Filtered-For Content (FFC) - A Decentralized Social Media Architecture
Filtered-For Content (FFC) is a social media architecture that eliminates centralized content moderation, algorithmic amplification, and privileged user statuses (special people). It operates on a strictly follow-based model, meaning users only see content from accounts they follow. There are no algorithmic feeds, bans, shadowbanning, or engagement-based ranking. Instead of moderating or removing content, FFC provides user-driven filters that allow individuals to curate their own experience.
FFC shifts control from the platform to the user:
- No centralized enforcement: Platforms do not remove content except where legally required.
- No privileged users: No moderators, influencers, or verified accounts with special status.
- No algorithmic curation: No engagement-based feed ranking or forced content discovery.
- No engagement-driven virality: Content spreads only through direct reposting by users.
- Strictly follow-based content model - Users only see posts from accounts they follow
- User-driven content filtering - Customizable inclusion and exclusion tags
- Annual username leasing - Prevents spam and abuse with renewable usernames
- No algorithmic manipulation - Content discovery through organic means only
- Bot prevention measures - Financial barriers and rate limits to deter spam
- Private engagement metrics - Follower counts and engagement stats are opt-in
- Pure syndication model - Content spreads only through direct user reposting
- Equal user status - No special privileges or verified accounts
- Legal compliance - Only illegal content is removed, everything else is user-filtered
- Multiple discovery methods - Following, public lists, random browsing, and opt-in pools
- Users only see posts from accounts they follow—there is no algorithmic feed, trending page, or forced discovery.
- Discovery happens through following users, subscribing to public lists, or manually browsing.
- Reposting (platforming) is syndication: If a user reposts content, it appears in the feeds of their followers.
- No content is removed unless it is legally required to be taken down (e.g., CSAM, direct threats).
- Users control inclusion and exclusion tags, filtering their own content experience.
- Shadowbanning and deplatforming do not exist—users determine what they see, not the platform.
- Usernames are leased annually for a fixed cost to prevent spam and abuse.
- If a renewal lapses, the username enters a grace period before becoming available to others.
- Users can interact via unique ID numbers if they do not purchase a username.
- Platforms may allow username transfers at their discretion.
- No engagement-based ranking—users are discovered organically through:
- Following others
- Browsing public lists
- Random profile browsing
- Viewing recent signups by tag (opt-in)
- Discovery pools (users voluntarily opt in to be browsable)
- Since content is strictly follow-based, bot accounts have no visibility unless real users follow them.
- Platforms may implement additional friction-based deterrents:
- Charging for usernames (creates a financial barrier to mass bot creation).
- Follow rate limits (prevents bot networks from growing rapidly).
- Basic human verification (e.g., CAPTCHA, email verification).
- Follower counts, engagement metrics, and reputation signals are opt-in and private by default.
- These metrics do not influence content visibility or discovery.
- Platforms may introduce user trust indicators (e.g., account age, verification), but they must be purely informational and not affect visibility.
- Moderation:
- Traditional C&M: Centralized enforcement via bans, removals, and shadowbans
- FFC: No discretionary moderation—users filter their own feeds
- Content Removal:
- Traditional C&M: Illegal + subjective rule-based takedowns
- FFC: Only illegal content is removed, everything else is user-filtered
- User Control:
- Traditional C&M: Users have limited control over content governance
- FFC: Users define their own experience with inclusion/exclusion filters
- Algorithmic Feed:
- Traditional C&M: Content visibility dictated by engagement-based ranking
- FFC: No algorithmic curation—strictly follow-based like RSS
- Special Users:
- Traditional C&M: Verified accounts, moderators, and influencers get privileges
- FFC: No special users—everyone operates under the same rules
- Spam/Bots Handling:
- Traditional C&M: AI-driven detection and bans
- FFC: Natural friction via paid usernames, follow limits, and human verification
- Discovery:
- Traditional C&M: Algorithmic promotion of trending content
- FFC: User-driven discovery via lists, random browsing, and opt-in exposure
- Engagement Metrics:
- Traditional C&M: Public likes, shares, and follower counts shape visibility
- FFC: Opt-in and private by default, no influence on visibility
- Virality Model:
- Traditional C&M: Algorithm-boosted virality
- FFC: Pure syndication (reposts only seen by direct followers)
Traditional social media platforms claim to help users discover content, but in reality, they restrict user control by forcing algorithmic feeds, trending topics, and engagement-based visibility. This "discovery" comes at the cost of user autonomy—platforms decide what you see, how often you see it, and who gets amplified. FFC rejects this trade-off by giving users complete control over their experience while still enabling organic discovery through following, lists, and opt-in exposure. Instead of algorithms dictating your feed, you choose what you see and how you discover new content. FFC-based platforms are necessary to restore autonomy, transparency, and fairness to the social web.
- No content moderation disputes: Users filter what they want instead of platforms enforcing rules.
- No algorithmic manipulation: Content spreads through user-driven syndication only.
- User-controlled experience: Every user curates their own feed.
- Natural spam deterrents: No engagement-based incentives for bots.
- No special privileges: All users operate under the same rules.
FFC fundamentally redefines social media by making it decentralized, user-controlled, and free from corporate-driven (or special-interest-driven) content curation.
Contributions are welcome! See the CONTRIBUTING file for details.
This project is dual-licensed, MIT for code, and CC BY 4.0 for content (e.g. docs, images). You can choose either license, depending on your use case. See the LICENSE file for details.