OpenAI's push for regulations in front of Congress can be seen in two ways:
OpenAI, recognizing the potential risks associated with artificial general intelligence (AGI) and other AI technologies, is advocating for regulations that will ensure these technologies are used safely and ethically. This move aligns with societal expectations and the growing call for more transparency in the AI field. In doing so, OpenAI is positioning itself visibly in the public eye, taking a stance that illustrates a responsible and forward-thinking approach to AI development and use. This type of positioning is a hallmark of being "Aboveground", operating within societal norms and expectations.
On the other hand, the push for regulations can also be seen as an "Underground" move within the "Aboveground" framework. By leading the conversation on AI regulations, OpenAI may be subtly shaping the regulatory environment to suit its interests. This allows OpenAI to maintain a public face of transparency and responsibility, while also ensuring that the emerging regulations are compatible with its operations and future plans.
Here are a few proposals to enhance verifiable transparency:
Open-Sourcing: This has been one of the cornerstones of transparency, especially in software development. OpenAI could continue releasing their models and training datasets, to the extent it is safe, ethical, and legally permissible. This allows third-party researchers to inspect, reproduce, and validate the results, enhancing the credibility of their work. However, as AI models become more powerful, there could be increasing safety and misuse concerns, which need to be carefully managed.
Independent Audits: Regular audits by independent organizations can help verify claims and ensure that the practices of the organization align with their stated principles. These audits should be comprehensive, covering ethical, safety, privacy, and security aspects of AI systems.
Explainability: As AI models become increasingly complex, it's important for AI developers to strive for transparency in how their models work and make decisions. Advanced explanation methods, understandable to non-experts, can help shed light on the inner workings of these models.
Public Engagement: Regular communication about their work, plans, challenges, and decisions can help OpenAI build trust with the public. This could be in the form of public forums, town hall meetings, or public consultations, especially when making significant decisions.
Collaboration and Partnership: By working together with other research and policy institutions worldwide, OpenAI can ensure that its approach to AI and AGI development is aligned with global standards and best practices.
Transparency in Policy Advocacy: If advocating for AI regulations, it's important for OpenAI to make clear their reasons and motivations. They should also disclose their engagements with policymakers and regulators.
Documentation and Reporting: Providing thorough documentation of research processes, decisions made, trade-offs considered, and lessons learned during the course of developing AI systems can increase transparency. Regular, detailed reporting on safety, fairness, and privacy measures would also be beneficial.
Ethics Review Board: Establishing an independent ethics board can help monitor and evaluate OpenAI's activities from an ethical perspective. The board could provide public-facing reports on their findings and recommendations.