Best of: The future of educational technology

Episode Summary

Executive Summary: Dan Schwartz argues AI in education should augment teachers, not replace them. The conversation explores how generative AI changes teaching, learning, assessment, and edtech business models, while stressing that the best uses are those that deepen understanding, motivation, and creativity rather than simply automate instruction or grading.

Main Topics: AI as a catalyst for educational change (Priority: 5/5): Schwartz says ChatGPT has made people realize education will change, triggering questions about human roles, jobs, and school design. Teachers vs. technology (Priority: 5/5): Altman voices skepticism about tech standing between students and teachers; Schwartz agrees teachers remain central and computers should support, not substitute, human instruction. Motivation and learning through computer interaction (Priority: 4/5): Examples include students teaching an intelligent agent and the protége effect, where learners work harder when helping another entity learn. AI, cheating, and assignment design (Priority: 5/5): The discussion focuses on whether students using ChatGPT is cheating or a new reality, and how educators should redesign tasks to measure novelty, judgment, and conceptual understanding. Assessment and grading in the AI era (Priority: 5/5): Schwartz argues assessment should increasingly support instruction in real time, while also evaluating strategic choices, creativity, and conceptual reasoning rather than just right answers. Edtech industry incentives and risks (Priority: 4/5): He worries companies optimize for efficiency and local school needs instead of deeper learning, even as AI creates opportunities for better educational tools. Embodiment and physical learning (Priority: 4/5): Schwartz connects learning to physical and perceptual experience, arguing AI should eventually integrate with virtual/embodied environments to better support complex understanding.

Key Arguments: AI makes the need to rethink education unavoidable because it exposes how much of schooling depends on routine tasks that machines can now do. Teachers are still essential because students learn not just from content delivery but from observing how teachers think and model reasoning. Computers can add value when they create motivating, social, or immersive learning experiences, such as teaching an AI agent or using virtual environments. The most effective AI uses in education will likely be those that help teachers manage projects, personalize support, and design richer learning experiences. ChatGPT is not simply a cheating tool; its impact depends on assignment design and whether students are asked to show prompts, drafts, and reasoning. Assessment should shift from only evaluating correct outputs to evaluating choices, strategy, creativity, and conceptual understanding. Educational institutions and companies must be careful not to optimize for efficiency alone, because speed and automation can crowd out deeper learning. Embodied and physical interaction matter for learning; future AI systems should better incorporate action, perception, and virtual manipulation. Many students and teachers are already using GenAI, so schools must adapt policies and pedagogy rather than pretend the technology is absent.

Data Points: K-12 teachers using GenAI: about 60% - Schwartz cites survey averages for pre-K through 12 teachers. K-12 students using GenAI: about 30% - Schwartz cites survey averages for K-12 students. College faculty using GenAI in teaching: about 30% - Schwartz cites survey averages for higher education faculty. College students using GenAI: about 80% - Schwartz cites survey averages for college students. Edtech unicorn valuation growth: $54 billion valuation companies last year - Schwartz notes the rise of large edtech companies compared with zero unicorns 20 years ago. Time savings from ChatGPT use in the class example: about half the time - Altman reports students using ChatGPT finished the assignment in roughly half the time. Grade difference with ChatGPT use: about 2 points out of 100 - Altman says users scored slightly lower, roughly 90 to 88. Number of students in Schwartz's high school teaching example: 130 kids - He describes managing project-based learning for a large class. Teaching the AI agent duration before obsolescence: a couple of days - Students learned the reasoning quickly enough that the tool was no longer needed. Response time in the statistics example: five minutes - Schwartz asks students to quickly answer a conceptual question via email.

Pivotal Quotes: "When I'm in a classroom, and this has been since I've been in third grade, I am watching the teacher trying to understand how they think about the information and how they struggle with it to understand it and then try to relay it to me." — Russ Altman: He explains why he values human teachers over technology as a mediator. "We allow ChatGPT because my fellow instructor and I decided, and this was the quote, we want to be part of the future, not part of the past." — Russ Altman: He describes allowing AI use in his ethics writing class. "The concern is the computer does all the work." — Dan Schwartz: He frames the key educational risk of technology that reduces student thinking.

Implications: Schools should treat AI as a redesign opportunity: preserve teacher-centered learning, update assessment, and use AI for motivation, personalization, and embodied learning rather than simple automation.