The only way to advance AI safety is to increase human interactions, human values and societal governance to promote a reinforced human feedback loop, much like we do with traditional AI training methods.
The post Trolley Problem, Safety Versus Security of Generative AI appeared first on SecurityWeek.