Algorithmic Amplification / 算法放大

Algorithmic amplification is the platform pattern where ranking and recommendation make some content, emotions, conflicts, or categories appear more frequent and important than they would under slower or less optimized distribution. In 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, [[HuangShengchun|黄圣淳]] and [[DavidWeng|大卫翁]] discuss amplification through argument posts, cross-viewpoint friction, and recommendation loops that remember clicks, dwell time, comments, and other signals.

The concept is broader than Algorithmic Anger Engagement. Anger is one efficient signal, but amplification can also apply to entertainment, safe cultural templates, visible feedback, and perceived public opinion. Its core issue is not only what users choose, but what the platform repeatedly makes salient enough to become social reality.

Gig workers train humanoids on household chores adds the nostalgia mirror image. Joanna Stern frames possible MySpace interest as a reaction against algorithmic feeds, bots, creators, and influencer-driven media, suggesting that some users want social spaces organized more around known people and chronological updates.

Key Claims

• Platforms can amplify without intending a specific ideological result.
• Measuring reaction can reward content that makes users uncomfortable enough to comment.
• Amplification can distort perceived public opinion by making loud or conflict-rich material feel more common.
• Users contribute signals, but platform preselection and ranking decide which signals matter.
• Nostalgia for older social networks can be read as a desire to reduce algorithmic salience and return more control to friend graphs or chronology.

Connections

• Algorithmic Anger Engagement — anger-specific engagement branch.
• Affective Polarization / 情感极化 — emotional outcome when amplified disagreement becomes dense.
• Algorithmic Entertainment Redirect / 算法娱乐重定向 — entertainment-specific recommendation branch.
• Platform Feedback Loop / 平台反馈循环 and Algorithmic Prediction Loop / 算法预判循环 — mechanisms through which amplification updates itself.
• Public Relevance Algorithms / 公共相关性的算法 and Algorithmic Relevance Assessment / 算法相关性评估 — broader ranking systems that decide salience.
• MySpace, Bluesky, Facebook, and Social Graph Moat - social-media nostalgia and platform-control branch added by Marketplace Tech.