Abstract:
The book "Artificial Historians" (2026) is dedicated to a fundamental rethinking of who and how history is created in the age of artificial intelligence. The author argues that AI is already today a full-fledged "historian": algorithmic recommendation systems, social networks, credit scoring, mapping services, and even CAPTCHA continuously analyze past data and, based on this, shape people's access to education, healthcare, justice, and information. "Artificial historians" operate invisibly, behind interfaces, but their "works" – ranked news feeds, product suggestions, reliability assessments—directly affect the lives of millions. The study focuses not just on critiquing technologies, but on analyzing the logical, epistemological, and ethical foundations that underlie both human and machine historiography. The methodology of the reviewed study is based on a comparative analysis of awarded historical works and the outputs of generative AI models, using Aristotelian logic, the concept of "endoxa," modality, and warranted theory. The scientific novelty of the book lies in the radical expansion of the concept of "historian": the authors convincingly argue for the first time in historiographical discourse that AI algorithms already perform a historiographical function and propose to view them not as a threat, but as an object of active professional intervention. The key contribution of the reviewed book is a detailed comparison of logical markers in human historiography (questions, modal constructions, metadiscourse, nonlinear temporality, the role of the author's body) with their absence or imitation in texts generated by AI. The authors find that standard AI models tend to present information as absolute truth, obscuring their logic and not signaling the incompleteness of knowledge. The main conclusion of the book is intriguing: the danger of artificial intelligence for history lies not in its technical imperfection, but in the passivity of the professional community. Historians' refusal to participate in the design, regulation, and auditing of AI leads to a loss of control over how current models construct the past. Using the example of the philosophy of Indigenous Australians (Kaurna, Yuin), the authors propose an alternative model—a "living" history, where time, place, and ethical responsibility are interconnected, which opens the way for the creation of more just and multidimensional "artificial historians" of the future.
Keywords:
artificial intelligence, Artificial Historians, endoxa, large language model, modality, metadiscourse, technogenic historiography, Kaurna, popularity bias, methodology of history