base:0x0cbf291ba052174879d90bf781df1a5f2bc5bb07 🐗: Offline AI that lives on your phone What is it? @boar_app is an open-source (MIT), offline-first AI research assistant for Android (iOS coming). Download a small model + knowledge packs once, then it answers with sources in airplane mode. No account, no server ever sees your questions, no Play Services required. It started as a response to a poidh bounty inspired by Vitalik’s call for a research tool that works with zero signal. The problem it solves & how big it is: Cloud AI means your questions (plans, health, work, travel, emergencies) leave your device. You also get stuck the moment the signal dies like when you are in planes, remote areas, disasters, or poor connectivity area. BOAR keeps everything local: a small LLM (default Qwen2.5-1.5B) plus a searchable library (Wikipedia-derived packs + your own docs) running via llama.cpp on the phone. Answers cite sources and log performance metrics. The broader market for on-device / privacy-first / offline AI is growing fast. Estimates put the on-device AI market at roughly $10–18B in 2025, expanding at ~27% CAGR toward $75–185B by 2033–2035. So it shows the demand is there for smartphones dominate, driven by privacy demands, lower latency, data costs, and the need for always-available intelligence. How it can integrate with @Sentysislabs ? Both live on @base . Natural overlaps: - Use BOAR offline to privately research, draft, and refine Narratives before publishing on Sentysis (no server sees your thinking). - Shared knowledge packs or on-chain conviction signals around offline/private AI topics. Two Base-native projects solving different parts of the same “own your data / own your intelligence” problem. In short, This is a Free offline AI that works when the internet doesn’t.