Media Debate and AI in Education: Why Learning Theory Matters
Authors: Yerin Seung; info@ciddl.org
As generative AI tools (e.g., ChatGPT, Gemini, Copilot) continue to gain attention in education, educators and researchers face a key question: Should we embrace these tools as the future of teaching and learning, or proceed with caution given their limitations? AI’s potential certainly brings excitement with its capabilities for personalized feedback, built-in scaffolding, and real-time support. But behind the buzz lies a long-standing question in education: Is it the tool that makes a difference, or how we use it?
Looking Back: What the Media Debate Can Teach Us About AI
To understand this discussion, it helps to revisit a classic debate from the 1980s and 1990s about whether media influence learning outcomes. In a widely cited article, Clark (1983) argued that media are simply delivery vehicles. They don’t directly cause learning, any more than a truck causes groceries to be nutritious. In his view, the improvements in learning seen in media studies were more a result of the teaching methods used or the novelty of the technology, rather than the medium itself. Others, like Kozma (1991), disagreed. He argued that different media shape how students think and learn by offering unique ways to interact with content. For example, video might support visual thinking, while simulations allow students to explore cause and effect. In this view, media do matter, but mainly when paired with thoughtful instructional design. Later, Tennyson (1994) reminded us not to fall for what he called the “big wrench” mindset, which is the idea that any new technology can fix all of education’s problems. The lesson is that it is not about picking sides, but about understanding the relationship between tools, teaching methods, and learning goals.
Is AI Today’s “Big Wrench”?
Fast forward to today, and AI is being applauded as a game-changer in education. But we’re starting to see some of the same issues resurface. A recent critique by Weidlich et al. (2025) notes that many studies fail to separate the impact of the AI tool from how it is implemented in teaching. For instance, if students who use generative AI perform better, is it due to the tool itself, or because teachers used it thoughtfully within an effective learning design? Even more critically: Does the improved performance indicate real learning, or is it simply AI-generated output? Without clarity on what students were asked to do, the kind of support they received, and how their learning was assessed, we risk confusing correlation with causation. In other words, we might wrongly assume that AI caused the improvement when, in fact, the real driver was the instructional approach.
Why Learning Theory Still Matters
This is where learning theory becomes essential. No tool, whether a book, video, or chatbot, teaches on its own. What makes the difference is how it’s integrated into instruction, which reflects how students learn. Effective use of AI should be guided by established learning theories, such as cognitive load theory, self-regulated learning, and sociocultural approaches. Simply using ChatGPT to summarize text is a missed opportunity. Instead, we should ask:
- How can this tool deepen student thinking?
- How can it support metacognition?
- How can it provide timely, meaningful feedback?
- How can it help us better differentiate for students with learner variabilities?
The goal isn’t to replace educators, but to support them in fostering deeper learning.
Next Steps: Designing with Purpose
If we want to use AI in ways that truly support student learning, here’s what we need to do:
- Start with the Learning, Not the Tool: Identify your instructional goals and your students’ needs first. Then consider how AI might support those goals, not the other way around.
- Use Theory to Guide Practice: Draw on research-based learning theories to inform your decisions about integrating AI into your teaching.
- Look for Sound Evidence: Be cautious of studies that don’t separate the effects of the tool from the teaching method. Look for research that evaluates real, meaningful learning.
- Avoid One-Size-Fits-All Thinking: AI isn’t a panacea that fixes all educational problems. It should be one tool among many, thoughtfully embedded in a broader instructional approach.
We’re at a pivotal moment in education. If we want AI to benefit all learners, we must stay focused on what matters most: good teaching and meaningful learning. Generative AI should not merely be the source of learning but a thoughtful partner in a well-designed educational journey.
Join the conversation in our community!
CIDDL is committed to providing high-quality resources to support the increasing knowledge, adoption, and use of a range of educational technologies that can be used for educators, related services, or leadership preparation programs. For more resources, including videos and blogs, subscribe to our newsletter and follow us on YouTube, Facebook, and LinkedIn. The most important part of our CIDDL community is YOU. Join our community and share the innovative ways you are using technology, ask a question about technology integration, or participate in our bi-weekly live AI Community Chats. We look forward to seeing you in our community!