Letter to the Editor: Artificial wisdom and the transience of truth: Ethical and temporal reflections on artificial intelligence-human inquiry into history

Abstract

Zhou et al published a study in the recent issue of the World Journal of Gastroenterology, which aimed to examines the concept of artificial wisdom (AW) as applied to historical medical inquiry, in response to a recent artificial intelligence (AI)-human analysis of Alexander the Great’s cause of death. While acknowledging the innovative integration of generative AI with clinical reasoning, the letter highlights key epistemic and ethical limitations. It argues that AI systems, inherently shaped by transient data and iterative obsolescence, cannot access enduring or timeless truths, particularly in historically remote contexts where evidence is fragmentary and context-dependent. The persuasive fluency of large language models risks creating an illusion of certainty, conflating probabilistic synthesis with wisdom. Drawing on principles of ethical AI use, the letter emphasizes transparency, accountability, and human moral stewardship as essential safeguards. Ultimately, it proposes a shift from the notion of AW toward epistemic stewardship, recognizing that truth evolves with time and that wisdom resides not in algorithms, but in ethically grounded human judgment.

Key Words: Wisdom; Alexander the great; Intellectual audacity; ChatGPT; Ethics

Core Tip: This letter challenges the concept of artificial wisdom (AW) by emphasizing that truth is transient, not eternal, and therefore cannot be captured by algorithms that age with time. It argues that artificial intelligence can simulate reasoning but not embody wisdom, which requires ethical constancy, humility, and temporal awareness. True progress lies not in AW, but in responsible epistemic stewardship that honors both ethics and impermanence.

TO THE EDITOR

I read with great interest the article by Zhou et al[1] published in the recent issue of the World Journal of Gastroenterology. The authors’ intellectual audacity deserves appreciation-they invite artificial intelligence (AI) into the realm of historiographic medicine and attempt to fuse it with human clinical discernment. Yet, this ambitious integration, while imaginative, also reveals epistemic and ethical fissures that merit deeper consideration before artificial wisdom (AW) can be accepted as a credible path to knowledge.

AW or algorithmic persuasion?

The authors define AW as a synthesis between AI’s computational breadth and human moral reasoning. In practice, however, their model seems more curatorial than co-creative: ChatGPT generates a hypothesis, and human reviewers correct its excesses. This inversion of intellectual hierarchy transforms human oversight into post-hoc editing rather than shared reasoning. Wisdom, as Aristotle conceived it, is phronesis-context-sensitive moral judgment born of experience. By contrast, ChatGPT’s probabilistic fluency produces semblance rather than substance: Eloquent but unanchored. What emerges is not wisdom, but algorithmic persuasion-the appearance of thought without the weight of understanding.

The mirage of methodological rigor

While the study adopts the language of scientific inquiry, its design remains ontologically unstable. Prompts, bias control, and reproducibility parameters are not disclosed in detail, making independent verification impossible. The authors quantify “source reliability” using numerical scoring, yet citation fidelity cannot substitute for epistemic integrity. Their own Table 2 reveals fabricated and misapplied references, some entirely non-existent. Such hallucinations, when unaccompanied by ethical commentary, risk legitimizing synthetic scholarship. A truly wise methodology would not only detect errors but confront the moral implications of machines manufacturing falsehoods with human endorsement.

Ethics: The missing pillar of AW

In our previously published work on “Ethical use of AI in infectious diagnostic decision and therapeutic stewardship”[2], we proposed four essential pillars for ethical AI integration: Transparency, accountability, contextual judgment, and continuous human oversight. Zhou et al[1] mention these concepts but do not operationalize them. By presenting ChatGPT as a co-investigator rather than a probabilistic tool, the study risks anthropomorphizing computation and diminishing human accountability. No ethical audit or governance structure for AI use in scholarly reasoning is described. Absent such safeguards, AW risks collapsing into epistemic populism-where fluency masquerades as truth.

History, medicine, and the transience of truth

The most striking limitation, however, is philosophical: The study treats time as transparent rather than transformative. To assign a modern microbial etiology to a death in 323 B.C. is to presume that truth is timeless and transferable, uncorroded by context. Yet truth, like tissue, decays with time. Historical evidence is perishable; so too are the algorithms we train upon contemporary data. AI, a technology defined by iterative obsolescence, cannot apprehend eternity-it mirrors only the transient truths of its training epoch. To ask a generative model to diagnose Alexander the Great is to ask the mirror to remember the face that once looked into it.

AW, if it is to exist at all, must therefore be anchored not in temporal data but in ethical constancy-the human capacity to recognize that truth is dynamic, partial, and revisable[3]. Time erodes certainty; humility preserves meaning. By neglecting this temporal dimension, the study risks presenting the speculative as stable, and the provisional as permanent.

Attribution and accountability in the age of AI

The authors commendably acknowledge ChatGPT’s role, yet the extent of AI contribution to content generation remains undefined. In the post-truth digital era, disclosure of AI participation is an ethical necessity. The World Association of Medical Editors[4] has issued clear guidance: AI cannot be an author, but its involvement must be transparently declared. Without explicit attribution, readers cannot discern where human reasoning ends and machine synthesis begins. Transparency is not merely bureaucratic-it is moral.

From AW to epistemic stewardship

The authors’ aspiration to model AW could evolve into something more robust: Epistemic stewardship. In this framework, AI serves as a heuristic companion, while humans retain custodianship of interpretation and ethics. The goal is not to manufacture certainty but to curate understanding responsibly. A practical pathway forward could include: (1) Prompt and parameter disclosure for reproducibility; (2) Ethical audit trails documenting hallucinations and reviewer interventions; (3) Philosophical delineation distinguishing insight from inference; and (4) Interdisciplinary oversight involving ethicists, historians, and clinicians alike.

Only when ethical reflexivity accompanies computational ingenuity can the concept of wisdom-artificial or otherwise-be meaningfully pursued.

The broader lesson

The study of Alexander’s death becomes a parable for our age: Power constrained not by enemy swords but by its own illusion of invincibility. AI, like Alexander, commands vast territories of information but falters before the boundaries of mortality-its knowledge expires with each software update. Time humbles all empires, digital and human alike. If wisdom has any artificial form, it is the recognition of transience: That no algorithm can capture the heartbeat of history, and no dataset can decode the silence between centuries.

CONCLUSION

Zhou et al[1] open a stimulating conversation on integrating AI into historical medical reasoning. Yet, their model of AW remains more metaphor than method, more rhetorical promise than ethical praxis. To equate machine synthesis with human discernment risks mistaking fluency for philosophy. Wisdom, whether natural or artificial, must remain accountable to truth-but truth itself is not static. It is a river shaped by time, not a monument carved in code.