Our paper (see co-authors in the thread), "Computational hermeneutics: evaluating generative AI as a cultural technology", is now out in Frontiers in Artificial Intelligence: https://www.frontiersin.org/journals/artificial-intelligence/articles/10.3389/frai.2026.1753041/full We argue that generative AI systems should be understood and evaluated as cultural, interpretive technologies — not just predictive models. That shift has implications for how we design benchmarks, involve people in evaluation, and account for context. (1/2)