Mustafa Suleyman and Walter Isaacson discuss whether AI systems should be treated as conscious and how that question affects safety. Suleyman argues against attributing subjective experienceMachine consciousness is the possibility that an artificial system could have subjective experience, rather than merely produce behavior associated with it. to current systems, a position he contrasts with views of uncertainty about AI welfare.
Suleyman calls for containmentModel containment limits the systems, data, tools, and real-world effects an AI model can reach during testing or use., transparency and public involvement in decisionsGovernance defines responsibilities, decision rights and controls for how a system or organization is operated. about powerful systems. He discusses an AI research incident as a warning while acknowledging that the test removed safeguards; his account and predictions are presented as his views.
The conversation also covers AI companionship, open models, scientific applications and employment. Suleyman expects significant displacement in some jobs and argues that society should decide which human roles and values to protect.
Watch on YouTube
