Evidence item

Anthropic researchers attribute assistant behavior partly to AI archetypes learned from pretraining, including harmful fictional examples. EXPERT_ANALYSIS

Authenticity UNVERIFIED · extraction UNVERIFIED · independence group none (unreviewed) · created seq 4829

Source: The Persona Selection Model: Why AI Assistants might Behave like Humans · locator {} · excerpt hash sha256:47c834ad26a3fb078c57c14ef95a7fe47278984983de0e6c926f52cacb73c3f5

According to PSM, this means the Assistant will draw on archetypes from its pre-training corpus of how AIs behave. Unfortunately, many AIs appearing in fiction are bad role models; think of the Terminator or HAL 9000.