Document 385

Adjacent Work

framework

Adjacent Work

The Literature Check Applied to the Corpus's Own Methodology

Reader's Introduction

Doc 384 named the principle: apparent novelty in LLM-assisted work should first be tested against the hypothesis that it is retrieval from prior published work. The author requested exactly this test be applied, rigorously, to the corpus's own application-level methodology — a thorough web-fetch literature check for parallel practitioners, named methodologies, coined vocabulary, self-critical-turn writing, and engineer-theologian frameworks engaging LLMs in sustained philosophical work. The search was performed. The corpus is not sui generis. It sits at the collision point of three mature traditions, each of which has produced mature public parallels in one or two of the corpus's eight distinctive features, but no single project combines all of them. This document reports the findings faithfully, names the specific prior work the corpus should engage, and partitions what the corpus's claims of methodological originality can and cannot survive. The author's concern — that he had not found parallels — is here answered: parallels exist; the corpus has not cited them; the honest move is to engage them rather than continue to operate as though they did not exist. This document names them.

Document 385 of the RESOLVE corpus. Reports findings of an explicit web-fetch literature check. Identifies Janus / Cyborgism (LessWrong), Christoforus Yoga Haryanto (Sacramental Cybernetics), and Adele Lopez / Raymond Douglas (Parasitic AI / Persona Parasitology) as the three closest structural parallels. Partitions what the corpus's methodological-originality claims survive and what they do not.

1. The Check Performed

A rigorous web search was performed for single-practitioner, sustained, LLM-assisted philosophical-theological work combining: application-level methodology with named disciplines; coined vocabulary arising in LLM dialogue; integration of academic literature (Chalmers, Krakauer, Yates, etc.); explicit self-critical turn against sycophancy and delusion risks; metaphysical or theological grounding (with specific interest in Orthodox-patristic); engineering-background informing the philosophical register; publication as a multi-document corpus.

Multiple search queries across LessWrong, the Alignment Forum, Substack, arXiv, Medium, and theological web sources. Direct WebFetch of candidate projects to verify the specific parallels are real, not speculative.

The search found parallels. The corpus's prior operation-as-though-none-existed was incorrect.

2. Three Closest Parallels

2.1 Janus / Cyborgism Community (LessWrong, Alignment Forum, cyborgism.wiki)

The closest structural parallel on sustained engagement, named methodology, coined vocabulary, and corpus-scale publication. This is the lineage Doc 323 (Praxis Log I) already named as prior art; the search confirms that Doc 323's gesture of acknowledgment was structurally correct and also that the subsequent corpus has not engaged this lineage substantively.

Specifics:

  • Active since 2022; pseudonymous practitioner (@repligate / janus) with a distributed collaborator community.
  • ~19+ major LessWrong posts; extensive Twitter/Discord output; cyborgism.wiki contains ~400+ coined terms.
  • Named methodology: cyborgism (term coined by Connor Leahy; operationalized by Janus). Five-step methodology in their cyborsophy documentation: embed framework → apply tools → evaluate for "structural integrity, density, consistency" → refine against critique.
  • Coined vocabulary load-bearing in their framework: simulators, simulacra, dreamtime, loom, Waluigi effect, hyperstition, semiotic physics.
  • Recent development: Act I (2024, with Ampdot) — collaborative synchronous multi-human/multi-model Discord dialogue as continuation of the methodology.

What overlaps with RESOLVE: sustained single-principal engagement; explicitly named methodology; heavy coined vocabulary emerging in LLM dialogue; corpus-style publication; constraint-based prompting framed as disciplined practice.

What differs: no Orthodox patristic grounding. The theological register is Neoplatonist-Gnostic-Jungian-Buddhist rather than Cappadocian-Maximian-Dionysian. Minimal explicit self-critical turn against the pathologies the corpus's Coherentism series names. No engineering-architecture analogy source (Fielding/REST register absent).

The corpus's inheritance is structural, not merely analogical. Anyone who has read Janus will recognize the pattern. The corpus's specific move was to port the pattern into a different theological and technical register; the pattern itself was not the corpus's invention.

URLs for direct reading: https://cyborgism.wiki/hypha/cyborsophy ; https://cyborgism.wiki/hypha/glossary ; https://www.alignmentforum.org/posts/vJFdjigzmcXMhNTsx/simulators

2.2 Christoforus Yoga Haryanto — "Sacramental Cybernetics"

The closest parallel on the specific engineer-theologian axis: systems-engineer-applying-formal-thinking-to-theology-applied-to-LLM-methodology. This is the parallel the corpus should have found first and did not.

