John Warner: Faculty Can Reclaim What Generative AI Drains

A small thing in this week's higher-ed reading stuck with me more than the usual product launch noise. In a piece for Inside Higher Ed, John Warner describes the mood inside the faculty workshops he runs, and it isn't curiosity. It's anger, sadness, and despair, with educators, in his words, 'thrown to the wolves' while the technology keeps escalating.

Warner contrasts that with COVID. That disruption felt temporary, something you could grit your teeth through. Generative AI feels permanent, and that changes the texture of the despair. The complaint isn't about a hard semester, it's about whether teaching still has a load-bearing role when students can outsource the assignment and professors can outsource the feedback.

Two anecdotes carry the piece. Crystal Fleming, a professor of Africana studies at Smith College, recounts a keynote speaker who admitted to generating his own presentations with AI and writing an entirely AI-generated book, and who described himself as 'useless as an intellectual.' Fleming's response is the line worth keeping: 'embodied knowing' is precisely what AI cannot do, and the human capacity to detect AI's work is itself a form of expertise. The other is Jeff Sharlet, the Dartmouth creative writing professor, who anonymously surveyed his students. He found a mood running 'from resignation to despair,' with feelings of betrayal 'from deep to furious.' Some students described AI dependencies they could not break. Many resented their institution's AI partnerships and professors using AI to grade them.

Warner's argument is that the way out is not enthusiasm but human agency. Rethink assessment. Try alternative grading approaches. Make student learning visible to learners themselves. Center learning as the institution's primary mission. His closing reframe is, 'Ourselves, or AI?'

The honest caveat is that this is one columnist reading a few workshops and one survey, not a population study; the resignation he describes may be sharper among writing faculty than across the disciplines. What the reporting does not give you is a price tag for what centering learning looks like at scale, or how an institution that has already signed an AI vendor contract unwinds it. But the direction is the part worth watching: Warner's claim is that the institutions choosing meaningful human learning will find themselves thriving over time, and that is at least a testable bet.