# The sophist in the server: Rhetoric, Reasoning and Scientific Judgment in the Age of LLMs

## TL;DR

This paper discusses how large language models can convincingly argue both sides of an issue, which may blur the line between persuasive rhetoric and scientific evidence.

## Contribution

The paper introduces a critical analysis of how LLMs impact scientific judgment and proposes ways to maintain rigorous standards in research.

## Key findings

- LLMs can generate equally persuasive arguments for conflicting hypotheses.
- This capability risks confusing rhetorical skill with scientific validity.
- Scientists must balance LLM fluency with epistemic discipline.

## Abstract

Large language models can defend a hypothesis and its opposite with equal eloquence, raising a new risk for scientific reasoning: confusing rhetorical plausibility with evidential support. This essay analyzes how LLMs reshape creativity, judgment and responsibility in research—and how scientists can engage their fluency without surrendering epistemic discipline.