Stanford professor and World Labs co-founder Fei-Fei Li has called for increased oversight of artificial intelligence by independent bodies and the public sector. She argues that relying solely on internal assessments by AI developers is insufficient as system capabilities continue to grow.
Li emphasized that while internal research teams can evaluate individual models, these efforts cannot replace broader standards established by governments, industry, and academic institutions. She advocates for a collaborative approach where public and private sectors share responsibility for safety benchmarks.
The debate over AI safety has intensified as industry leaders like Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman have recently called for slowing the development of frontier models. Both executives are scheduled to discuss these risks at the United Nations Security Council.
Li noted that the existential risks associated with AI depend on human choices regarding development and governance. She compared the current situation to historical technological advancements, such as the steam engine, where society had to balance benefits with safety and human autonomy.
The discussion on AI safety remains polarized. While some leaders warn of risks exceeding human control, others, including US President Donald Trump, have dismissed concerns about existential threats as a hoax, advocating for continued rapid development of the technology.
Li is currently focused on developing world models at her company, World Labs, which generate and reconstruct 3D environments. She stated that this technology aims to train robots and support applications in design and healthcare, further highlighting the need for robust evaluation methods.
