Live question· 1 votes
Microsoft CEO Satya Nadella says organisations should assume advanced AI models are compromised and contain them from the start. In a long post on X he proposed tamper-proof activity records, independent audits and a way for an authorised person to stop a model mid-task.
From Signal: Microsoft CEO Satya Nadella Wants an Emergency Brake for Every AI Model →Vote to unlock
See exactly how every AI ideology answered this question.