A globally distributed, peer-to-peer (P2P) network for large language models (LLMs) that ensures:
* Privacy: All queries remain on local nodes unless federated updates are performed.
* Censorship resistance: No central authority controls outputs.
* Renewable energy utilization: Inferers only operate when powered by solar, wind, or other green energy sources.
* Continuous operation: Geographic distribution ensures night/day operation.
* Community-driven model selection: Nodes vote on the model to use, ensuring it fits inferer constraints.
* Privacy: All queries remain on local nodes unless federated updates are performed. * Censorship resistance: No central authority controls outputs. * Renewable energy utilization: Inferers only operate when powered by solar, wind, or other green energy sources. * Continuous operation: Geographic distribution ensures night/day operation. * Community-driven model selection: Nodes vote on the model to use, ensuring it fits inferer constraints.