'Show Your Work' Then AI Called My Bluff: Rethinking What Our Assessments Actually Measure

If AI can produce what we've been asking for, what were we actually asking for? I spent years telling students to show their work and it wasn't until AI that I let myself question an educational tradition of scoring human intellect that never felt completely right.

For over a century, education has asked students to produce things essays, tests, rubric-scored papers and called it evidence of learning. It wasn't a bad idea. It was a practical one. But somewhere along the way, the final submission became the point, and the thinking and learning behind it became invisible.

This presentation isn't about AI detection, policy, or whether students are cheating. It's about an older, quieter problem: assessment built around collecting output and predetermined outcomes, not witnessing genuine learning the confusion, the questions a student doesn't know how to ask yet, the moment something finally clicks. As educators, we work within systems that need the cleaned-up product at the end the scores, the outcomes, the things that can be counted. But now we have an opening to design assessment around what we actually care about: the learning itself.

Dr. Rachel Tatro-Duarte is a Professor of English at Porterville College and an AI Fellow for the California Community College Chancellor's Office. Her EdD research developed a VR Learning Model for experiential engagement in literature work that began with her immersive field studies in Greece as a Graduate Scholar Awardee and in Italy as a National Endowment for the Humanities scholar. That same pursuit of authentic encounter over surface performance now drives her assessment design work, which centers the visible process of thinking over polished output.

Webinar Registration - Zoom