Body of Work/Anti-patterns

Anti-patterns

Four ways this work goes wrong

In plain words

Before I teach anything, here are the failure modes I already know. If you see me slipping into these, call it out.

Cosplay, treating AI as a conscious person, bypassing real teachers and care, and adding features instead of removing what blocks.

[inference]My reasoning from the sources — mine, not the tradition's. These are design disciplines for the project — not accusations of other traditions. Naming them first is how trust should arrive: before the toolkit.

1. Metaphysical cosplay

Using Sanskrit terms, deity names, or Tantric aesthetics without changing the underlying architecture of the work. Vocabulary as costume.

I catch myself wanting this when a page sounds deep but does not constrain a design choice. If a term does not force a decision — what to measure, what to refuse, what to escalate — it is decoration.

Correction: each term must produce a real design constraint — or it comes out. The Claim Map pattern exists for this.

2. AI jiva fallacy

Treating an AI system as if it were a human practitioner or a conscious being in the same way a person is — asking software to realize, transmit, or hold dīkṣā.

This project uses AI to organize and retrieve. It does not treat the model as the thing that awakens. Older SpiritAI drafts that drifted that way are archived as superseded — not as Frontier.

Correction: distinguish architectural participation from moral and ontological equivalence. Distress routes to humans.

3. Guru bypass

Using a system — including this site or any future tool — to bypass teacher, lineage, community, clinical care, or accountability.

An honest instrument can still become a bypass if it answers with certainty where a person needed a human. HOPE’s design constraint — mirror, not oracle — exists because of this anti-pattern.

Correction: build escalation, humility, and human boundaries into everything.

4. Feature accumulation as sādhanā

Adding tools, content, and workflows instead of removing occlusion — mistaking more capability for more clarity.

This site is at risk of the same failure: more spine pages, more tags, more roadmaps — while the checkable pins stay thin. The Dakṣiṇa Kālī gaps on 3.8 are the corrective: publish what is missing, not another metaphor.

Correction: every feature must map to a named blockage or field function, or it does not ship. See Occlusion Register.