Editorial notes
Classification rationale, evidence status, source limitations.
Classification rationale
A historical text-generation and portrait-animation experiment in AI representations of Jesus, distinct from later conversational systems and the Lucerne installation. Generated religious-style language materially shapes the documented output; the record does not establish scripture adoption or autonomous religious authority.
Evidence status
A surviving creator-maintained repository and dated creator accounts document the released texts and presentation. Contemporary reporting, a critical sermon and a later comparative study document attributed reception without independently verifying training or scripture adoption.
Classification confidence: Moderate
Source limitations
Training details come from Durendal; the repository contains no training code or model weights. Contemporary coverage often repeats his account, and the later study analyzes published documents. The release dates do not establish an operating lifespan, and the cited reception does not establish adoption of the texts as scripture.
Evidence profile metadata
base_status | Reviewed candidate |
|---|---|
review_flags | Prompt-context-hidden, Participation-unclear, Synthetic-media-caution |
provenance_strength | Preserved artifact |
artifact_verifiability | User-supplied |
prompt_visibility | Partial |
context_visibility | Partial |
external_coverage_quality | Moderate |
external_coverage_weight | 3 |
independent_sources_count | 4 |
source_relationship_risk | None |
The pinned creator-maintained GitHub revision preserves six published text artifacts and repository history; this supports artifact identity and publication, not reproduction or independent authentication of model generation. References 1, 2 and 4 belong to one creator lineage. Four external reporting/reception families are counted: The Christian Post’s original email interview, TechXplore’s critical reading, the United Church of Hinesburg’s critical sermon, and Çiftçi and Yeşilyurt’s later scholarly interpretation. This is not a count of independent confirmations of training, model behavior, religious adoption or technical success. References 3, 5 and 7 rely on the creator lineage for development details; the sermon is primary evidence of its own critical reception, not adoption. Coverage is moderate, weight 3, based on substance rather than URL count. The disclosed dependencies are resolved by attribution and do not establish unresolved circular validation. Topic prompts and selected generation/presentation details are described, but complete model, training code, weights and generation context are unavailable. Human narration and portrait selection are explicit. Commit dates describe the surviving repository record, not an operating lifespan.
Edition status
| en | published |
|---|---|
| ai | published |
Related entries
- related case: Deus in Machina (AI Jesus)
- related concept: Religious chatbots and scripture interfaces
- related concept: Human–AI coauthorship
Redirects and former locations
- None recorded.