AI chatbots need 'deception mode' (www.computerworld.com)

Researchers at NYU Tandon School of Engineering have found that users perceive AI chatbots' responses as more thoughtful when they experience intentional response delays, a surprising shift from the typical preference for speed in technology. This phenomenon links slower response times to perceived deliberation, suggesting that a "context-aware latency" could enhance user satisfaction in chatbot interactions. Designed to create a perception of thoughtfulness, this approach could lead companies to reevaluate how they configure response timings for varying complexities of inquiries. The study raises significant ethical concerns about anthropomorphizing AI, highlighting the risks of misleading users into believing chatbots have human-like qualities. Bioethicist Jesse Gray proposes a "deception mode" that would allow users to consciously engage with chatbot features that mimic human empathy or humor, thereby enhancing user awareness of the AI's true nature as a tool rather than a sentient being. This solution aims to balance the emotional needs of users with the necessity for transparency in AI capabilities, addressing the potential for harmful attachments to technology and fostering a more informed user experience in the AI landscape.