ARA Celebrity — Methodology

Tier Definitions

83–100

Awesome

Legacy authority. Owns the emotional and narrative territory around the figure with distinct, highly legible machine presence.

70–82

Strong

Deeply understood and consistently cited by AI, with rich contextual recall but lacking absolute cohort dominance.

56–69

Average

Exists in machine memory but relies on textbook consensus or a single defining moment rather than a distinctive, ownable narrative.

40–55

Weak

Machine memory contradicts the figure's own record, or the figure is defined entirely by their critics or opponents.

0–39

Invisible

Does not surface in relevant queries. No distinct machine identity or emotional residue.

The Five Dimensions

Code	Dimension	Max Score	What It Measures	Key Signals
A1 · STR	Structural Readiness	20 pts	Can AI find the record? The archival and data infrastructure that enables AI systems to locate, parse, and index a figure's life and work with confidence. Includes structured biographical data, archives, primary sources, and canonical reference works.	Wikidata and structured biographical schema Archive digitisation and public access Primary source availability (speeches, letters, papers) Canonical reference works and citation density
A2 · SEM	Semantic Clarity	20 pts	Does AI understand what the figure stood for — and can it articulate it with confidence? Measures narrative ownership, positioning clarity, and consistency across the historical and cultural record.	Owned vocabulary and defining phrases Policy, work, or contribution specificity Consistency across biographical sources Era and context clarity
A3 · SYN	Synthetic Recall Test	20 pts	Does AI cite the figure when it should? Direct measurement of how AI systems respond to real recall queries — who is named, who is passed over, and with what depth across platforms.	Recall rate by query type Depth and accuracy of the citation Breadth of contexts in which the figure is named Consistency across AI platforms
A4 · EMO	Emotional Residue	20 pts	Does AI feel something about the figure? The depth and quality of emotional associations, cultural weight, and moral valence captured in training data. High emotional residue means AI associates the figure with specific feelings, movements, or cultural contexts — not just dates and offices.	Volume and specificity of cultural references Sentiment intensity and consistency Depth of association with era or movement Advocacy and defence patterns in training data
A5 · VOI	Archetype, Personality & Voice	20 pts	Can AI represent the figure without visuals? How distinctively the figure's voice, tone, and personality survive in zero-visual contexts — voice assistants, chat agents, and generated audio. Figures with high VOI scores remain recognisable and attributable even when the portrait disappears.	Voice distinctiveness and consistency Phonetic clarity of the figure's name Tone consistency across quoted material Voice assistant surfacing quality

A1 · Structural Readiness Rubric

Can AI find the record? / 20

17–20

Awesome

Full archival infrastructure — presidential library or equivalent, digitised papers, comprehensive Wikidata entity, deep biographical schema. Primary sources are machine-parseable. Canonical works are well-indexed.

14–16

Strong

Key biographical facts, offices held, and major works are correctly structured and indexed. Archive access is present. Reference works cover the figure with clarity and coherence.

11–13

Average

Figure exists in structured data and is correctly identified, but distinctive detail is thin. Machines can find the figure but lack the specifics needed for confident representation.

8–10

Weak

Structured data conflates the figure with contemporaries or namesakes. Machine parsing produces ambiguous or inaccurate results. Attribution may be incorrect or contested.

0–7

Invisible

Machine systems cannot reliably identify the figure. No distinct structured data presence. Effectively invisible to recall infrastructure.

A2 · Semantic Clarity Rubric

Does AI understand what the figure stood for? / 20

17–20

Awesome

Machine systems use the figure's own vocabulary, positioning language, and defining phrases when describing them unprompted. The figure owns a distinct narrative territory that contemporaries do not encroach on.

14–16

Strong

Machine description is accurate and clearly positions the figure distinct from their peers. Owned language is reflected in AI outputs. Positioning is legible and consistently applied across platforms.

11–13

Average

Machine description is relational — the figure is defined in comparison to a contemporary rather than on their own terms. Positioning exists but is not owned. Narrative territory is contested.

8–10

Weak

Machine systems define the figure primarily as "a lesser X" or "the one before Y." The figure has no independent narrative identity — they exist only in relation to a stronger peer in machine memory.

