Calculated Publics / 计算出的公众

Calculated publics are publics assembled by algorithmic inference rather than by explicit membership or shared public media. 159.算法的六副面孔：它是如何从处理数据，变成定义我们是谁的 presents this as the sixth face of algorithms: platforms decide that “people like you” exist, then route content through that inferred similarity until users can feel themselves inside a group they did not consciously join.

The source contrasts calculated publics with networked publics. A networked public forms when people gather around an interest, identity, creator, or social relationship. A calculated public can be assigned by models using purchase history, watch time, location, gendered language, search patterns, or similarity to other users. The danger is that users may start treating the platform’s temporary classification as the world’s natural structure.

164.算法的“兔子洞”：为什么你总在看完新闻后滑向娱乐？｜对谈黄圣淳教授 adds a perception-of-public-opinion extension. 黄圣淳 notes that algorithmic environments may not directly change political views, but they can make users overestimate how many people support their own view. That makes calculated publics a reality-imagination problem as well as a targeting problem.

167.柏拉图、卢梭、哈耶克、阿伦特四大哲学家会如何解释算法时代？｜串台独树不成林 adds an Arendtian extension through Algorithmic Public Appearance / 算法公共显现. The episode argues that public metrics and inferred groups can make a person appear as a type, label, or data point rather than a full “who” acting and speaking in public.

Key Claims

• Algorithmic grouping can give users identity before they have named it themselves.
• “People like you also liked” is not just convenience; it is a social classification act.
• Calculated publics can amplify cultural bias by presenting statistical regularities as neutral truth.
• The concept helps explain why algorithmic feeds can intensify Group Polarization / 群体极化 and Information Cocoon / 信息茧房 without perfect isolation.
• The source still treats calculated publics as partial and unstable because users often move among platforms and contexts.
• Episode 164 adds that inferred publics can distort perceived majority support even without proving a direct attitude-change effect.
• Episode 167 adds that calculated publics are also appearance systems: they decide whether people show up as persons or as classified fragments.

Connections

• Public Relevance Algorithms / 公共相关性的算法 — umbrella category.
• Algorithmic Labeling and Personalization As Social Identity — related identity-compression and data-story patterns.
• Information Cocoon / 信息茧房, Group Polarization / 群体极化, and Tribal Truth / 部落真相 — downstream belief and belonging dynamics.
• Algorithmic Diversity Dividend / 算法多样性红利 — countermeasure when multiple platforms expose users to different calculated publics.
• Xiaohongshu, Douyin, TikTok, YouTube, and Spotify — platform contexts where inferred publics shape recommendation.
• Filter Bubble / 过滤气泡, Incidental Exposure / 偶然暴露, and Affective Polarization / 情感极化 — episode 164’s exposure and public-opinion-perception extension.
• Algorithmic Public Appearance / 算法公共显现, Hannah Arendt / 汉娜·阿伦特, and Platform Feedback Loop / 平台反馈循环 — episode 167’s public-appearance extension.