Tag
A creator explains how the X (Twitter) recommendation algorithm actually works, breaking it down into components like Thunder, Phoenix, and Simcluster to help readers reverse-engineer virality.
YouTube Shorts will downrank recommendation weight for "clipped and reposted" content, further prioritizing original content. Channels built on pure reposts, lightly modified, and templated content will be hit first.
YouTube announces a new AI-powered feature called custom feeds that lets users build personalized video recommendation feeds by describing their preferences, leveraging Google's Gemini model and similar to tools on other social platforms.
Spotify is launching the 'Taste Profile' feature in the US for Premium subscribers, allowing users to adjust their recommendation algorithm via natural language commands to refine music and podcast suggestions.
Social media platforms like Threads, Instagram, and TikTok are introducing tools that allow users to personalize their recommendation algorithms, giving users more control over their feeds.
A developer analyzed X's open-sourced recommendation algorithm and built an AI agent using Claude that autonomously creates and posts content, reportedly achieving 6 million views in two days.
OpenAI outlines the design philosophy behind Sora's feed system, prioritizing creativity, user control, connection, and safety through steerable ranking algorithms and robust content moderation guidelines.