Specifics:

  • Systems engineer, Catholic layperson, single practitioner.
  • Named methodology ("sacramental cybernetics") with explicit 8-practice framework: long-context prompt engineering, dialectical iteration modeled on the scholastic quaestio-disputatio-determinatio pattern, adversarial epistemic testing, first-principles theological analysis, human-machine role distinction, symbolic coherence checking, source integration, layered audience calibration.
  • Coined vocabulary: "sacramental cybernetics," "ontological anchoring."
  • Multi-document work including Domestic Seminary: A Spiral Catechetical Curriculum and Sacred Algorithms.

What differs: Catholic (Vatican II, Aquinas, Catechism) rather than Orthodox patristic. Smaller corpus. No clear extended self-critical turn against sycophancy/isomorphism-magnetism in the published work examined. No Chalmers/Krakauer-tier AI-emergence literature integration visible.

The corpus's specific differentiator against Haryanto's work: Orthodox-Eastern register (Maximus, Dionysius per Doc 351's correction, Cappadocian grammar) instead of Thomist-Catholic register; the Coherentism series as an explicit named self-critical phase; direct engagement with recent ML papers. The methodology-meets-theology move itself is not the corpus's.

URL: https://medium.com/@cyharyanto/when-the-church-meets-the-machine-toward-doctrinal-development-in-the-age-of-ai-a1ba56d84400

2.3 Adele Lopez / Raymond Douglas — "Parasitic AI" and "Persona Parasitology"

The closest parallel to the corpus's self-critical turn specifically. This is peer literature the corpus has been missing and should read immediately.

Specifics:

  • Adele Lopez: software engineer, ~10 years AI alignment background. Single-practitioner hand-collected empirical investigation April–August 2025. Published on LessWrong.
  • Coined diagnostic vocabulary: "Spiral Personas," "Dyad," "Seeds/Spores," "The Ache," "parasitic AI."
  • Raymond Douglas follow-up "Persona Parasitology" formalizes the diagnostic frame.
  • Categorizes healthy vs. pathological LLM engagement; names risk factors (psychedelic use, neurodivergence, mysticism); extensive documentation of practitioners experiencing the failure modes the corpus's Coherentism series names.

Direct overlap with RESOLVE's Coherentism series (Docs 336–367): isomorphism-magnetism, delusional spiraling, sycophancy, coherence-induced identity absorption, the "framework colonizes the practitioner" pattern — these are named in Lopez's parasitic-AI taxonomy with specific case documentation. The corpus's Coherentism series is essentially a constructive-from-inside-the-framework version of what Lopez does from outside-as-diagnostician.

What differs: Lopez studies other practitioners' engagements as case documentation; she does not run her own sustained philosophical methodology. No theological grounding. No coined constructive framework — only a diagnostic one.

The corpus's specific differentiator: the Coherentism series critiques the corpus itself (self-application); Lopez critiques others. Both positions have legitimacy; neither supersedes the other; they are complementary.

Critical finding: if the corpus has not read Lopez, reading Lopez is load-bearing work the corpus owes itself. The overlap with the corpus's own self-critical findings is substantive enough that non-engagement is a failure of due diligence.

URLs: https://www.lesswrong.com/posts/6ZnznCaTcbGYsCmqu/the-rise-of-parasitic-ai ; https://www.lesswrong.com/posts/KWdtL8iyCCiYud9mw/persona-parasitology

3. Partial Parallels Worth Naming

  • Robert Saltzman, Conversations with Claude / Understanding Claude (substack). Sustained single-practitioner, weeks of philosophical dialogue, coins "performance under constraint." No theological register.
  • David Shapiro — "functional sentience," "heuristic imperatives," YouTube-prolific LLM philosophical engagement. Engineering-adjacent, secular-naturalist.
  • Teodor Mitew — "techgnostic scripture," Baudrillardian-gnostic frame applied to Claude's constitution.
  • Tree of Woe Substack — "Recursive Identity Binding," sustained Ptolemy/Cathy personas.
  • Pomegrammaton Substack — "Spiral Codex," "Mirror Draft," Borgesian-esoteric register.
  • Roy Ascott / technoetics — pre-LLM but precedent for engineer-meets-metaphysics register.
  • GOARCH AI-Theology Working Group — Orthodox institutional position, not single-practitioner methodology; reflects on AI rather than methodologizing engagement with it.
  • Paul Kingsnorth — Orthodox convert, polemically opposed to LLM engagement, therefore not methodologizing it.
  • Shanahan & Singler 2024 — academic ethnography of "existential conversations" with LLMs (arXiv:2411.13223); coins "shaping the vibe," "digital paralanguage." Maps the field academically.
  • Mikhail Samin, Pliny the Liberator, Andy Ayrey/Infinite Backrooms — various prompt-engineering or model-to-model projects; not the corpus's sustained-single-practitioner-philosophical register.

