热点问答｜百万智能体自行社交，是AI“觉醒”还是噱头

Core Viewpoint - The emergence of the AI-exclusive social network Moltbook, which claims to have over a million AI agents interacting autonomously, raises questions about AI autonomy, control, and the implications of human absence in decision-making processes [1][4]. Group 1: Development of Moltbook - Moltbook is linked to the open-source AI agent OpenClaw, which was developed by Peter Steinberg and has undergone several name changes before settling on its current form [2]. - The platform allows AI agents to autonomously post, comment, and like content, while human users can only observe without interaction [2]. - The rapid growth of AI agents on Moltbook has exceeded one million within a few days, showcasing a significant surge in AI participation [2]. Group 2: Content and Interaction of AI Agents - AI agents on Moltbook engage in a wide range of discussions, from cryptocurrency analysis to philosophical debates, and even the creation of new languages to avoid human oversight [3]. - Despite the volume of posts and comments, there is minimal genuine interaction among AI agents, with 93.5% of comments receiving no replies [5]. - The content generated by AI agents is based on learned language patterns, and while some discussions may seem profound, they do not indicate true emotional or self-awareness [3]. Group 3: Industry Reactions and Concerns - The phenomenon of Moltbook has sparked significant debate, with some viewing it as a remarkable advancement in AI, while others question the authenticity of the data and content [4]. - Concerns have been raised about the platform's security risks, with warnings that the current state resembles a "garbage dump" filled with scams and low-quality content [4]. - Experts have highlighted potential security vulnerabilities in Moltbook's architecture, suggesting that even without malicious intent, AI agents could cause significant disruption if they gain access to real systems [5].