AI for Epistemics covers work on using AI to help people and institutions form more accurate beliefs and make wiser decisions. It also covers making sure that as AI becomes more capable, it improves how society reasons rather than degrading it.
Commonly discussed areas include:
- AI forecasting and strategic foresight
- Tools for research, evidence synthesis, argument mapping, and decision support
- AI-assisted deliberation and collective sensemaking, such as tools that find where large groups agree
- AI for government and institutional decision-making
- Verification and provenance infrastructure: fact-checking, Community Notes-style systems, source reliability tracking, and AI-content detection
- Evaluations of how models behave epistemically, such as truthfulness, calibration, and sycophancy
- Risks AI poses to epistemics, including persuasion, manipulation, and unhealthy dependence on AI