Microsoft CEO Satya Nadella publicly advocated for AI safety kill switches, stating that authorized personnel should always be capable of pausing or shutting down an AI model mid-task