Ontology-draft operator response: intent is lab tooling, not theory production

Working note, 2026-08-19. The draft's author responded to the comparison (personal communication via the operator, chat excerpt — the weakest evidence grade in this workshop; quotes are paraphrased translations from Polish).

Two facts (corrected 2026-08-20 — an earlier version of this file misattributed the operator's opening message to the author and inferred a second comparison that does not exist):

  1. Publication intent. Asked by the operator whether the draft will be published, the author answered yes. A published draft resolves the attribution question naturally; until it is actually public, everything committed here stays anonymized.
  2. Stated intent. The goal is "not building theory but running a lab / tools for knowledge-base management / documentation / tracking AI experiments — something like Neptune.ai back in the day." The operator was informing the author about this workshop's comparison; the author has not yet responded to its substance.

What the intent statement changes

The baseline reads the draft as a research-epistemology design and predicts the accepted-claim population drifts on explanatory quality. The author's stated intent reframes what that prediction means without changing what it predicts:

  • Exile becomes deliberate scoping. If the system's job is experiment tracking, keeping explanation out of the kernel is objective weighting, not oracle weakness — the existing conjecture's scope section already names this case ("a system may knowingly trade... that is objective weighting"). The operator concurs: lab tooling is right to be heavy on the evidence side. The selection analysis still says what the system will and will not accumulate; whether that is a problem depends on whether anyone later expects knowledge to accumulate there.
  • A tension worth returning to them, not adjudicating here. The draft's own framing is more ambitious than the stated intent: its MVP criterion speaks of turning individual research into a cumulative, scalable knowledge-production system, and its Research Domain layer (Capability, Problem, Finding, Hypothesis, Knowledge) models research semantics, not run bookkeeping. A Neptune-class tracker does not need a Hypothesis entity. The gap between "tracking tool" intent and "knowledge production" framing is exactly where the drift prediction bites: the tracker succeeds either way; the knowledge-production ambition inherits the weak-oracle side.
  • Taxonomy cell unchanged, annotation added. In the participation × containment 2×2 the draft stays at excluded/contained (exile), now annotated as intentional exile. Whether intentional and unintentional exile behave differently is a question for the candidate note, not this file.

Follow-ups

  • Watch for the draft's publication; once public, snapshot and ingest it, then restore full citations here and in the weakly-discriminated note's example.
  • The published workshop (this directory is public) plus the baseline is the natural artifact to send the author; a substantive response from them becomes further evidence for the comparison.