Key Takeaways:
- In his latest blog, faculty member Nathan Pritts explores AI and how learning comes from engaging with ideas, struggling through challenges, and developing understanding, not just getting answers from AI.
- AI can reduce routine mental work and free learners to focus on higher-level thinking, but excessive reliance on AI may short-circuit the cognitive processes necessary for learning.
- The value of AI is determined less by the tool itself and more by how and when it is used within the learning process.
- When learners allow AI to do too much of the work, their personal engagement and ownership of learning can weaken.
- Learners should question AI outputs, revise them, and integrate them into their own understanding rather than accepting responses as final products.
- AI can improve assessment performance, but outcomes depend heavily on prior knowledge, self-regulation, and learning conditions.
- AI is already part of education, but meaningful learning only occurs when individuals remain actively engaged in interpretation, synthesis, and judgment.
--
How to Use AI Without Letting It Do the Thinking for You
A lot of the conversations around artificial intelligence (AI) and learning begin with questions about capability. What can AI do? Can it tutor students? Can it explain complex ideas accurately? Can it improve performance on assessments? Increasingly, the answer to all three is yes.
But I don’t think these are the questions we need to be asking. They direct our attention toward what AI can produce, rather than toward what happens cognitively when learners engage with those outputs. Before asking if AI can do something, we need to understand whether we should be using it for particular tasks.
Learning has never been about getting the right answers, or even getting things done on deadline. It depends on the processes through which individuals encounter, struggle with, and ultimately make sense of new information, skills, and competencies. When AI enters that process, it can offer support, but it can also offer shortcuts.
Most of the time, we don’t even notice it happening.
The Complexities of AI
One of the dynamics of AI use is cognitive offloading, the process by which tools take over certain mental tasks. When AI takes over routine or extraneous work, it can free learners to focus on interpretation, synthesis, and evaluation. When it replaces the very processes through which understanding develops, though, it can short-circuit learning.
Research on human interaction with advanced AI systems has emphasized this uneven dynamic, sometimes referred to as a “jagged technological frontier,” in which tools perform certain tasks with remarkable fluency while failing unpredictably in others. The challenge, then, is not simply to use AI effectively but to recognize where its strengths and limitations intersect with the learner’s own developmental needs. Sometimes that means asking AI to show you a few different ways something could be organized and then doing the harder work of deciding what actually makes sense.
Research on AI-assisted learning reflects this complexity. In controlled studies, students who had access to AI support often performed better on assessments than those who did not. However, those gains depended heavily on factors such as prior knowledge, self-regulation, and the conditions under which the AI was used.
The implication is that AI use changes how the thinking happens. And whether that supports learning depends on how deliberately we design for it. In other words, the tool shouldn’t be steering the process. We need to take responsibility for that.
How Learners Can Navigate AI
What really moves the needle is how learners are guided to use these tools and at what points in the learning process they are introduced. The difference between productive support and cognitive disengagement often lies in timing and intention rather than in the presence or absence of the tool itself. It’s how we use AI that matters.
The challenge is not simply that AI can produce answers. It is that it can do so in ways that alter the learner’s relationship to effort, responsibility, and meaning-making. When that relationship changes, learning changes with it. We’ve seen this most clearly with writing, where AI reveals a breakdown in how we assess thinking and learning.
But learning with AI does not operate in binary terms, and blanket rules tend to obscure the very nuances that determine whether learning is supported or diminished. As higher education research increasingly emphasizes, effective integration of AI requires attention to discipline-specific practices, intentional pedagogical design, and the preservation of human-centered forms of engagement rather than universal mandates.
"The question isn’t whether AI becomes part of the process. It already has. The question is whether we’re willing to stay with the work long enough to make it our own."
A more productive approach begins with a different kind of question. Instead of asking whether AI can be used, we might ask when it should be used and for what purpose within a given process. In some moments, speed and structure may be beneficial, allowing learners to organize ideas or explore alternative perspectives. In others, particularly when understanding is still forming, the same efficiencies may undermine the very cognitive work that needs to occur.
This is why many people report a sense of discomfort when using AI, even when the results are useful. There is often a moment of uncertainty, an awareness that something about the process feels incomplete or misaligned. It’s a small moment, but it’s hard to ignore once you notice it. It reflects an intuitive recognition that producing an answer is not the same as developing understanding. When AI displaces too much of the thinking process, the connection between the learner and the work begins to weaken.
Moving Forward With AI
It may seem obvious, but in order to learn, you need to think. And you can do this with AI by sticking with your own insights and ideas through the interaction with the tool; questioning outputs; revising them; and integrating them into one’s own developing understanding rather than accepting them as finished products. It just requires additional, intentional work that preserves the need for interpretation, synthesis, and judgment.
The question isn’t whether AI becomes part of the process. It already has. The question is whether we’re willing to stay with the work long enough to make it our own.
Want to learn more about using AI? Check out my other blog on what we’ve misunderstood about using AI for writing. Then, if you’re ready to see how you can correctly apply these tools to your learning experience, talk to an advisor about our degree programs today!
