What AI Regulation and Ownership Could Be

In a recent interview, B Cavello, the director of emerging technologies at the Aspen Institute, discussed the need for public input in AI development and regulation. She emphasized the importance of public AI, which centers public access, accountability, and sustainable public goods. Public AI aims to ensure free access to information, as well as the ability for users to build models in their local context.

Cavello also highlighted the significance of accountability, which involves preventing potential harms and encouraging the development of beneficial capabilities. She explained that treating AI as public infrastructure could lead to a more reliable and universally accessible technology.

The interview also touched upon various stakeholders and communities participating in public AI discussions. These include movements around AI sovereignty in Africa and Europe, as well as conversations in the US regarding a potential AI bubble and the need for public ownership or uptake of these technologies. Cavello expressed skepticism about proposals for financial ownership stakes in major AI companies, suggesting that focusing on power distribution, similar to windfall trust networks, might be a more effective approach.

When asked about the policy implications of these discussions, Cavello compared the situation to the renewable energy transition, where both technological advancements and supportive policies are necessary. She noted that Congress tends to be reactive, and public awareness and engagement will likely drive policy changes. As regular internet users, Cavello encouraged increased attention to public accountability in data center operations, transparency in technology companies, and set-asides for data centers that allocate a portion of their resources for public access.