This is not a product roadmap. It is a research agenda — a living document that defines the questions we are investigating, the status of each investigation, and the experiments that could falsify our current hypotheses.
We do not defend hypotheses.
— The Continuity Lab Manifesto, Article I
We design experiments that can falsify them.
Is there a statistically separable signal in human motion that persists across sessions, hardware, and environments — and can it be quantified as a continuity score?
Dependent on PES Benchmark stability validation. Current evidence: Cohen's d = 2.1, AUC = 0.94 (54 samples).
→ RN #001 — The Continuity ProblemUnder what conditions can a continuity proof be replayed, simulated, or adversarially synthesized? What is the minimum signal dimension that resists attack?
Requires quantification of anti-replay thresholds in a controlled Replay Benchmark.
If a human delegates agency to an AI, does continuity transfer? How do we define 'entropy of presence' for an agent with no biological substrate?
Core challenge: defining existence entropy for non-biological entities. This may become the defining question of the post-agent era.
If continuity is measurable and verifiable, can it be encoded as a cryptographic primitive — alongside identity, encryption, and consensus — in decentralized protocols?
Active prototype: Continuity Receipt format. Exploring ZK-friendly encoding of motion-derived continuity proofs.
→ RN #001 — The Continuity Problem