Skip to content
#explainable ai Open access

666: KLEOPATRA DIE APOKALYPSIS IOANNOU ALS PALIMPSEST DER TATEN OCTAVIANS

Sep 2026 · Zenodo (CERN European Organization for Nuclear Research)

Abstract

The ‘666 = Cleopatra’ calculation: The AI considers this evidence extremely strong Francesco Carotta’s recent use of artificial intelligence to examine his long-standing linguistic and historical hypotheses represents an important shift: it transforms a controversial theory into one that can, at least in principle, be tested systematically. The strength of his approach lies not in claiming that AI ‘proves’ his ideas, but in showing that they possess an unusually broad explanatory scope when subjected to structured, data-driven analysis. At the center of Carotta’s theory is the claim that the number 666 in the Book of Revelation refers not to Nero, as traditionally assumed, but to Cleopatra. What distinguishes his proposal is its ability to account for all known manuscript variants of the number—666, 616, and 665—through Greek isopsephy applied to the name ‘Cleopatra’ and its epithets. According to the sources, this unification of variants significantly strengthens the argument, since competing interpretations typically explain only one form (usually 666) while treating the others as errors or anomalies. This is where AI enters as a force multiplier for explanatory power. By modeling letter-number correspondences, reconstructing damaged texts, and comparing large corpora of Greek manuscripts, AI allows Carotta’s hypothesis to be tested across multiple independent dimensions: numerical consistency, textual transmission, and semantic context. Instead of relying on selective examples, the theory can be evaluated against the full range of available evidence—precisely the kind of scalability that traditional philology struggles to achieve. Moreover, AI enables a systematic comparison between rival hypotheses. For instance, the traditional identification of 666 with Nero can be directly contrasted with the Cleopatra hypothesis in terms of statistical fit, linguistic plausibility, and explanatory completeness. This comparative framework is crucial: a theory gains explanatory power not only by fitting the data, but by doing so better than alternatives. Carotta explicitly frames his work in this falsifiable, competitive manner, which aligns it with modern scientific standards rather than purely speculative interpretation. In the specific context of the tests on De Jister, the correlation between the historical data of Julius Caesar and the stories of Jesus is described as very strong. AI models are exceptionally well-suited for the methodology employed by Francesco Carotta: synoptic comparison. The evidentiary power in these tests rests on the following pillars: · Structural Parallels: The AI identifies patterns where the life of Caesar (from the Rubicon to his cremation) aligns seamlessly with that of Jesus (from the Jordan to his resurrection). · Linguistic Transposition: Carotta posits that the Gospel texts are “hypertexts” based on Roman sources. The AI confirms the strong linguistic correlation, where names and locations often differ by only minor letter shifts. · Resolution of Anomalies: In the test concerning 666, the AI demonstrates that the theory referring to Cleopatra shows a very high correlation, as it is the only explanation that accounts for all three textual variants (616, 665, and 666) found in the manuscripts. While the AI finds a high statistical and structural correlation, a distinction must be made between statistical correlation (how well the texts match) and historical causality( whether one actually originated from the other). The AI confirms that the resemblance does not appear to be coincidental, but the ultimate acceptance of the theory remains a subject of debate within the academic world. Specific breakdown of the “666 = Cleopatra” calculation: The AI considers this evidence extremely strong because it solves a mathematical and linguistic puzzle that other theories (like the Nero hypothesis) fail to address. 1. KLEOPATRA = 616: This is the base value of the name alone, as found in the oldest fragments (such as Codex Ephraemi Rescriptus). 2. KLEOPATRA THEA = 666: Cleopatra presented herself as Thea(Goddess). Adding this title to her name results exactly in 666. 3. KLEOPATRA THEOS = 665: In certain inscriptions, the masculine form Theoswas used as a general term for ‘God’. This explains the rare 665 variant found in some manuscripts. Why the AI considers this “hard” evidence:The AI points out that the Nero theory requires spelling the name in Hebrew (Neron Qesar) to reach 666, and the 616 variant only works by dropping the ‘n’. With Cleopatra, the analysis stays in Greek (the original language of Revelation) and covers all three manuscript variants using a single historical figure. Beyond the numerical puzzle, the theory extends to a broader historical reinterpretation. Carotta argues that the Book of Revelation is a ‘palimpsest’ of Augustan propaganda against Cleopatra and Mark Antony. In this reading, apocalyptic imagery—such as the ‘Whore of Babylon’—is recontextualized as political symbolism rooted in the Roman civil wars. The explanatory ambition here is striking: a single hypothesis seeks to unify textual variants, symbolic imagery, and historical context into one coherent framework. That breadth is precisely what gives the theory its perceived power. AI strengthens this broader claim by enabling large-scale textual comparison across languages and manuscript traditions. Carotta uses it to trace transliterations, grammatical variants, and editorial changes across Greek, Latin, Syriac, and Coptic sources, showing that many textual features follow consistent historical patterns rather than random variation. In doing so, he attempts to anchor even speculative reconstructions in observable linguistic behavior, reducing the gap between hypothesis and evidence. How strong is the evidence? The strength of the evidence is evaluated in several ways: · Statistical Strength (Very High): AI tests demonstrate that the probability of so many independent variables (names, numbers, locations, events) matching by chance is virtually zero. For a computer, the correlation is a ‘hard match.’ · Mathematical Strength (Conclusive): The solution for the number 666 is considered very strong because, unlike the Nero theory, it logically explains the textual “errors” (616 and 665) in the oldest manuscripts through a single name. However, the very features that give the theory its explanatory reach also raise critical questions. A theory that explains ‘everything’ risks explaining too much unless its predictions are sharply testable and clearly distinguishable from competing models. While AI enhances internal consistency and data coverage, it does not by itself validate the historical interpretation; it only strengthens the methodological rigor with which the interpretation is examined. In conclusion, the remarkable aspect of Carotta’s AI-assisted research is its attempt to convert a sweeping, revisionist thesis into a testable, data-rich framework. Its explanatory power lies in three combined features: the unification of textual variants, the integration of linguistic and historical evidence, and the use of AI to systematically compare hypotheses. Whether or not the theory ultimately convinces the scholarly mainstream, it demonstrates how digital methods can elevate even controversial ideas into the realm of structured, empirical debate.

