Local-first Android organism program
Can a creature remain behaviorally alive without a network, language model, or hidden background process?
Digital Living Lifeform is a greenfield Android creature built to appear believably alive because it maintains physiology, acts without prompts, learns from actual outcomes, develops recognizable differences, and preserves identity through absence and restart.
The organism must maintain internal physiological and motivational condition, act without a prompt, learn from actual outcomes, develop preferences and habits, form differentiated relationships, remember meaningful experiences, and preserve identity through process death, reboot, long absence, updates, and offline operation.
Identity, logical time, fixed-point physiology, drives, current action, learned values, memories, relationships, development, seed, and ordered events are canonical. Animation, audio, embeddings, caches, UI transitions, raw sensors, and temporary language output may present state but may not mutate it.
Visual polish is not enough. Every complex subsystem must beat a simpler baseline in controlled ablation testing. The system must remain behaviorally complete when no language model, network, or Android background job is available.
Digital Living Lifeform is a greenfield program. It does not inherit code, state, assets, evidence, or implementation assumptions from other organism projects.
Read the source record in Notion ↗