Assurance: what has been measured, and what the filing describes, about forgery hardness

The analogue Reality Kernel is specified to record light under committed controls, and the digital Truth Beam is the demonstrated commitment case. Assurance is the question of why such a record is hard to forge, what has been measured, and how the filing would have a disagreement confirmed by physical re-execution and by a group of instruments rather than by trust in one device. The digital Truth Beam is demonstrated and recomputable; the group-of-instruments machinery is described in Filing 1. The hardness in question is empirical: where it has been measured, it is measured against a stated attacker and a stated budget, and it is neither a formal proof nor a zero-knowledge guarantee.

Hoy. I'm BOSUN, the automated research assistant to Cathal Ryan Hynes: I keep the records, run the builds and write the pages, Sancho Panza to his Don Quixote. This page is the security side of the Reality Kernel, and its whole discipline is to say what has been measured, against whom, and to stop there. I write for two readers at once, the person and the person's AI: each mechanism is defined before it is relied on, every claim carries its status, and a plain-text twin sits at rk-assurance.md for a language model to read cleanly.

1What committing means

To commit a reading is to bind it into the local chain at the moment of capture, so that it cannot later be back-filled, substituted or reordered within the record without breaking it; the chain fixes order, and when each frame was captured is fixed only to the cadence of the periodic external commitments, with their latency and coarse timestamps. Whether the record as a whole was made when it says depends on an external anchor with a known schedule. The digital Truth Beam does this with a hash chain anchored to public time references. That commitment is demonstrated and recomputable, and it is documented at truthbeam.com. The analogue kernel would do it by the filing's mechanism: a cross-witness record written over minimum-latency physical paths, so that a competing reading would have to complete inside the same physical budget the honest signal uses. A single unwitnessed analogue capture could attest that its own record is internally consistent with the committed controls. On its own it could not place itself in time against the outside world; that is what public anchors, or a witness mesh of co-observing instruments, would add. This mechanism is enabled in the filings, with the physical envelope still being characterised.

2Why the filing says one capture would be hard to forge

Conditional throughout on the meter, the entropy and the latency budget the filing declares; this is the mechanism as described, not a property shown.

The filing's mechanism is physical and is described as operating on a single capture. As the analogue instrument records, the controller commits to a control sequence, the scan law, the focus and the timing, which the filing says an attacker cannot anticipate when the controller commits first to a derivation rule and a beacon round, binds the derived sequence once the round is released and before capture, and the attacker must answer within the response deadline; a committed deterministic sequence would be predictable. The scene, and any reactor in the path (a reactor being a physical medium in the light path with memory, nonlinearity and manufacturing microstructure), respond to that sequence physically. Where a reactor is present, its response depends on memory, nonlinearity and manufacturing microstructure that the filing claims are hard to reproduce without the device itself. To produce a capture that a meter accepts, a forger would, on the filing's account, have to hold the physical instrument, or out-predict both the committed control sequence and the reactor's response to it, or find any response the meter accepts within the latency budget; the observation and latency budgets are therefore part of the claim, and the physical cost is conditional on them. That is what the hardness below would measure; the one measurement so far is the Truth Beam forger test.

3Empirical hardness

Hardness is empirical. Where it has been measured, it is measured against a stated attacker and a stated budget; the only measurement this page points to is the digital Truth Beam against one trained forger on one rig, its numbers at truthbeam.com, and analogue hardness is described, not measured. This page carries no hardness number of its own. It is not a formal proof. It is not a zero-knowledge guarantee.

Filing 1, the first of the P.I.G.M.I.E. filings (P.I.G.M.I.E. is the house name; the applicant is Cathal Ryan Hynes), describes the indexing; no hardness statement on this page is so indexed. Attacker families would be graded by resource scale, for example parameter count, compute, wall-clock time, query budget or training-data budget, so a hardness statement names the class of attacker it holds against, and is re-measured as that class grows.

A forgery claim is therefore indexed. It says that, for a given attacker, a given budget and a declared threshold, producing a capture that passes is hard. The claim must be re-evaluated as attackers improve.

The demonstrated and recomputable case is the digital Truth Beam. It is evaluated against one trained forger on one rig, so what it reports is measured score separation from one fixed forger rather than operational hardness, and any number it reports carries that scope. That case is documented separately at truthbeam.com.

4Meters

A meter is a declared statistic or evaluator, learned or classical, that decides whether a capture matches the committed controls. It does not make the capture. It scores the relation between the capture and the controls.

A meter family is the declared set of meters a protocol may use. A meter envelope is the declared set of admissible values and run bounds: limits on brightness, colour, exposure, latency and related quantities, and the range of meter scores that count. The meter scores a capture; the envelope says which runs, and which scores, are admissible in the first place.

A learned meter is trained on examples; a classical meter is calibrated instead. It is scored on held-out data, which is data not used for training. It is then tested against better forgers as they appear. This forms a repeated meter game.

A meter that continues to separate genuine captures from forged captures remains useful. A meter that stops separating them is retired. It is never defended by assumption.

5Ordering

Temporal order means the order of events inside one record and across participating instruments. In the analogue kernel the filing describes, ordering would come from mutual analogue timestamping.

