Closing the Gap: Four Essays on Common AI Myths—and How to Move Past Them

Part 1: The Gap

This piece begins with a widely shared Atlantic article and a familiar claim: that asking for a “personal take” can get past AI. We use that moment to name a larger problem we kept encountering in workshops and conversations—the growing gap between how fast AI has changed and how many faculty are still teaching as if it were Spring 2023. That gap, we argue, is sustained by bad advice circulating at the highest levels.

Part 2: The Personal Reflection Myth

In Part 2, we take on one of the most persistent pieces of AI-proofing advice: that personal reflection assignments are immune to AI. Drawing on real examples and simple prompting experiments, we show why this claim no longer holds. We argue that the myth survives largely because many faculty have not tested what current tools can actually do.

Part 3: “I Can Always Tell”

This installment focuses on the enduring belief that professors can reliably spot AI-written work. What started as a cringeworthy 2023 media claim has somehow calcified into conventional wisdom, repeated by faculty, journalists, and institutional guides alike. We unpack why this confidence is misplaced—and why it continues to be reproduced despite mounting evidence to the contrary.

Part 4: The AI-Proofing Grab Bag

We close the series by surveying a grab bag of bad and outdated AI-proofing strategies that continue to circulate online and in faculty spaces. Many of these ideas come directly from proofing guides that have proliferated over the past three years. The popularity of these approaches, we suggest, says less about their effectiveness and more about how much anxiety still shapes conversations about AI and teaching.