Summary
A member of OpenAI's non-profit board, Paul Christiano, has warned that OpenAI is not on track to reduce the risk of "catastrophic" loss of control to an acceptable level, citing a meaningful risk of rapid AI acceleration leading to irreversible loss of control.