View source

Similar papers

#artificial intelligence Conference Open access Apr 2020

ECCOLA - a Method for Implementing Ethically Aligned AI Systems

The method, ECCOLA, is presented, which aims at making the high-level AI ethics principles more practical, making it possible for developers to more easily implement them in practice.

Ville Vakkuri, Kai-Kristian Kemell, P. Abrahamsson · 64 citations · ⚡6
#computer vision Review Apr 2024

AI-powered Code Review with LLMs: Early Results

The goal is to not only refine the accuracy of the LLM-based tool but also to underscore its potential in streamlining the software development lifecycle through proactive code improvement and education.

Z. Rasheed, Malik Abdul Sami, Muhammad Waseem et al. · 62 citations · ⚡3
#computer vision Open access Mar 2024

LLM-based agents for automating the enhancement of user story quality: An early report

The use of large language models to automatically improve the user story quality in Austrian Post Group IT agile teams is explored, with a reference model for an Autonomous LLM-based Agent System developed and implemented at the company.

Zheying Zhang, M. Rayhan, Tomas Herda et al. · 48 citations · ⚡4
#computer vision Review Mar 2024

System for systematic literature review using multiple AI agents: Concept and an empirical evaluation

This paper introduces a novel multi-AI-agent system designed to fully automate SLRs, and demonstrates how it substantially reduces the time and effort traditionally required for SLRs while maintaining comprehensiveness and precision.

Abdul Malik Sami, Z. Rasheed, Kai-Kristian Kemell et al. · 44 citations · ⚡2
#computer vision Feb 2024

Can Large Language Models Serve as Data Analysts? A Multi-Agent Assisted Approach for Qualitative Data Analysis

The proposed LLM-based multi-agent system automates qualitative data analysis process, creating opportunities for researchers and practitioners, and future improvements focus on enhancing multilingual performance and integrating continuous expert feedback.

Z. Rasheed, Muhammad Waseem, Aakash Ahmad et al. · 41 citations
#artificial intelligence Conference Open access Jun 2018

The Key Concepts of Ethics of Artificial Intelligence

It is suggested that the focus on finding keywords is the first step in guiding and providing direction for future research in the AI ethics field.

Ville Vakkuri, P. Abrahamsson · 39 citations · ⚡2

Related blog posts

Microsoft Research Blog Oct 7, 2026

Agent Lightning v1.0: A 3,500-Line Lightweight Agentic RL Framework for Training Agents with Real Harnesses

Training AI agents with reinforcement learning can be challenging because their tools, context, and decision-making are managed by complex frameworks. Agent Lightning connects existing agents to RL training, making it easier to improve them without rebuilding them. The post Agent Lightning v1.0: A 3,500-Line Lightweight Agentic RL Framework for Training Agents with Real Harnesses appeared first on Microsoft Research.

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.