About the Episode
About the Episode:
Join in as hosts Timothy and Jamie dive into the evolving role of AI in education, specifically exploring how large language models like ChatGPT are shaping learning environments. The episode centers on two fascinating but conflicting studies—one from Harvard and another from the University of Pennsylvania—about the impact of AI tutors on student performance.
Key Takeaways
- AI tutors, like ChatGPT, can improve students' performance on practice problems and multiple-choice tests.
- There are inconsistencies in learning outcomes depending on the type of study, age group, and context of AI use.
- Active learning and human interaction still play a vital role in deep understanding, which cannot be fully replicated by AI.
- The long-term role of AI in education should focus on balancing human instruction with AI assistance, not replacing traditional methods.
Episode Summary
How does AI impact learning in the classroom?
This episode tackles the complex question of how AI, particularly large language models (LLMs) like ChatGPT, can be used in educational settings. Timothy and Jamie review two studies—one from Harvard and the other from the University of Pennsylvania—that offer conflicting insights into the efficacy of AI tutors. While both studies indicate that AI can improve performance on practice work, the real question is whether these gains translate into meaningful learning outcomes.
The Harvard study, which involved 194 students in a physical sciences course, found that students who used an AI tutor (TeachGPT) outperformed those in a traditional active learning environment. The AI tutor boosted test scores significantly, from a baseline of 2.8 to over 4.3. The University of Pennsylvania study, which involved 839 Turkish high school students in a math course, had more nuanced results. While students with AI tutors performed better on practice problems, they did no better than traditionally taught students when it came to the final exam. These contrasting findings led to a discussion on the need for more research and consistency in how AI is applied in education.
Is there a downside to using AI in education?
Both hosts raise the issue of over-reliance on AI in education. Jamie emphasizes that while AI can help with immediate problem-solving, it might not foster the deep learning necessary for critical thinking and long-term retention. This brings up the classic "calculator dilemma"—just as calculators changed how students approached math, AI tools like GPT could lead to students bypassing the mental work necessary for understanding fundamental concepts.
Another important factor is accuracy. Timothy explains that while LLMs can generate helpful responses, they are prone to "hallucinating" or creating false information when left unchecked. This is why ensuring that AI tutors use a reliable source of information (such as retrieval-augmented generation models) is critical. Additionally, they point out that the social component of learning—engaging with peers and professors in real time—is irreplaceable.
How can AI and human instruction work together?
The conversation ends with a discussion on balancing AI tools with traditional learning environments. While AI tutors can provide individualized support and improve performance on specific tasks, human interaction fosters collaboration, deeper understanding, and the development of soft skills. More studies need to be done to fully understand the most effective ways to integrate AI into the classroom without diminishing the value of human interaction and active learning.
Ultimately, the conclusion is that AI can supplement education but not replace the rich, social learning environment provided by lectures and peer interaction. While AI tutors may boost test performance, they do not offer the same holistic educational experience that comes from engaging with a community of learners.
Links to articles:
Harvard
Connect With Our Co-Hosts:
About The Enrollify Podcast Network: The EduData Podcast is a part of the Enrollify Podcast Network. If you like this podcast, chances are you’ll like other Enrollify shows too!
Some of our favorites include Generation AI and The Higher Ed Geek.
Enrollify is produced by Element451 — the next-generation AI student engagement platform helping institutions create meaningful and personalized interactions with students. Learn more at element451.com.
Element451 is hosting the AI Engage Summit on Oct 29 and 30. Register now for this free, virtual event.The future of higher ed is being redefined by the transformative power of AI. The AI Engage Summit brings together higher ed leaders, innovators, and many of your favorite Enrollify creators to explore AI’s impact on student engagement, enrollment marketing, and institutional success.
Experience firsthand how AI is improving content personalization at scale, impacting strategic decision-making, and intuitively automating the mundane tasks that consume our time. The schedule is packed with real examples and case studies, so you leave knowing how to harness AI to drive meaningful change at your institution.
Whether you’re looking to enhance student outcomes, optimize enrollment marketing, or simply stay ahead of the curve, the AI Engage Summit is your gateway to the next level of higher education innovation. Registration is free, save your spot today.
.svg)
.svg)
