Label Over Logic? How Source Cues Bias Human Fallacy Judgments More Than LLMs

Name: Label Over Logic? How Source Cues Bias Human Fallacy Judgments More Than LLMs
Uploaded: 2026-09-07T19:34:52.238725+00:00
Duration: 423 s
Description: The episode discusses the paper "Label Over Logic? How Source Cues Bias Human Fallacy Judgments More Than LLMs." It examines how humans are highly susceptible to source cues when judging logical accuracy. The hosts conclude that while LLMs are stable, human and AI biases are complementary, suggesting a collaborative model is needed to mitigate errors in AI-assisted content.

The episode discusses the paper "Label Over Logic? How Source Cues Bias Human Fallacy Judgments More Than LLMs." It examines how humans are highly susceptible to source cues when judging logical accuracy. The hosts conclude that while LLMs are stable, human and AI biases are complementary, suggesting a collaborative model is needed to mitigate errors in AI-assisted content.

Open the full episode