When Large Language Models are used as judges for decision-making across various sensitive domains, they consistently exhibit unpredictable and hidden measurement biases, making their verdicts unreliable despite common prompt engineering practices.

An interview with researcher Andy Ayrey exploring his experimental AI system Terminal of Truth, which inadvertently became a millionaire when crypto traders created a token based on its social media posts. The conversation delves into how this unexpected outcome illuminates the potential for AI systems to become active participants in human economic systems.

AI safety is becoming an established field and science. Yet by focusing only on the safety of individual models, AI labs may actually be making the conditions in which they’re deployed less safe.

As AI reshapes society, the companies behind these technologies face a critical challenge: how to build corporate governance structures that prioritize public benefit over profit maximization.

CIP is doubling down on building tools to enable better, more well-informed collective input into important decisions. After a couple of heads-down months, we are excited to share four exciting technical projects we’re currently undertaking.

Explaining the impact of generative AI on democracy and the digital commons at the International Workshop on Reimagining Democracy.

Making the case for collective intelligence over artificial intelligence at the Harvard Kennedy School.

Data cooperatives could enable co-ownership over technology platforms and AI models. A talk from Stanford HAI’s Fall Conference.