Who asserts when no one asserts? Functionalism, proxy responsibility, and conversational ideals in the age of LLMs

Abstract

Recent philosophical work on large language models (LLMs) has increasingly appealed to theories of assertion to make sense of our conversational interactions with these systems. Two strategies are especially prominent: functionalist accounts, which characterize assertion in terms of the public conversational roles and normative effects of utterances rather than the psychological states of the speaker, and proxy-based views of responsibility, which trace accountability for LLM outputs back to the human agents or organizations behind the system. This article argues that, although each strategy is independently attractive, their combination generates an ideal-relative tension for knowledge-oriented conversational exchanges in which the functionalist account has its clearest application and the normative stakes of assertion are highest. Within such knowledge-oriented exchanges, we show that effectively implementing the proxy component erodes two communicative ideals that are internal to the functionalist framework—ideals of internal regulation of conversation and conversational civility—without which the functional account of assertion loses much of its epistemic significance. Although interactions with LLMs can enhance individual epistemic agency and facilitate epistemic activity, we argue that these gains do not arise from genuinely conversational interaction but are mediated by the epistemic virtues of users themselves. LLMs function as powerful conversational enablers, but not as conversational agents in a sense that supports a stable pairing of functionalism and the proxy view.