Project MIRROR: Can AI Learn to Prioritize Evidence Over Social Approval? A proposed longitudinal experiment investigating independent reasoning in continuously learning AI systems. I would like to propose a research question that builds upon existing studies…