Motivation by Design: Self-Determination Theory and the Problem of AI Tutors

When a student sits down with an AI tutor, the immediate question is usually about accuracy. Will the system explain the quadratic formula correctly? Will it catch the misplaced comma in a paragraph? These are engineering problems, and they matter. But once the math is right and the grammar is fixed, a deeper question emerges: does the student actually want to keep working? Motivation is not a bonus feature in education. It is the engine. Without it, even the most sophisticated tutoring system becomes a very expensive paperweight.

For decades, psychologists have tried to map what keeps people engaged in difficult tasks. One of the most durable frameworks to emerge from that work is self-determination theory. Developed by Edward Deci and Richard Ryan, the theory argues that human motivation is not just about rewards and punishments. Instead, people are driven by three basic psychological needs: autonomy, competence, and relatedness. When these needs are met, people engage with tasks willingly and deeply. When they are thwarted, motivation collapses. As outlined in their foundational paper, "Self-determination theory: A macrotheory of human motivation, development, and health," these three needs are essential for healthy development and sustained effort (Deci & Ryan, 2008).

Applying this framework to artificial intelligence in the classroom is no longer a theoretical exercise. A recent synthesis of research trends in the field, "Artificial Intelligence in Education (AIED): Publication Patterns, Keywords, and Research Focuses," highlights a growing emphasis on integrating motivational and psychological frameworks into AI-driven educational tools (Chen et al., 2023). Researchers are recognizing that building a smart algorithm is only half the job. The other half is ensuring the algorithm interacts with students in a way that respects how human motivation actually works.

The Need for Autonomy in a Directed System

Autonomy, in the context of self-determination theory, does not mean doing whatever you want. It means feeling a sense of volition—that your actions align with your own interests and values. This presents an immediate tension for AI tutors. By definition, a tutoring system directs a student toward a specific learning goal. It sequences problems, selects readings, and dictates the pace. If an AI agent controls every step, the student becomes a passenger rather than a driver.

Teachers face this problem daily, but an AI agent can make it worse if designed poorly. A system that forces a student through a rigid sequence of multiple-choice questions without offering any choices strips away autonomy. To support the need for autonomy, an AI assistant might offer meaningful options. A student could choose which historical figure to write about, select between two different methods for solving a physics problem, or decide whether to review a concept before attempting a quiz. These choices do not change the learning objective, but they change the student’s relationship to it. The AI acts less like a drill sergeant and more like a guide offering paths up the same mountain.

Competence and the Danger of Getting Stuck

The second basic need is competence—the desire to feel effective and capable when interacting with the environment. In a classroom, competence is built through optimal challenge. Tasks that are too easy breed boredom; tasks that are too hard breed anxiety. Human teachers constantly read the room, adjusting their explanations based on confused faces or restless energy. An AI agent must replicate this sensitivity using data.

If an AI tutor repeatedly serves problems that a student cannot solve, it actively damages the student's sense of competence. Conversely, if the system lowers the bar so much that success requires no effort, the student learns nothing and feels no genuine accomplishment. The design challenge is calibration. The AI must track performance closely enough to keep the student in a space where effort leads to progress. When a student makes an error, the system’s response matters immensely. A blunt notification that an answer is wrong provides no path forward. A well-designed AI agent breaks the problem down, identifies the exact point of failure, and offers a targeted hint, allowing the student to experience the satisfaction of figuring it out themselves. That experience of overcoming a hurdle is what builds competence.

Relatedness and the Limits of the Machine

The third need, relatedness, is perhaps the most complicated for artificial intelligence. Relatedness is the need to feel connected to others, to care and be cared for. Learning is inherently social. Students work harder for teachers they respect and alongside peers they like. An AI agent is not a person. It does not care about the student, and pretending that it does—through overly enthusiastic chatbot personas or simulated empathy—often feels hollow to learners.

However, an AI agent does not have to replace human connection to support it. Instead, it can coordinate it. When an AI system assists a teacher, it can free the teacher from grading routine assignments or delivering basic lectures. That reclaimed time allows the teacher to sit with a struggling student, ask how their weekend was, and provide the genuine human attention that fulfills the need for relatedness. Furthermore, AI agents can facilitate peer collaboration by grouping students with complementary skills or coordinating shared projects, turning the technology into a bridge between people rather than a wall isolating the student behind a screen.

Designing for Motivation

The integration of self-determination theory into educational technology is not about making software friendlier. It is about recognizing that cognition and motivation are inseparable. A student who feels controlled, incompetent, or isolated will not learn effectively, regardless of how accurate the AI’s database is. The research trends identified in the AIED literature review suggest that the field is moving past the novelty of machine learning and toward the harder, more necessary work of psychological alignment.

For teachers and designers, self-determination theory offers a concrete checklist. Before deploying an AI tool in a classroom, ask three questions. Does this system give students meaningful choices? Does it calibrate difficulty so students can experience genuine success? Does it enhance human connection rather than replacing it? If the answer to any of these is no, the technology may be delivering information, but it is failing to educate. Motivation is not magic. It is a structure built from autonomy, competence, and relatedness. Any AI agent entering a classroom must be built to support that structure, or it will inevitably tear it down.

Sources

• Self-determination theory: A macrotheory of human motivation, development, and health
• Artificial Intelligence in Education (AIED): Publication Patterns, Keywords, and Research Focuses