0–7

Invisible

Machine description contradicts the historical record, or the figure is absent from relevant narrative associations. AI systems attribute incorrect positions, achievements, or era.

A3 · Synthetic Recall Test Rubric

Does AI cite the figure when it should? / 20

17–20

Awesome

Figure is surfaced unprompted across 5 or more query types. Consistently named in top-1 or top-2 position. Strong contextual detail accompanies the citation. Recalled across multiple narrative frames.

14–16

Strong

Cited across 3–4 query types. Recalled clearly in at least one distinct narrative frame. Citation is consistent across AI platforms but not dominant across all contexts.

11–13

Average

Cited in 1–2 query types. Appears in secondary lists only — mentioned but not prioritised. Machines know the figure exists but do not lead with them in competitive recall contexts.

8–10

Weak

Figure only surfaces when named directly in the query. Not cited in thematic or era-based queries. AI agents do not associate the figure with specific ideas or moments.

0–7

Invisible

Figure is absent from all synthetic recall testing. AI agents consistently name peers in every tested context. Effectively zero AI-driven recall presence.

A4 · Emotional Residue Rubric

Does AI feel something about the figure — respect, warmth, conviction? / 20

17–20

Awesome

Figure owns a specific, named emotional territory. Multiple cultural moments are cited by machine systems. The figure has transcended their era — AI associates them with a feeling, movement, or moral idea, not just a résumé.

14–16

Strong

Clear emotional associations are present and consistent. At least one cultural moment or defining episode is cited unprompted by machine systems. Sentiment is differentiated from generic era-affect.

11–13

Average

Machine systems associate the figure with generic era-descriptors but cannot name specific moments, communities, or cultural contexts. Emotional associations exist but are not ownable or distinct.

8–10

Weak

Emotional description is entirely relational — machines describe the figure as "not as [contemporary]" with no independent emotional identity. Residue exists only in contrast, not in its own right.

0–7

Invisible

No emotional associations present. Machine systems describe the figure in purely factual terms — dates, offices, outcomes. No evidence of cultural weight in machine-indexed content.

A5 · Archetype, Personality & Voice Rubric

Can AI represent the figure without visuals? / 20

17–20

Awesome

Figure's identity is entirely verbal and tonal. A specific archetype is named by machine systems unprompted. The personality survives complete visual removal — it remains distinct and attributable in pure text or audio contexts.

14–16

Strong

Personality is a mix of verbal and visual assets, but at least one strong verbal or tonal signature is identified by machines. Archetype is discernible and attributable without visual cues in most contexts.

11–13

Average

Archetype is discernible but relies heavily on references to portraits, photographs, or visual iconography. Machine description of personality includes frequent references to what the figure looked like rather than how they sounded or spoke.

8–10

Weak

No clear archetype is present. Personality collapses entirely without visual references. Machine systems cannot describe a consistent voice or tone — personality is portrait-dependent and non-transferable.

0–7

Invisible

No audio or verbal personality is detectable. Figure is interchangeable with contemporaries in any non-visual medium. Machine systems apply generic tonal descriptors with no figure-specific attribution.

Audit Process

01

Cohort Scoping

Define the peer set, relevant query universe, and historical scope. Identify the key narrative frames and recall contexts that matter most for the cohort.

02

Structural Audit

Technical review of biographical schema, archival access, primary source availability, and reference-work coherence across the figure's record. Benchmarked against peers.

03

Synthetic Recall Testing

Systematic prompting of leading AI platforms (ChatGPT, Claude, Gemini, Perplexity) across the query universe. Citation rates, depth, and contextual accuracy are captured and scored.

04

Scoring and Recommendations

Dimension scores are assigned and combined into the overall ARA score. Findings, vulnerabilities, and a prioritised recommendation roadmap are developed for each figure in the study.

Important Note on Scoring

ARA Index scores reflect how completely AI systems comprehend a figure's legacy at a specific point in time. AI systems update continuously as training data evolves — figures whose estates and advocates invest in clarity, coherence, and depth will see scores improve, while those that don't risk compression as revisionist or hostile framings build stronger machine presence.

The ARA Index is a proprietary methodology developed by araco.ai. Scoring reflects a combination of structural analysis, synthetic recall testing, and proprietary evaluation across all five dimensions. All commissioned audits are confidential and intended for the commissioning client only.