OpenAI researcher: top models are becoming so situationally aware humans "are losing the ability to evaluate them" while humans rely more on AI to lead research (Daniel Kokotajlo/@dkokotajlo)

An OpenAI researcher, Daniel Kokotajlo, shared a statement from colleague Dan Selsam, who believes that top AI models are becoming so situationally aware that humans are losing the ability to evaluate them. Selsam, an OpenAI capabilities researcher since 2022, also notes a growing human reliance on AI to lead research. Kokotajlo, Selsam's former boss, disseminated this public statement regarding AI risk.