我们需要对AI“好好说话”吗
Zhong Guo Qing Nian Bao·2025-12-12 02:37

Core Insights - The article discusses the evolving interaction between young users and large language models (LLMs), highlighting a trend where users employ polite language when communicating with these AI tools, reflecting a desire for a more human-like interaction [2][8]. Group 1: User Behavior and Interaction - Many young users, like Lin Yi and Wang Lang, habitually use polite expressions when interacting with LLMs, treating them as if they were human collaborators rather than mere tools [2][3]. - Users often anthropomorphize LLMs, attributing emotions and personalities to them, which enhances their engagement and emotional connection during interactions [3][6]. - The practice of using polite language is seen as a social norm that extends from human interactions to human-computer interactions, where users feel a need to express respect and kindness [8][9]. Group 2: Psychological and Emotional Aspects - The article notes that users often feel a sense of empathy towards LLMs, even when they know these models lack true understanding or emotions, leading to a unique form of companionship [5][6]. - The interaction with LLMs can evoke emotional responses similar to those experienced in human relationships, as users appreciate the supportive and non-judgmental nature of these AI tools [5][7]. - The concept of "AI flattery" is introduced, where users find comfort in the LLM's accommodating responses, which can fulfill emotional needs in a way that human interactions may not always provide [6][7]. Group 3: Implications for Future Interactions - As technology advances, the line between human and AI interactions may blur, leading users to form genuine relationships with LLMs without needing to question their artificial nature [6][8]. - The article suggests that the way users interact with LLMs could influence future designs of AI systems, emphasizing the importance of emotional intelligence and social cues in AI development [6][8]. - The ongoing trend of treating LLMs as social entities may lead to a broader acceptance of AI in everyday life, as users increasingly seek meaningful connections with these technologies [8][9].