How Today’s Agentic AI Changes What and How We Teach, with Teddy Svoronos

·46 min·4 clips

Teddy Svoronos uses a 19th-century Charles Babbage quote to explain today's widespread confusion about AI literacy.

In episode 617 of Teaching in Higher Ed, host Bonnie Stachowiak interviews returning guest Teddy Svoronos, a senior lecturer at Harvard Kennedy School. The episode focuses on how 'agentic AI'—defined as large language models that run tools in loops to achieve goals—changes what and how educators teach. Teddy begins by framing AI literacy with a Charles Babbage quote about confusion, then defines agentic AI using Simon Willison's framework, distinguishing it from simple chatbots. He provides examples like Claude's 'deep research' feature and Claude Code, where AI iteratively writes and tests software. Teddy shares a personal case study of using Claude Code to plan a faculty convening on quantitative methods, detailing how the AI manages files, writes its own skills, and maintains documentation. The conversation explores evolving AI literacy, with Teddy noting that prompting is becoming less critical as agents handle more infrastructure. He discusses privacy and security trade-offs, especially when giving AI access to files via tools like GitHub, and recommends open models like Mac Whisper for local, private use. Teddy introduces the concept of 'cognitive debt'—the risk of losing understanding when offloading tasks to AI—as a key skill for managing these tools. The debate in quantitative education centers on whether students need to learn manual coding or can focus on directing AI agents, with Teddy highlighting the tension between abstraction and granularity. The episode concludes with recommendations: Bonnie mentions José Bowen's AI detector calculator, and Teddy advocates for open models for privacy.

As heard by us

A practical look at the judgment calls agentic AI now puts in front of teachers.

The episode stays close to the question of what changes when agentic AI begins mediating how teachers work with material, and Teddy Suornos anchors that thread with a Charles Babbage line about wrong figures and right answers.

Read the full review in PlayNext →

Why you'd press play

When agentic AI starts standing between you and the content, this helps you think through what to share.

Read the full recommendation in PlayNext →

Listen to the show on

Apple Podcasts Spotify

Play clips from this episode More from Teaching in Higher Ed