AI Futures Project lead Daniel Kokotajlo warns that pursuing certain tractable AI governance interventions will likely face adversarial action · Digg