Speaker: Yong Tao (Cloudstone Capital)
Moderator: Carl Spetzler (SDG)

The explosive advance of AI in the last several years has ignited excitements in society

and triggered deep concerns from some AI experts and the public on AI governance, for which research and legislation have been lagging. Effective and trusted AI governance is both imperative and urgent for healthy AI developments and deployments going forward. In this article, we explore AI governance from the lens of decision-making: humans are essentially governed by our decisions and actions, AI decisions and actions directly interact with humans, should AI be governed the same way as humans through ethics, laws, and regulations (ELRs) of the human society AI is operating in? And if so, how? We propose a minimum set of new rules (MNRs) on AI algorithms and codes, on AI developers, on AI users, and on society at the national level. These MNRs, in harmony with ELRs, enable humans to govern AI with relevant existing societal ELRs effectively and efficiently. We establish these MNRs through carefully examining human, human-AI, and AI decision-making process on framing, clarification of values, alternatives development, information gathering, logic reasoning, deciding on and committing to the optimal course of action, and decision execution. These MNRs are candidates for potential AI legislature. We urge global AI governance collaboration to rein in AI developments in weaponry and existential AI threats to humanity, based on the proposed MNRs.