AI can either deplete or energize higher education learners through resource appraisal, thereby offering a comprehensive approach to understanding and sustaining academic well-being in AI-mediated environments.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
In the modern university, the act of learning has quietly changed its texture. For decades, education has relied on understanding how students stay motivated, how they engage with difficult material, and how they maintain their mental well-being. These are familiar concepts: motivation is the reason a student starts a task, engagement is the effort they put into it, and well-being is the sense that they are functioning healthily. Yet, a new force has entered the classroom, one that does not just sit on a shelf but actively participates in the thinking process. Generative artificial intelligence, the technology that can draft essays, explain complex ideas, and offer instant feedback, has moved from the fringe to the center of daily academic life. It offers speed and clarity, but it also brings a paradox. A student might finish an assignment faster while thinking less, or they might use the same tool to deepen their understanding and verify their own ideas. The question facing educators and researchers is no longer just whether this technology works, but how it changes the internal fuel that keeps a learner going.
A new perspective from the University of Phuong Dong proposes a way to measure this internal fuel, calling it "Learning Energy." This concept is not a single feeling like excitement or tiredness, but a dynamic resource that allows a person to start, sustain, and regulate their learning. The author, Van Huong Nguyen, suggests that this energy is renewable and socially shared. It is not something a student carries alone in their head; it is generated and drained by how they interact with the technology, their peers, their teachers, and the rules of their institution. The paper argues that we cannot understand the impact of artificial intelligence by looking only at the final grade or the speed of the work. Instead, we must look at whether the technology helps a student think clearly, feel safe enough to take risks, persist through difficulty, connect with others, and remain the true author of their own work.
The research begins by observing that existing ideas about learning are too fragmented to explain what is happening now. We have theories about motivation, about how people manage their resources, and about how social connections help us act, but they are often studied in isolation. The paper weaves these separate threads into a single, coherent framework. It suggests that learning energy has five distinct but connected dimensions. First is cognitive energy, the mental clarity needed to process information. Second is emotional energy, the sense of safety and confidence that allows a student to handle uncertainty. Third is motivational energy, the force that keeps a student working when a task gets hard. Fourth is social-relational energy, the readiness to learn that comes from feeling connected to peers and teachers. Finally, there is agentic energy, the capacity to remain the decision-maker and the accountable author of the work.
The framework explains that artificial intelligence can either build up this energy or drain it, depending on how it is used and how the student perceives it. When a student uses the technology to clarify a confusing idea, get a quick example, or rehearse a difficult concept without fear of judgment, the tool acts as a generator. It lowers the barrier to starting a task and can make the student feel more capable. However, the same tool can become a drain if it leads to over-reliance. If a student simply copies the output without thinking, or if the technology replaces the need to verify facts, the mental effort required to learn disappears, and with it, the sense of ownership. The paper notes that this can lead to a state of "efficient isolation," where a student produces work quickly but feels disconnected from their peers and unsure of their own abilities.
To make sense of these different outcomes, the author outlines four possible profiles of how students experience this environment. One group, lacking both strong human support and effective use of the technology, may feel depleted and anxious. Another group, supported heavily by teachers and peers but with less reliance on the technology, may show resilience. A third group, using high levels of artificial intelligence but lacking human connection, may find themselves in a state of efficient isolation, where speed comes at the cost of deep learning and social belonging. The ideal profile, described as "energized co-agency," occurs when high-quality artificial intelligence is combined with strong human support. In this scenario, the technology helps generate ideas, but the student, guided by teachers and peers, verifies, critiques, and refines those ideas. The result is a learning process where the student feels clear, confident, connected, and fully in charge.
The paper emphasizes that this energy is not a fixed trait but a cycle that rises and falls. A student might start a task with high energy, using the technology to get a quick start. As the work progresses, if the demands for verification and critical thinking grow without support, that energy can deplete. However, if the student pauses to reflect, discusses the work with a peer, or receives clear guidance from a teacher, the energy can be regenerated. The framework suggests that sustainable learning depends on this ability to recover. It is not enough to simply have access to powerful tools; the educational environment must provide the conditions for the student to recover their mental and emotional resources. This includes clear policies that reduce anxiety about rules, assessment methods that value the thinking process over the final product, and a culture where students are encouraged to question and verify what the technology tells them.
The author proposes that this new way of looking at learning offers a solution to the confusion surrounding artificial intelligence in education. It moves the conversation beyond simple questions of whether the technology is good or bad. Instead, it asks whether the specific way a student uses the tool helps them think, feel, and act as a responsible learner. The research suggests that when artificial intelligence supports a student's need for autonomy, competence, and connection, it fuels their learning. When it replaces their choices or isolates them, it drains them. The paper does not claim to have solved the problem of how to teach with artificial intelligence, but it provides a vocabulary and a map for understanding the hidden costs and benefits of the technology. It invites educators to design learning experiences that do not just produce faster results, but that preserve and renew the human capacity to learn, think, and flourish in a world increasingly shaped by machines.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.