Is AI Quietly Eroding the Magic of Teaching?

At the Krause Center for Innovation (KCI), we talk a lot about using technology with purpose. For us, the gold standard of education will always be empowering student agency and cultivating durable skills like critical thinking, collaboration, and intrinsic motivation.

But as we jump into Year 2 of our Silicon Valley Artificial Intelligence Project (SVAIP) initiative, a fascinating new study out of the University of Pennsylvania caught our eye, and honestly, it gives us a lot to think about.

Researchers ran a massive, randomized controlled trial across 14 middle and high schools in Turkey, looking at what happens when teachers are given custom AI  tools to help design their classes. 

You’d think more productivity for teachers would mean better outcomes for kids, right?

Well, the data told a surprisingly cautionary story:

• Motivation took a hit: Students in the AI-assisted courses reported a notable drop in intrinsic motivation. They rated their classes as less interesting, less important, and less enjoyable.
• The “cookie-cutter” effect: Usage logs showed that teachers mostly used AI for quick task delegation like generating slides, syllabi, and exams with a median of just two prompts. The result? Homogenized materials that lacked the teacher’s unique, signature voice. Students noticed the lack of personal touch.
• The equity trap: For classrooms with lower-performing teachers, relying on AI as a cognitive shortcut actually dragged down student exam scores and tanked their academic confidence.
• The “Awe vs. Awareness” curve: Interestingly, the teachers with the most prior experience with AI became the most pessimistic about it by the end of the semester. The initial “wow” factor wore off, and they became aware of the unintended side effects.

Let’s Talk: What’s Your Experience?

We are actively working on the ground to build the frameworks and guardrails that keep AI acting as an iterative design collaborator, not a low-effort substitute for human-centered pedagogy with our Silicon Valley Artificial Intelligence Project (SVAIP).

Following up on our first three SVAIP reports, we’re releasing a new research paper this fall that dives deep into these dynamics. But we want to pause and ask the educators in our community:

• Does this match what you’re seeing in your own schools? 

• Have you noticed a shift in student engagement when materials lean heavily on AI? 

• Or are you seeing something completely different, where AI is actually amplifying your unique voice and giving you more time to build deep relationships with your students?

Let’s get a conversation going in the comments. How are you keeping the human magic alive in the age of AI?

Dig into the data: Read the full University of Pennsylvania research paper here.

See our ongoing research: Check out our journey so far and read our three SVAIP initiative reports here.