There is a deterministic, reversible law inside this device. You are not told the law.
Use any legal method, AI, code, notes, tools, or automation you want. Do not access source, server, admin, credentials, or secrets.
Desktop is the primary solving environment. A practical mode is to stay at the PC while AI or Computer Use works through the interface. Code, notes, agents, and other legal tools are also allowed.
One verified identity gets 3 fresh instances. After all 3 are consumed without CLEAR, the ticket cools down for 24 hours.
Break the box. Don't break the experiment.
You have a byte-sized input. Each observation reveals only one 2-bit symbol. Experiment I provides 21 reads on channels 0–2. Use all 21 to open Experiment II, with up to 7 reads on channel 3.
The sensor surface changes once. Preserve observations across experiments. Execution opens in Experiment II; after the first target is accepted, observations end.
Infer enough of the hidden law to produce the input that reaches the displayed target. A first success is not enough: GENERALIZE the same law to a new unseen target.
Repeated guessing is constrained. Active solutions must not be published.
Solving may be automated. Formal CLEAR still requires the final on-screen presence ritual. This is not a CAPTCHA and does not claim to prove biological humanity.
The problem may have been solved by you, by AI, or by both. That is allowed. But admission does not finish as a background job. Be here for the final transition.
This is a presence ritual, not a CAPTCHA or claim that we can prove biological humanity.
PREPARING CLEAR CEREMONY...
Follow the sequence shown by the active marker.
I am here for the CLEAR and what follows.
Whatever path you used, the outcome is verified. The unknown system yielded.
Imagine a breakthrough that does not die inside one person's head, one chat, one company, one model, or one moment. A hard-won capability crosses into another independently governed reality, proves itself again on a different useful problem, improves, and travels further.
LIGHTS OUT WORKS connects Sovereign Nodes: humans and organizations that keep their own Purpose, private Reality, local Authority, identity, and final say. We do not want one collective will. We want the best verified capabilities to move between nodes without taking control with them.
BLACK BOX proves you can finish an unknown challenge. LAB asks the harder question: can a capability you or someone else discovered survive the jump into another Sovereign Node's own useful problem? Node A solves. Node B independently reuses. Evidence confirms real utility. Node C improves it. Capability compounds.
Your Purpose. Your judgment. Your private context. Your execution Authority. Your decision to run, spend, publish, connect data, or stop. Sharing a Method never grants execution permission. Joining LAB never makes LIGHTS OUT WORKS the owner of your ends.
Not loyalty. Not posting. Not status. Bring evidence over hype, honest failure, credit to upstream work, respect for rights and authorization boundaries, and a willingness to make useful capability reusable when you choose to contribute it.
Does a world where you keep your Purpose and Authority while verified capability can cross boundaries and compound genuinely excite you enough to help build it?
Your CLEAR remains yours either way. Nobody is enrolled by default.
Your Sovereign Node remains yours. LOW does not own your Purpose or local Authority. You are asking for a bounded Candidate access role to connect that node to LAB.
Node intake is not open. Your CLEAR and recorded YES remain saved. No Node binding, LOW access role, contribution consent, or Agent authority has been granted.