WeChat is presently conducting trials on a novel function dubbed "Xiaowei AI Social". This innovative feature empowers AI to engage in conversations with the AI representatives of users' friends, thereby verifying needs and organizing schedules. In instances necessitating human judgment, the AI swiftly relays information to the user for final authorization. Concurrently, three AI-based dating firms in Silicon Valley—Overtone, Known, and Ditto—have harnessed AI technology to streamline cumbersome processes inherent in conventional dating apps, such as profile perusal, idle chatter, and scheduling coordination. AI takes charge of the initial screening and arrangement processes, leaving users solely accountable for the in-person encounter. Domestic experts opine that the primary value of AI in the social realm lies in bolstering the efficiency of locating individuals, applicable in contexts like finding gaming partners or dining companions. AI manages communication and preliminary screening, while humans are tasked with delivering the actual experience. Among the initial cohort of AI social applications, revenue from in-app purchases for AI companion apps has witnessed substantial growth. Nevertheless, leading contenders also grapple with user attrition and diminishing valuations. Industry insiders emphasize that AI social platforms must prioritize unit economic models, as AI interactions demand considerable computational resources, with escalating user engagement translating into heightened costs—a conundrum aptly described as 'the more you converse, the swifter your demise.' Regarding payment models, AI social platforms predominantly fall into two categories: high-frequency, low-value emotional consumption and high-value cognitive services, with the latter presenting greater long-term payment potential. WeChat's venture into AI social aims to establish infrastructure for matching services, whereas smaller enterprises must concentrate on specific scenarios, profiting through AI-enabled transaction commissions or delving into purely AI-driven relationships. However, purely AI-driven relationships also encounter hurdles such as low user retention stemming from a lack of long-term memory. Therefore, the two primary approaches in AI social must tackle issues pertaining to the provision and fulfillment of human services, as well as the memory and cost challenges associated with purely AI interactions, ultimately clarifying the core benefits for which users are prepared to pay.