# Computational hermeneutics: evaluating generative AI as a cultural technology

## TL;DR

This paper introduces computational hermeneutics as a new way to evaluate generative AI by focusing on meaning and context rather than just accuracy.

## Contribution

The paper proposes computational hermeneutics as a novel framework for evaluating generative AI systems through interpretive cultural challenges.

## Key findings

- GenAI systems function as 'context machines' that inherently address interpretive challenges like situatedness, plurality, and ambiguity.
- The paper introduces three principles for hermeneutic evaluation: iterative benchmarks, inclusion of people, and measuring cultural context.
- Computational hermeneutics shifts AI evaluation from standardized accuracy to contextual meaning.

## Abstract

Generative AI (GenAI) systems are increasingly recognized as cultural technologies, yet current evaluation frameworks often treat culture as a variable to be measured rather than fundamental to the system's operation. Drawing on hermeneutic theory from the humanities, we argue that GenAI systems function as "context machines" that must inherently address three interpretive challenges: situatedness (meaning only emerges in context), plurality (multiple valid interpretations coexist), and ambiguity (interpretations naturally conflict). We present computational hermeneutics as an emerging framework offering an interpretive account of what GenAI systems do, and how they might do it better. We offer three principles for hermeneutic evaluation—that benchmarks should be iterative, not one-off; include people, not just machines; and measure cultural context, not just model output. This perspective offers a nascent paradigm for designing and evaluating contemporary AI systems: shifting from standardized questions about accuracy to contextual ones about meaning.