从AI到爱有多远:AI伴侣的情感慰藉与风险考量
From AI to Love: How Far Is the Distance? Emotional Comfort and Risk Considerations of AI Companions
摘要: 随着AI的物质性基础和技术可供性不断地更新迭代,现如今AI伴侣作为人类交往中的新对话主体,日益参与到社会互动的方方面面。本文对现有的AI伴侣应用进行考察发现,用户与AI伴侣的关系走向通常经历了好奇、孤独心理驱动的初见——自适应交流与去社交压力吸引下的发展——从情绪倾诉到情感依赖的深化三个阶段。此过程中,用户享受AI伴侣的定制化、陪伴可得性以及服务性所带来的交流快感,但同时也潜藏着隐私泄露、自我认知失衡以及浅层交流等危机。这场人类与AI伴侣情感互动何如,仍是一场没有保证的冒险,由人类自然语言与0/1机器语言所建立的对话关系,其最终走向何方依旧在重重迷雾之中,难以明确预见。
Abstract: With the continuous iteration of AI’s material foundation and technological affordances, AI companions, as new conversational subjects in human interactions, are increasingly participating in all aspects of social interaction today. This study examines existing AI companion applications and finds that the relationship between users and AI companions typically goes through three stages: initial contact driven by curiosity and loneliness; development attracted by adaptive communication and reduced social pressure; and deepening from emotional venting to emotional dependence. During this process, users enjoy the communication pleasure brought by the customization, accessibility of companionship, and service-oriented nature of AI companions. However, there are also hidden risks such as privacy leakage, self-cognition imbalance, and superficial communication. The interaction between carbon-based organisms and silicon-based organisms remains an uncertain adventure. The final direction of the conversational relationship established by human natural language and 0/1 machine language is still shrouded in mystery and difficult to predict clearly.
参考文献
|
[1]
|
张意茹. 生成式AI伴侣的媒介可供性对用户情感建构的影响研究——以“星野”App为例[J]. 科技传播, 2025, 17(14): 44-48.
|
|
[2]
|
匡文波, 姜泽玮. 聊天机器人用户的算法想象与人机关系: 基于算法民间理论的进路[J]. 现代传播(中国传媒大学学报), 2024, 46(9): 132-140.
|
|
[3]
|
Mitchell, M. 人工智能聊天机器人为何会撒谎[J]. 科技导报, 2025, 43(15): 12-14.
|
|
[4]
|
张晓辉, 孙菁苓. 对虚拟AI言说: 用户对聊天机器人的情感联系探析——以软件Replika为例[J]. 现代传播(中国传媒大学学报), 2023, 45(9): 124-133.
|
|
[5]
|
杨钊, 仲佳. 青年群体“AI伴侣”走热的生成逻辑、伦理风险和引导方向[J]. 宁夏社会科学, 2025(3): 146-155.
|
|
[6]
|
苏歆海, 钱铭钢, 刘成. 个性化人机信息交互中的大模型微调技术研究[J]. 科技视界, 2025, 15(14): 13-15.
|
|
[7]
|
郭陆祥, 王越余, 李芊玥, 等. 大语言模型智能体操作系统研究综述[J/OL]. 计算机科学, 1-14. https://link.cnki.net/urlid/50.1075.TP.20250722.1601.005, 2025-11-06.
|
|
[8]
|
付志远, 陈思宇, 陈骏帆, 等. 大语言模型安全的挑战与机遇[J]. 信息安全学报, 2024, 9(5): 26-55.
|
|
[9]
|
宋美杰, 刘云. 交流的探险: 人-AI的对话互动与亲密关系发展[J]. 新闻与写作, 2023(7): 64-74.
|
|
[10]
|
Internet Matters. 什么是Replika? [EB/OL]. https://www.internetmatters.org/zh-CN/advice/apps-and-platforms/entertainment/replika/, 2025-10-10.
|
|
[11]
|
单珣, 王珏. 生成式人工智能交互中意识形态感性化认同的行为机制及其调制——以高易感性青年群体为例[J]. 中国文化研究, 2025(2): 69-81.
|
|
[12]
|
Duffy, C. (2024) “There Are No Guardrails”. This Mom Believes an AI Chatbot Is Responsible for Her Son’s Suicide. CNN Business. https://edition.cnn.com/2024/10/30/tech/teen-suicide-character-ai-lawsuit/index.html
|
|
[13]
|
张平, 裴晓军. 准社会关系理论在传播研究中的价值、意义与方法[J]. 东南传播, 2014(7): 2-4.
|
|
[14]
|
约翰·杜翰姆·彼得斯. 对空言说: 传播的观念史[M]. 邓建国, 译. 上海: 上海译文出版社, 2017: 25.
|