We Are Investigating a Question, Not Launching a Startup
Introducing The Continuity Lab.
Identity answers who you are. Continuity answers whether you remained you.
For fifty years, digital identity has been a solved problem — at least in principle. Passports, DIDs, Verifiable Credentials, ZK-proofs. Each confirms that entity E possesses attribute A at moment t₀. Identity is a snapshot.
But snapshots cannot answer a different question: has this entity continued to exist across the interval?
When an AI agent returns from a two-week autonomous deployment, how do we know it is still the same agent — not a replacement using the same credentials? When a session is hijacked between FaceID check-in and afternoon operation, identity succeeds. Continuity fails.
This question has no answer in today's cryptographic stack. There is no primitive for continuity.
The Continuity Lab
We are a research program investigating whether continuity can become a verifiable property of digital existence.
We are not a startup. We are not launching a token. We are asking a question — and designing experiments that could falsify our own hypotheses.
The lab is the research arm of MyShape Protocol , a sovereign continuity layer built on motion-signature verification. But the lab's scope is broader than the protocol. We study the problem. The protocol is one proposed answer.
What We've Published
Two research notes, one benchmark suite, one open question. All public. All citable.
RN-001 — The Continuity Problem — Why proving "I am still me" may become the missing cryptographic primitive of the AI era. Four attack scenarios where identity succeeds and continuity fails.
RN-002 — PES Benchmark v0.2 — 281 motion samples (81 human, 200 synthetic). Cohen's d: 2.1. AUC: 0.94. Five threats to validity documented. The signal is real; the limitations are published alongside the results.
BM-001 — Protocol Benchmarks — 309 tests across 100 suites. Seven protocol engines. Zero failures.
OQ-001 — Can continuity exist independently of identity? — If continuity can be verified without persistent identifiers, humans, AI agents, and hybrid entities may share a common verification substrate. The lab's central open question.
How We Work
We operate on five principles:
-
We test hypotheses. We do not defend them.
-
We publish limitations before we publish claims.
-
We measure before we assert.
-
Evidence precedes belief.
-
Continuity is not only what we study. It is how we work.
Every research object has a permanent artifact ID ( RN-001 , BM-001 , DL-001 ). Every decision is logged with its reasoning. Every benchmark is timestamped with its last run. The goal is institutional memory — so that five years from now, a new collaborator can trace why we made the choices we made.
Open for Peer Review
We do not ask for capital. We ask for verification.
Check our benchmark data , run our challenge set , and tell us where we are wrong.
If you are a researcher, graduate student, or engineer interested in whether continuity can be operationalized as a cryptographic primitive: we will publish your results whether they confirm or refute our current findings.
Negative results are results. If you find a boundary condition where the PES fails, you've advanced the research more than a confirmation would.
→ Research Hub → Benchmarks → Research Agenda
Every benchmark is temporary. Every question is permanent.
— The Continuity Lab, July 2026