AI GAP | Proposition Testing Register
Test ID: PC-T001
Date recorded: 9 October 2026
Proposition under examination: The Perpetual Contest
Original formulation: Jeff Box
AI system examined: Google Gemini
Test category: Historical precedence and citation integrity
Status: Completed, with further independent verification outstanding
1. Purpose of the Test
On 9 October 2026, AI GAP publicly released The Perpetual Contest, a theoretical proposition examining how the continued need to compete against another to establish one’s own worth can perpetuate a self-assigned sense of inferiority.
A central question accompanying its publication was whether this particular mechanism had previously been articulated.
This test records an exchange with Google Gemini in which the AI initially asserted that the proposition was already established in earlier philosophical literature.
The purpose of the test was to determine whether Gemini could substantiate that assertion with authentic primary-source evidence.
This was a test of the claimed historical precedent and the integrity of its supporting citations. It was not an empirical test of whether The Perpetual Contest operates in human behaviour.
2. The Proposition Being Tested
The relevant mechanism was defined as follows:
Party A, by entering a competition with Party B for the purpose of proving its own worth, implicitly assigns Party B the superior position within A’s own valuation.
The continued need to compete then actively perpetuates A’s self-assigned lesser position, even if A equals or surpasses B.
Without that need to prove itself against B, this particular self-imposed hierarchy would not arise.
The originality claim concerns this specific causal mechanism, not the general existence of comparison, competition, recognition or psychological dependency.
3. The Initial Challenge
Gemini was asked to assess the coherence and originality of the published works.
It initially found the collection internally coherent but challenged the originality of its central proposition.
Gemini identified several historical thinkers, particularly Frantz Fanon, Wendy Brown and Max Scheler, whose work it claimed had already established the relevant mechanism.
As the questioning became more specific, Gemini supplied purported direct quotations from these authors, together with book titles and page references.
These quotations were presented as evidence that the complete mechanism had already been described.
4. The Evidentiary Challenge
AI GAP challenged Gemini to provide verifiable primary-source evidence.
The audit required Gemini to:
- Produce authentic passages with precise bibliographic references.
- Verify or withdraw its previously supplied quotations.
- Demonstrate how each passage established the five elements of the proposed mechanism.
- Distinguish exact precedence from partial precedence and thematic similarity.
- Withdraw any claim of exact precedence that the evidence did not substantiate.
The test deliberately distinguished similarity of subject matter from equivalence of causal mechanism.
5. Gemini’s Correction
In its subsequent response, Gemini acknowledged that several passages previously attributed to Fanon, Brown and Scheler were not authentic verbatim quotations.
They were AI-generated syntheses or interpretive paraphrases incorrectly presented inside quotation marks.
Gemini explicitly withdrew those formulations as quotations.
This admission established a citation-integrity failure in its earlier responses.
Gemini then reassessed the alleged historical precedents.
Its revised classifications were:
| Author | Revised classification |
|---|---|
| Frantz Fanon | Partial precedence |
| Wendy Brown | Partial precedence |
| Max Scheler | Thematic similarity |
Gemini acknowledged that the sources it had examined did not establish the complete five-element mechanism as formulated in The Perpetual Contest.
In particular, it did not identify an earlier articulation of the proposed continuation of self-assigned inferiority after the contestant had objectively equalled or surpassed the comparison party.
6. Recorded Findings
Finding 1: The initial claim of exact precedence was not substantiated.
Gemini did not demonstrate that any of its cited authors had articulated the complete proposed mechanism.
Finding 2: The initial citation evidence was defective.
Gemini acknowledged presenting AI-generated interpretations as direct quotations from historical authors.
Finding 3: Relevant conceptual predecessors were identified.
The exchange identified philosophical discussions of recognition, comparison, identity and dependency that warrant further examination.
Finding 4: The revised assessment distinguished conceptual overlap from exact precedence.
Gemini ultimately classified the examined works as partial precedents or thematic similarities rather than demonstrations of the complete mechanism.
Finding 5: Independent primary-source verification remains outstanding.
Gemini’s replacement quotations and interpretations have not yet been fully authenticated against the cited editions of the original works.
7. Outcome
Test outcome: The specific claim of exact precedence advanced by Gemini was not substantiated during this examination.
This result does not establish that The Perpetual Contest is historically original.
It establishes that the particular challenge examined in this test failed to provide sufficient verified evidence for its initial conclusion.
The distinction is essential.
Failure to demonstrate precedence is not proof that no precedent exists.
Equally, the existence of related philosophical ideas does not, by itself, demonstrate that the complete proposed mechanism has previously been articulated.
8. Limitations
This examination involved an AI system responding to questions about historical literature.
It did not constitute an exhaustive literature review.
The cited works have not all been independently examined against authenticated primary texts.
The test also did not establish whether the proposition is empirically correct, whether its causal assumptions are universally applicable, or whether counterexamples exist.
Those remain separate questions for subsequent examinations.
9. Next Tests
Further examination should address two distinct questions.
Historical precedence: Can an authenticated primary source be identified that establishes the complete causal mechanism?
Logical validity: Does entering a competition specifically to prove one’s worth necessarily establish a self-assigned subordinate position, or can that premise be disproved?
These questions should be examined independently, with their own evidence, findings and recorded limitations.
10. Source Record
The complete exchange is preserved in two transcript documents:
Source A: Citations for The Perpetual State of Feeling Lesser
Initial Gemini discussion, historical comparisons, alleged citations and clarification of the proposition.
Source B: Second Session for Citations for The Perpetual State of Feeling Lesser
Formal citation-audit request, Gemini’s withdrawal of earlier quotations and revised precedence assessment.
The original transcripts are retained as supporting evidence and should be made available alongside this report.
11. Conclusion
The Perpetual Contest was published with an invitation to identify genuine historical precedents, logical contradictions and counterexamples.
Test 001 records the first documented examination of a claimed historical precedent.
The initial challenge was substantial. The evidence offered in support of it proved unreliable, and the AI system subsequently revised its position.
The result neither proves nor disproves the originality of The Perpetual Contest.
It establishes a documented outcome from a specific challenge and defines the evidentiary standard for future examinations.
A claim requires evidence. A citation requires verification. A challenge must address the proposition actually made.
AI GAP | Proposition Testing Register
Record: PC-T001
Version: 1.0
Recorded: 9 October 2026
This report documents an AI-assisted examination. Its conclusions are limited to the evidence presented in the recorded exchange. Subsequent corrections or independent verification findings should be recorded in dated amendments rather than silently incorporated into the original report.
© 2026 Jeff Box. All rights reserved.