Public and expert opinions on artificial intelligence are evolving quickly, but there isn't enough systematic data tracking how people feel about AI's moral implications, perceived consciousness, and societal impact. Without this information, policymakers and developers risk creating AI systems that don't align with public expectations, especially as AI becomes more advanced and integrated into daily life.

Tracking Shifting Perceptions Over Time

One way to address this gap could be through a recurring, nationally representative survey in the U.S., with potential to expand globally. The survey would measure how attitudes change year-over-year, with core questions staying consistent for comparison while allowing new questions to address emerging AI developments. For example, it might track whether people believe advanced AI can have moral responsibility or consciousness, and how these views differ between the general public and AI researchers.

Comparing Expert and Public Perspectives

A unique aspect of this approach would be surveying both general populations and AI experts (like researchers at major conferences) to identify where their views align or diverge. While the public might focus on existential risks, experts might have more nuanced technical concerns. This comparison could help bridge communication gaps between AI developers and society.

Practical Applications and Adaptability

The data could help multiple groups:

• Policymakers could use it to shape regulations that reflect public concerns
• Developers might adjust their approaches based on ethical expectations
• Researchers could study how AI perceptions evolve alongside technological progress

To stay relevant amid rapid AI advances, the survey could use a modular design - keeping some foundational questions while adding timely sections about new developments like generative AI or autonomous systems.

Starting with a U.S.-focused version before expanding internationally might make this more manageable as an initial approach. Partnering with academic institutions or existing survey organizations could help with implementation while maintaining rigorous methodology.