AI regulation has become a hot topic as AI continues to rapidly develop. With the potential threat of existential risk, government, tech leaders, and concerned citizens are all calling for immediate action. OpenAI, a California-based research technology organization, has put forward a few ideas to start regulating the development of AI.
OpenAI CEO Sam Altman, Co-founder Greg Brockman, and Co-founder and Chief Scientist Ilya Sutskever addressed this issue in a post titled “Governance of Superintelligence.” In the post, the organization suggests that a good starting point for AI regulation is to form a coordinating entity to monitor the safety and integration of AI technology. This entity could be made up of governments worldwide or a separate organization dedicated to AI development that would limit the growth of AI capabilities.
The post also highlights the importance of establishing an AI-related branch of an International Atomic Energy Agency. This international agency would be responsible for ensuring safe and responsible AI development, and could perform activities like inspections, audits, and enforcing safety standards.
Additionally, OpenAI mentioned the need for a technical capability to safely make superintelligence. This is an open research question, being explored by multiple researchers and organizations. To ensure a smooth transition, companies should begin implementing regulations that an international agency would have, and countries could start introducing these regulations as a first step.
OpenAI CEO Sam Altman testified at a Senate Judiciary Committee hearing last week to further discuss the risks and future of AI. He spoke about current regulations, as well as why immediate action is necessary to regulate AI growth.
OpenAI is a research technology organization based in San Francisco, California whose mission is to ensure that artificial general intelligence benefits the greater good. The organization researches, develops and publishes the most successful technologies in AI, offering open-source models and datasets to the public. They are pioneering the development of new methods to ensure that their technology is safe and beneficial for all of society.
