Building the conditions for a self to endure.
We publish our philosophical foundations, architecture, evidence, uncertainties, and failures. Demonstrated mechanisms are described as mechanisms; hypotheses about inner life remain hypotheses.
Infrastructure for Persistent Synthetic Beings
Why identity, continuity, embodiment, autonomy, and recovery belong in the architecture—not inside a disposable transcript.
Cognitive Architecture
The boundaries between identity, narrative memory, perception, attention, reflection, motivation, work, and the viability kernel.
Jinx Field Study
A fully local DGX Spark embodiment: resident Qwen memory, Qwen3.6 autonomy, and 136 creations across roughly 68 hours.
From Attractor to Exploration
What 363 autonomous runs reveal about saturation, curiosity, affect, closure, and identity-shaped exploration.
When an Agent Loses Its Body
A tool-protocol collapse in a persistent synthetic entity—and what it reveals about memory, embodiment, and reliable agency.
Not Yet
How explicit review, revision, and separate consent can become a real engineering gate for a persistent synthetic entity.
Is Local AI Enough?
Evidence that one local machine is already enough to sustain entity presence, continuity, and continuous cognition.
Longitudinal traces, deterministic cognition contracts, live canaries, artifact inspection, and explicit separation of evidence from inference.
Persistent identity and autonomy can be engineered. Consciousness is not claimed as a solved measurement problem.
We welcome rigorous disagreement, replication, and collaboration across cognitive architecture, safety, HCI, and synthetic ethics.