TechFlow news — Sam Altman, Greg Brockman, and Ilya Sutskever, founder, president, and chief scientist of AI company OpenAI, published a blog post titled "Governing Superintelligence" on their official website.
The article notes that while AI systems will surpass human expertise in many fields and bring a more prosperous future, we must manage the associated risks. To address these risks, they propose three solutions:
1. A certain level of coordination among developers to ensure the safety of superintelligence and its integration with society;
2. The establishment of an "IAEA for AI"—a dedicated international regulatory body;
3. Advancing technical capabilities to ensure superintelligence safety.
OpenAI executives emphasize that development of models below significant capability thresholds should be permitted without burdensome regulation. However, preventing the emergence of superintelligence would require a global regulatory regime, which even then cannot guarantee effectiveness.