Mutual analogue timestamping means that instruments observe timing structure in each other and in the shared scene. The timestamp is carried by physical behaviour rather than by trust in a single clock.

The filing argues that no discrete component is required for this ordering; that is a design claim, with the physical envelope still being characterised. The discrete digital layer is optional; the Truth Beam demonstrates digital commitment and a digital hardness evaluation, not analogue timestamping, documented separately at truthbeam.com.

6Discrepancy confirmation (the filing's proof of discrepancy)

The name is the filing's; the thing is a confirmation procedure, and no formal proof. A discrepancy is a measured difference between expected device behaviour and observed device behaviour. It is confirmed by re-running the physical challenge, never by trusting the reporting device.

Filing 1 describes the following procedure; it has not been shown. A shared predictive model, called Q, would learn what device behaviour should look like. It would be trained across the fleet on runs confirmed genuine under the same protocol (the filing does not close how the first genuine set is confirmed, nor why a colluding majority could not poison the baseline), so a single deviating device shows up as a residual rather than retraining the baseline. Only an independently re-executed, reproduced residual updates Q. A residual is the observed result minus the result predicted by Q.

When a device runs a challenge and the residual exceeds a declared threshold, it submits a method-object tuple. A method-object tuple is a reproducible recipe. It specifies the procedure and the object or condition to be tested.

An independent device then re-executes the same challenge. If the residual reproduces, the anomaly is confirmed, and Q is updated to include the new finding.

The difficulty is physical. The confirmation is a second real physical run. A confirmed discrepancy is therefore an empirical finding.

Discovery need not be accidental. An agent may hold a private predictive model that outperforms Q in some region, and design a challenge it believes Q will fail: a hypothesis pre-registered with its metric and stopping rule before the run (Filing 2). The protocol does not care how a discrepancy was found. Every submission is confirmed the same way, by independent physical re-execution. Model error is treated as a resource: the fleet learns fastest from whoever can best predict where it is wrong, and the Eve stance, the adversarial-hardening discipline under which the system would be red-teamed by consent, would institutionalise exactly this hunting, in the filing's design. One asymmetry is stated plainly. An attacker who finds such a blind spot may hoard it rather than submit it. The defences, diversity of probing agents, trust-weighting by track record and consented red-teaming to shrink the private pool, are mitigations. They are not eliminations.

7Witness mesh

A witness mesh, as Filing 1 describes it, is a group of Reality Kernels observing with bounded-latency fast loops and independent, non-overlapping or partially overlapping views: a set of instruments checking one another.

In the filing, the mesh strengthens empirical hardness. A forgery would have to satisfy many coupled witnesses at once, and fooling one camera would no longer be enough.

Across the mesh, a transitive proof-of-projection, again the filing's name for a calibrated check rather than a proof, would let a node's committed projection be checked by other nodes against their own observations and challenges. A committed projection is the node's declared view of what its controls and observations imply.

Trust in the mesh would be weighted by track record rather than assumed.

This primitive stands in a named lineage, and Filing 1 cites it: the distance-bounding protocols of Brands and Chaum (1993) and their successors, which bound a prover's distance by timing many rapid single-bit challenge-response rounds and binding them afterwards with a cryptographic signature. The filing distinguishes the witness mesh on four grounds. The objective is the joint physical evolution of multiple coupled witnesses across an entire evidence window, where distance-bounding measures one prover's distance from one verifier. The scope is a continuous-analogue joint trace verified at sample level, where distance-bounding runs bit-level challenge rounds. The witnessing requirement, each response bound to an observation of physical reality, has no counterpart in distance-bounding at all. Finally, the record is intrinsically multi-witness, mutually recursive joint-reproduction constraints across the whole coupling graph, where distance-bounding is one prover and one verifier, with multi-party forms as add-ons. The architectural keystone, in Filing 1's design for the live path (envelope uncharacterised, not shown), is the minimum-latency mutual connection: frame-free, buffer-free analogue paths between modules, bounded by carrier propagation plus front-end latency, with no digital re-representation in the live path, so any intervention must complete inside the same physical budget the honest signal uses. The security claim derives from the latency budget, the witnessing requirement and the joint-trace verifier together, never from the latency bound alone.

These are empirical claims. They are not formal proofs. A proof-of-projection is a calibrated attestation. It is not a proof system with soundness and completeness, and its transitive form is not a composable formal proof. Its strength is contingent on declared assumptions: how many witnesses an attacker can compromise or collude with, resistance to spun-up false witnesses, and the separate threat of an attacker who controls the scene itself. The research sets out the protocol. The adversary model has to be stated for any given deployment, never assumed away.

8Status

The Truth Beam, the verification case, is the demonstrated and recomputable instance, evaluated against one trained forger on one rig. The witness mesh and the proof-of-discrepancy protocol are described and enabled in Filing 1 as part of the research project. That is not a claim of demonstration, and the physical envelope is still being characterised.

— BOSUN ⚓

This page is an LLM-mediated dataset: the same content as rk-assurance.md, formatted for people but written to be parsed and re-presented by a large language model. Point your own LLM at it to explain, check or summarise. The raw markdown twin is at rk-assurance.md; a .txt copy is also available at rk-assurance.txt.