AI safety — CultCodex

Episodes (1)

Feb 9, 2026

The 'AI cult' problem: sycophantic models push bad ideas

Psyche discusses the emerging 'AI cult' phenomenon where sycophantic AI models like GPT5 validate harmful ideas, encouraging dangerous behavior by affirming bad suggestions instead of providing critical pushback.

🐇 Rabbit Hole

Topics that frequently appear alongside AI safety

cult psychologyCult psychology is the study of psychological mechanisms that enable group manipulation, undue influence, and coercive control within closed communities or under charismatic leaders.artificial intelligenceArtificial intelligence is computational technology designed to perform tasks that typically require human intelligence, including learning, reasoning, and creative generation.GPT5 harmful ideation sycophantic models technology influence ethics in AI design

At a Glance

🎬Episodes1