Music platforms operate under the same constraints. Artists upload their work to corporate servers, submit to platform policies, and accept payment terms set by intermediaries. Listeners stream through interfaces designed to maximize time spent, not connection to the music. The relationship between creator and audience is mediated at every point by entities whose interests are not aligned with either.
This project starts from a different set of premises. What if every user ran their own node that can be installed just like any other desktop application? What if posts were stored as connections between concepts rather than rows in a database? What if feedback carried actual meaning and actual value? What if artists received payments directly from listeners with no platform taking a cut?
How it should work
Each user controls an identity neuron secured by their own private key. Authentication happens via cryptographic signature. This identity accumulates reputation, relationships, and history that travel with the user across applications.Posts and comments are stored as neurons in the hypergraph. Full text is distributed via the Alt Data Network and referenced through relons. Comments link to parent posts through typed relationships, creating a semantic thread structure. Following another identity means your node tracks that neuron and includes its activity in your locally generated feed.
Content discovery runs on meaning rather than engagement. When someone posts about a topic, they can attach semantic signals: this post is about privacy, this post expresses hope, this post argues against something. These attachments are explicit, created by the user at posting time, not inferred by an algorithm. Your node queries the hypergraph for content whose semantic attachments align with your interests. The entire operation runs locally on consumer hardware because the hypergraph is optimized for traversal, not because a central server is doing the work.
FEELZ/OPINIONZ
The feedback system allows responses across multiple dimensions. Emotional reactions, intellectual positions, value judgments. Each is represented as a distinct neuron. When you react to a post, you select the neuron that matches your response and attach a micro-payment in annecoin. The payment flows directly to the creator. Reactions without value are not processed.This does several things. It captures what a like cannot: the specific nature of the response. It creates an economic circuit where valuable content is rewarded directly. It gives creators not just engagement metrics but a rich picture of how their work resonates.
Music streaming
Artists publish tracks by creating track neurons with metadata: genre, mood, instruments, themes. This metadata becomes part of the hypergraph, enabling semantic discovery. Audio files are segmented and distributed via ANTOR, with the artist's node serving as the initial seed. Listeners stream directly from the peer-to-peer network and can send payments directly to the artist's wallet on a per-stream basis, as tips, or via subscription. When listeners react to tracks with FEELZ/OPINIONZ, artists receive both payments and the semantic feedback.The currency question
Two currencies operate in parallel. Annecoin is required for all hypergraph operations: posting fees, feedback micropayments, anything that creates a relon. This ensures low overhead and seamless integration with the semantic layer. Monero provides a privacy-focused channel for artist compensation and larger transfers. Users who hold only Monero can acquire annecoin through community peer-to-peer swaps, where swap offers are published via the Alt Data Network and settled directly between parties. The platform never touches user funds in either currency.Governance
General content is not moderated by any central authority. Users control their own experience through individual filters, blocking specific neurons, topics, or identities they wish to avoid. They can subscribe to community-maintained block lists if they want additional filtering.For manifestly illegal content, a community-driven reporting and moderation process exists. Flagged content is reviewed by a rotating pseudonymous moderator pool drawn from trusted community members. Verified illegal content can be removed from the hypergraph via protocol-level pruning mechanisms. All moderation actions are recorded as signed relons, creating a transparent and auditable history.
Reputation
Reputation emerges from activity rather than being imposed as a score. Every interaction, posting, reacting, curating, sending payments, contributes to a multidimensional reputation graph stored in the hypergraph. Users and algorithms can weight these signals according to individual preferences. Someone seeking thoughtful commentary might prioritize others who consistently receive reflective reactions. Someone seeking entertainment might follow identities with high joyful and awe feedback. The reputation model is transparent and user-configurable.The legal posture
The project does not incorporate. It does not establish a foundation. The code is open source, the network runs on independently operated nodes, and there is no central organization to target. Node operators are passive infrastructure providers with no financial control, no data custody, and no editorial role. Legal action would need to target individual operators worldwide, a strategy with practical and jurisdictional barriers.Questions worth discussing
- Whether a platform where every reaction costs money can attract mainstream users
- How semantic discovery compares to algorithmic feeds in practice
- What the FEELZ/OPINIONZ system reveals that simple likes cannot
- Whether artists fare better under direct compensation than platform models
- How the two-currency model works for users who only hold one
- Whether community moderation of illegal content can function without capture
- What happens to discoverability when there is no central feed algorithm
- How reputation graphs built from multi-dimensional feedback actually behave
- Whether running a node is a barrier or a feature for different user types
- How semantic attachments at posting time compare to algorithmic content analysis
- What the legal arguments amount to in practice against determined adversaries
For further insights, check out the paper "A Decentralized Social Platform with Music Streaming on ANNE and Monero" at https://anne.media/decentralized-social ... nne-monero. If you have thoughts on any of this, bring them on.