Safe Pareto Improvements Fundamentals Program

• Teach participants about Safe Pareto Improvements (SPIs) as a robust intervention strategy to mitigate risks from AI conflict.
• Cover game theory foundations, bargaining problems, and canonical SPI examples including surrogate goals and delegated game-playing.
• Provide weekly readings, exercises, Slack discussions, and office hours over four weeks with 5-7 hours weekly commitment.
• Offer optional paid capstone projects and two final-week curriculum streams: conceptual or empirical research focus.

This program equips participants with knowledge of Safe Pareto Improvements, a promising approach to reducing risks from AI conflict, thereby contributing to safer AI development and long-term human flourishing.

• Interest in AI safety and long-term risk reduction.
• Background in game theory, bargaining, or related fields preferred.
• Ability to commit 5-7 hours per week for four weeks.
• Open to participants with varying levels of experience; no strict prerequisites.

Optional paid capstone projects available. No other compensation mentioned. Participants gain expertise in a cutting-edge AI safety strategy.

The Center on Long-Term Risk (CLR) focuses on reducing risks from AI conflict and other existential threats, promoting robust strategies for long-term positive outcomes.