Mesh LLM brings distributed AI inference to iroh's P2P network

tionis built a system for running large language models across distributed peers using iroh, eliminating the need for centralized inference servers. It's an early exploration of how P2P networking primitives could reshape who controls and runs AI workloads—relevant for makers wanting to avoid cloud costs or build resilient, decentralized AI features.