US Legislators Propose Moratorium on Artificial Superintelligence Development

Senator Bernie Sanders and Representative Greg Casar have introduced the 'Ban Artificial Superintelligence Act'. This proposed legislation aims to impose a moratorium on the development of Artificial Superintelligence (ASI). The bill defines ASI as models exceeding human cognitive abilities across all domains. Its proponents assert that such systems pose unquantifiable risks. The Act mandates a halt to ASI development until comprehensive regulatory frameworks are established.

The proposed legislation would prohibit the development of AI models demonstrating capabilities beyond human performance. These capabilities include reasoning, learning, and creativity. Furthermore, the bill specifies a threshold for compute power. It would ban models trained using more than 10^26 floating-point operations (FLOPs). This threshold is intended to prevent the creation of systems with unmanageable complexity and potential for autonomous action.

Key provisions of the Act also include requiring a pre-launch risk assessment for all advanced AI models. Developers would need to demonstrate safety and alignment before deployment. The legislation also proposes establishing a new federal agency. This agency would be tasked with overseeing AI development and enforcing the moratorium. It would also be responsible for licensing AI research and development activities.

Public discourse on ASI has intensified, with some experts advocating for preemptive regulation. The 'Ban Artificial Superintelligence Act' reflects a growing concern among policymakers regarding uncontrolled AI advancement. The bill's introduction initiates a legislative process. Its passage would significantly alter the trajectory of AI development in the United States. How might this proposal shape global AI governance conversations?