Fei-Fei Li on AI Policy: Science Over Science Fiction

Fei-Fei Li's Principles for AI Policy

Fei-Fei Li, a computer scientist at Stanford University and a founder of AI startups, has articulated three core principles to guide future AI policymaking.

These principles are being presented in anticipation of the upcoming AI Action Summit, scheduled to take place in Paris next week.

Grounding Policy in Scientific Reality

Li’s first principle emphasizes that policy development should be firmly rooted in “science, not science fiction.”

Policymakers, she suggests, ought to concentrate on the present capabilities of AI, rather than speculative, extreme future possibilities – be they optimistic or dystopian.

It is particularly important, according to Li, for those creating policy to recognize that current AI systems like chatbots and co-pilot programs do not possess genuine intelligence, intention, free will, or consciousness.

This understanding will help avoid unproductive debates about improbable outcomes and allow for a focus on pressing, real-world issues.

A Pragmatic Approach to Regulation

Secondly, Li advocates for a pragmatic, rather than ideological, approach to AI policy.

This means crafting regulations that aim to minimize unforeseen negative effects while simultaneously encouraging continued innovation.

Empowering the Broader AI Community

Finally, Li stresses the need to empower the entire AI ecosystem, encompassing both open-source communities and academic institutions.

Open access to AI models and the computational resources required to utilize them is, she believes, essential for continued advancement.

Restricting access will inevitably create obstacles and hinder progress, especially for academic researchers and institutions that often have limited funding compared to private companies.

Further details regarding the Artificial Intelligence Action Summit in Paris can be found in our complete coverage.