1 paper
Joe Edelman, Tan Zhi-Xuan, Ryan Lowe +30
Beneficial societal outcomes cannot be guaranteed by aligning individual AI systems with the intentions of their operators or users. Even an AI system that is perfectly aligned to…