Two Levels of Design
Discovery and coordination systems are designed at two different levels, and confusing them produces bad architecture. The first level is behavioral: given a set of independent participants, a behavioral standard specifies how each participant must act so that the participants interoperate. The second level is architectural: given a substrate through which requests and objects flow, an architecture specifies what the substrate itself enforces on every operation, independent of any single participant's discretion.
IETF DULT is a clean, well-scoped example of the first level. This article uses it as a reference point to explain what the second level looks like in the governed semantic discovery substrate disclosed in United States Patent Application 19/647,395. The two operate on different problems in different domains. The comparison is an analogy about levels of design, not a claim that the filing covers consumer location trackers.
What IETF DULT Actually Specifies
The DULT Working Group emerged after Bluetooth crowd-sourced finding networks made covert location tracking a cross-vendor problem: a tracker emitted by one vendor's device travels through the detection surface of every other vendor's devices, so no single vendor could address the harm alone. Apple and Google jointly published a draft in 2023 proposing a common advertising format and detector behavior, and the IETF DULT Working Group formalized the standardization track, with participation from major tracker manufacturers and civil-society contributors.
The specification's scope is deliberately behavioral and precise. It defines how a tracker advertises its presence so that any compliant detector can recognize it. It defines how a compliant detector should behave when it identifies a tracker moving with a person who is not the owner, including alert conditions and escalation to platform-level surfaces. It defines how an affected individual can obtain information sufficient to disable a tracker. The result is a behavioral interoperability standard: any compliant tracker and any compliant detector participate in a shared anti-stalking flow regardless of vendor.
This is genuinely useful work. It produces a cross-vendor baseline that did not previously exist and gives civil-society organizations a stable surface to evaluate compliance against. DULT does well exactly what a behavioral standard is meant to do. It does not, and was never chartered to, specify a governance substrate that enforces properties on every operation independent of each operator's discretion. That is a different level of design, and it is the level the filing addresses.
What Semantic Discovery Specifies
The filing discloses a governed semantic discovery substrate (Chapter 10 of 19/647,395). A discovery object, a schema-conformant semantic agent carrying typed persistent-state fields such as intent and accumulated memory, traverses an adaptive semantic index. At each anchor boundary the substrate performs a single three-in-one traversal step that fuses three operations that conventional stacks keep separate: search narrowing, semantic state update (inference), and execution admissibility evaluation. Search, inference, and governance are not sequential layers bolted together; they occur as one governed step at each boundary.
Governance is a constituent phase of that step, not a post-hoc filter. In the filing's terms, the model proposes and the substrate decides: a proposed transition is admitted, rejected, or decomposed by the substrate before it takes effect, and the admissibility evaluation carries bounded per-step overhead by construction. The substrate also performs policy-enforced alias resolution and records the lineage of each admitted transition, so the traversal history is itself verifiable.
The filing's post-PageRank relevance model follows directly. PageRank and its successors compute a query-independent global authority score from corpus link structure, which is manipulable through link schemes and does not compose with governance. In the filing, relevance is not a precomputed score. A semantic object is relevant to a query if and only if the three-in-one step admitted every transition on the governed path that reached it. As the specification states, "Relevance is not a score; it is an admissibility-verified traversal history." Because the path is initialized from the specific query and evolves through query-specific interactions at each anchor, relevance is inherently query-specific rather than a global ranking.
The Level Distinction, Made Concrete
Placing the two side by side clarifies what a governance substrate provides that a behavioral standard, by its nature, does not.
A behavioral standard constrains how independent participants act at their interfaces. Compliance is a property of each participant. If a participant is compliant, interoperation follows; if enforcement of a property must hold regardless of any participant's discretion, a behavioral standard cannot supply it, because it has no substrate of its own on which to enforce anything. This is not a criticism of DULT. It is the definition of what a behavioral standard is.
A governance substrate, in the filing's sense, constrains every operation that flows through it. The admit, reject, or decompose decision is made by the substrate at each traversal step, so properties such as policy conformance, lineage continuity, and admissibility of the reasoning path are enforced structurally rather than delegated to the good behavior of the participants. The filing also describes multi-party governed negotiation, in which multiple agents and substrates participate in a governed resource-allocation process whose record is auditable. That is coordination enforced by the substrate, distinct from coordination achieved by convention among independent actors.
The distinction is orthogonal to domain. DULT lives in consumer location tracking; Semantic Discovery lives in agent cognition and knowledge traversal. What transfers between them is only the level-of-design lesson: some guarantees can be achieved by a behavioral standard, and some require a substrate that governs every step.
For an Implementer
A skilled implementer can build the governed-substrate approach the filing discloses from its described components. Represent the discovery object as a typed record carrying at least an intent field and an accumulated-memory field, and traverse an anchored semantic index. At each anchor boundary, implement a single step that (1) narrows the set of reachable transitions (search), (2) updates the object's semantic state from the anchor neighborhood (inference), and (3) evaluates each candidate transition against policy, lineage, and entropy constraints, emitting one of admit, reject, or decompose (governance). Bound the admissibility evaluation so per-step overhead stays within a fixed budget.
Record each admitted transition in a lineage log so that the object's final state is an admissibility-verified reasoning chain rather than an unverified output, and derive relevance from that recorded traversal history rather than from a precomputed global score. Apply policy-enforced alias resolution during narrowing so that alias targets are governed on the same footing as direct targets. Reasonable variations include: alternative admissibility criteria (temporal validity, source trust, cross-domain coherence); decomposition that expands a rejected transition into a governed subtraversal; single-object and multi-object coordination, where multiple discovery objects or substrates negotiate a governed allocation with an auditable record; and substitution of the underlying index structure so long as anchors expose neighborhoods over which the three-in-one step can operate.
Disclosure Scope
The inventive subject matter referenced in this article, the governed semantic discovery substrate with its three-in-one traversal step, admit/reject/decompose governance as a constituent phase, policy-enforced alias resolution, admissibility-verified traversal history, and post-PageRank query-specific relevance, is disclosed in United States Patent Application 19/647,395. All statements about the invention are grounded in that filing.
References to IETF DULT (Detecting Unwanted Location Trackers), to the DULT Working Group, and to the vendors and civil-society participants associated with it are external context describing a real, independently developed standard. They are provided only to illustrate the distinction between a behavioral interoperability standard and an architectural governance substrate. Nothing here asserts that 19/647,395 claims consumer location-tracker detection, anti-stalking flows, or any DULT-defined behavior, and nothing here should be read as characterizing the internal architecture, roadmap, or intentions of any named organization beyond the public, widely understood scope of the DULT standardization effort.