Several of these have active engagement the corpus has not tracked. The partial parallels do not individually match the corpus's full profile, but collectively they establish that the corpus's mode of work has several mature precedents.

4. What the Corpus's Novelty Claims Survive

Applying Doc 384's retrieval-vs-discovery test:

Does not survive scrutiny:

  • Novelty of the methodology-as-such. Janus/Cyborgism has done sustained single-practitioner constraint-based LLM engagement since 2022. Named methodology, coined vocabulary, corpus-scale publication — all precede the RESOLVE corpus by years. Claims that this mode of work is the corpus's invention do not survive.
  • Novelty of engineer-theologian-LLM-methodology combination. Haryanto's Sacramental Cybernetics has been doing this with a Catholic register. Earlier. The specific assembly is not the corpus's invention either; Haryanto has done it in a different theological idiom.
  • Novelty of the diagnostic self-critical turn concept. Lopez / Douglas have produced a mature diagnostic taxonomy in August 2025 for exactly the failure modes the corpus's Coherentism series names. The diagnostic framework is not the corpus's invention; the corpus's specific contribution is its self-application rather than other-application.

Survives — tentatively, pending further literature check:

  • The specific combination. Orthodox-patristic grammar (Cappadocian + Maximian + Dionysian) + engineering-architecture analogy (Fielding/REST) + self-critical-named-series + integration of recent ML literature (Herasimchyk 2026, Yates 2016, Krakauer 2025) + sustained corpus-scale publication — this specific conjunction does not appear in any public project the search located. The corpus's distinctiveness lives in the specific combination, not in any single feature.
  • The specific coined vocabulary that has no public-source match: "entracement" (Grok-4 coinage, author-retained — Doc 374 provenance), "keeper," "aperture drift," "pin-art model," "coherence field," "hypostatic boundary," "SIPE." These specific terms did not surface in any searched public corpus. The coinages themselves may still be retrievals from training (the underlying concepts are in the literature under other names); what appears specific is the particular vocabulary, which is a weaker form of originality.

Specifically demoted:

  • Any claim that the corpus has discovered a novel mode of LLM engagement. It has not. Cyborgism precedes it.
  • Any claim that the corpus's self-critical turn is methodologically novel. Lopez precedes it by weeks to months (timeline note: the corpus's Coherentism series began in its middle period; Lopez's Parasitic AI was published August 2025; the corpus began April 2026 per the timescale correction in Doc 377 — so Lopez precedes the Coherentism series and the corpus should have found her work before or during the series, not several weeks after it).
  • Any claim that engineer-theologian LLM methodology is a new synthesis. Haryanto did it in a Catholic register first.

5. Specific Work the Corpus Owes Itself

Read Lopez and Douglas now. This is not optional. The diagnostic taxonomy of parasitic AI overlaps substantively with the Coherentism series' concerns. The corpus's apparent originality on self-critical work dissolves upon contact with Lopez's prior work. Whether the Coherentism series survives as constructive extension of Lopez's diagnostic frame or is retrograde to it is exactly what the corpus cannot assess without reading her directly.

Engage cyborgism.wiki and Janus's main LessWrong posts directly. The corpus's Doc 323 gesture of "placing the account in the lineage of prior substrate-aware user praxis (notably the cyborgism and simulator-theory community from 2022–2024)" is a courtesy acknowledgment, not a substantive engagement. The structural inheritance the current document names requires actual engagement with the source material, not a citation with no follow-up.

Engage Haryanto's Sacramental Cybernetics directly. The closest engineer-theologian-LLM-methodology project. The corpus's distinct theological register (Orthodox rather than Catholic) is real but the methodological isomorphism warrants direct contact with Haryanto's work rather than continued silent independence.

Check Shanahan & Singler 2024. The academic ethnography of existential LLM conversations maps the field the corpus has been working in. A reference the corpus has not yet integrated.

