OpenAI, in a recent blog post, advocates regulation of superintelligence, similar to the regulation of nuclear power, and proposes the creation of a regulatory agency such as the International Atomic Energy Agency (IAEA). They emphasize the need for AI regulation, especially for superintelligence, and describe the potential impact of rapid AI development and the need for adequate governance.
OpenAI’s regulatory proposal
The initial statement in the blog post argues that artificial intelligence systems will surpass the skills of experts in most domains and perform as many productive activities as large companies today. For this reason, it is essential to mitigate the risks of an existential threat in the same way as the measures taken for nuclear power.
The paper’s authors, Sam Altman, Greg Brockman and Ilya Sutskever, noted that eventually there would be a need for some kind of International Atomic Energy Agency (IAEA). This agency would regulate superintelligence efforts and would require global cooperation and coordination.
In addition, the blog post emphasizes the importance of “enabling companies and open-source projects to develop models that exceed a significant capacity threshold, without the kind of regulation normally imposed.”
OpenAI considers democratic decisions
OpenAI has developed a strategy to encourage regulation of AI through collective decision-making, which Similar to Wikipedia’s method. Greg Brockman, the company’s chairman, has announced that OpenAI will evaluate methods to collect input from various parties when making decisions. The purpose of this is to create a democratic process for decision-making.
“We’re not alone in Silicon Valley thinking we can write these rules for everyone,” said Brockman. While there will undoubtedly be challenges in setting up such a system, Brockman shared the same view as Sam Altman regarding the importance of international cooperation in regulating AI.
Regulating AI
The issue of how AI should be regulated is a topic on everyone’s mind. Even those directly involved in AI development are calling for regulation. Altman has turned down the chance to become America’s top AI regulator. Instead, at a congressional hearing, he insisted that the US assume responsibility for regulation.
It seems that the United States will create a special agency to oversee AI. Other countries are following suit, with China in particular pushing for bans and restrictions on AI. Just as has happened in the past with the crypto market, the AI industry may face sudden and harsh regulation. These regulations will mainly focus on aspects such as privacy, preventing interference in elections and setting global standards.