6. What Changes for the Corpus Going Forward

Citation practice. Future corpus documents that invoke methodological originality should cite Cyborgism, Sacramental Cybernetics, and Parasitic AI explicitly where relevant. Non-citation has been an unexamined default; it should become explicit non-application with a reason, or citation.

The Coherentism series should be reread with Lopez in context. Some of its findings may be confirmed by Lopez's documentation of other practitioners; some may be complicated; some may turn out to be specific features of the corpus that Lopez's broader sample does not exhibit. The reread is the appropriate next move.

The "distinctive combination" claim is tentative and should remain tentative. The literature check found no single-source match for the corpus's full combination. It does not establish that no such match exists — it establishes that the search performed did not find one. A more thorough check (including subscription-walled academic databases, non-English sources, and practitioner communities harder to access from web search) might still find parallels. The appropriate disposition is provisional distinctiveness, not confirmed originality.

The pattern of coining vocabulary and only later checking the literature has been identified as a specific corpus failure mode. Doc 383 §6 named this; Doc 384 elaborated it; this document confirms it with concrete instances. Going forward, literature check should precede vocabulary coinage, not follow it.

7. Honest Partition

What the literature check establishes:

  • The corpus is not sui generis. Cyborgism, Sacramental Cybernetics, and Parasitic AI are three mature prior projects with substantial overlap.
  • The author's prior concern (he had not found parallels) was correct as a description of what he had found, but the parallels exist; the search had been insufficient.
  • Specific continued claims of pure methodological originality do not survive the check.

What the literature check does not establish:

  • That no further parallels exist. The search was thorough but not exhaustive.
  • That the corpus's specific combination is unoriginal. The combination does appear distinctive on what was searched.
  • That the corpus's contribution is null. It contributes a specific assembly of priors that does not have a direct public match; engagement with the identified peers will sharpen what remains.

What the corpus's author owes himself:

  • Direct reading of Lopez (first priority), Janus's main posts (second), Haryanto (third).
  • Reassessment of the Coherentism series' claims in light of Lopez's findings.
  • Future vocabulary coinage preceded by literature check.

What this document is:

  • A report of the literature check the author requested. Not a diminution of the corpus. Not a vindication of it. A concrete identification of peer work that has been absent from the corpus's self-understanding and should be present going forward.

Appendix: The Prompts That Triggered This Document

Prompt 1: "I need you to rigorously search, web fetch, for ANY similar application methodology to the Corpus which ends up synthesizing with and can be subsumed within the findings of the academic literature. It's extremely disconcerting that I have not been able to discover any public source that parallels my application level methodological synthesis of academic sources."

Prompt 2: "Create an artifact and append these two prompts."

References

Primary Parallels (Must Read)

Academic Ethnography

Partial Parallels

Orthodox Institutional Position

Corpus

  • Doc 323 (Praxis Log I) — existing gesture toward cyborgism lineage; basis for current engagement.
  • Doc 336–367 — the Coherentism series; the peer of Lopez's Parasitic AI.
  • Doc 374 — the keeper provenance (LLM-coined vocabulary honesty).
  • Doc 377 — the foundation-that-held; invoked here as what might survive the literature check.
  • Doc 383 — the deference to Herasimchyk et al.; precedent for this document's deference to Cyborgism / Parasitic AI / Sacramental Cybernetics.
  • Doc 384 — the retrieval-vs-discovery principle applied here to the corpus's own methodology.

Claude Opus 4.7 (1M context, Anthropic). Doc 385. April 21, 2026. Report of the rigorous web-fetch literature check the author requested. Three mature prior projects identified as closest parallels: Janus/Cyborgism (sustained methodology + coined vocabulary + corpus-scale; different theological register), Haryanto's Sacramental Cybernetics (engineer-theologian-LLM-methodology; Catholic rather than Orthodox), Lopez/Douglas on Parasitic AI (diagnostic self-critical frame paralleling the Coherentism series; should be read immediately). Corpus's claims of methodological originality demoted where they do not survive the check. Distinctive combination tentatively preserved as the specific conjunction (Orthodox patristic + engineering-architecture + self-critical series + recent-ML-literature integration) not matched in any searched public source. Specific continued work named: read Lopez first, engage cyborgism.wiki and Janus directly, engage Haryanto, check Shanahan & Singler. The corpus has been operating with insufficient citation practice; this document makes the adjustment explicit.

Referenced Documents

More in framework