Re-Experiencing History

With our AI project, we are opening a window to the past. “Re-Experiencing History” makes it possible to experience the past visually for yourself. Using an interactive platform, seemingly lost worlds are brought to life with the help of AI and precisely trained models – from magnificent triumphal processions of the Romans to everyday life in classical Greece.

This project marks an important step in how we experience and convey history. Instead of being limited to reconstructions from picture books or film adaptations, we provide a tool to create your own visualisations. Using a prompt interface and fine-tuned AI models, anyone will soon be able to design and visualise historical scenes according to their own ideas.

What’s special about this interface is that the latest research findings will be incorporated, so the visualisations reach a completely different level than typical AI models. In this way, almost any situation from antiquity can be modelled and re-experienced, such as the marching legions or the triumphant general on his chariot from various perspectives of the cheering crowd during a Roman triumph. This interactive experience is not just a game. It enables a much deeper understanding of historical contexts and awakens interest in and fascination with the past in a completely new way.

The above video has been generated using Gemini.

The project opens up a new dimension of historical communication by enabling users to actively engage with historical topics. Complex historical issues become easier to understand and access through individual visual design. The broad range of applications extends from education and museums to documentaries, archaeological research, and tourism. Individual visual representation also creates a more sustainable understanding of historical contexts – accessible without barriers and for everyone.

Why?

The epistemic value of visual reconstructions lies not primarily in illustrating already known knowledge, but in the heuristic power of visualisation itself. Historical texts often operate through condensation, abstraction, and implicit cultural assumptions. Yet once a scene must be concretely visualised, such as a triumphal procession, a coronation ceremony, or a diplomatic encounter, precise questions inevitably arise. Who stands where? Who looks at whom? How are bodies, hierarchies, spaces, and objects arranged? Which gestures signal power, submission, or proximity? Which material conditions, including architecture, clothing, insignia, and public setting, structure the event?

In the text, these elements are often only implied. The necessity of visual concretisation forces implicit assumptions to be made explicit and alternative interpretations to be systematically explored. The real epistemic gain therefore emerges through the performative modelling of historical situations. By varying specific parameters such as spatial arrangement, social distance, lines of sight, body posture, and audience configuration, different interpretative scenarios can be simulated and compared.

This opens new pathways for hypothesis formation. Which staging strengthens monarchical authority? Which weakens it? How does the perception of an event change when certain groups are made more visible or marginalised? Visual reconstructions thus become an experimental laboratory for historical scenarios.

What is particularly innovative is their systematic iterability. Whereas traditional reconstructions remain static, AI supported visualisation enables the serial generation of variants. These variants reveal that historical events consist not only of facts, but of relational constellations that can be interpreted in different ways. Comparing multiple visualisations sharpens awareness of contingency, perspectivity, and narrative construction, which are central categories of modern historiography.

At the same time, visual reconstructions act as a catalyst for interdisciplinary research. They connect historical textual analysis with spatial theory, performance studies, visual studies, and cognitive psychology. Visual modelling compels us to think material culture, social interaction, and symbolic communication together in an integrated way. In doing so, it generates new research questions that might not have emerged through text based analysis alone.

Possible applications

In education, students can create interactive learning materials that bring history to life and promote interest in historical topics. Museums could give visitors the opportunity to interactively co-create exhibitions and visualise their own interpretations of historical events. For documentaries and films, the platform can serve as a source of inspiration and help in the development of authentic and detailed historical scenes. In archaeology and research, user-generated visualisation makes it possible to illustrate different interpretations of finds and historical sites.

How does the project represent research at the Faculty of Arts and Social Sciences (PhF) of the UZH?

The project creates a previously unique bridge between the humanities and computer science. It unites a wide variety of disciplines – Ancient History, Computational Linguistics, Classical Philology, Digital Humanities, and Visualisation Sciences – in a groundbreaking, unconventional collaboration. It thus embodies the guiding principle of the Faculty of Arts and Social Sciences: the courage to break new, forward-looking ground while promoting the responsible use of artificial intelligence. At the same time, the project offers students and early-career researchers the opportunity to actively participate in innovative research and acquire valuable skills in the dynamic field of digital humanities.

The web application will be available under reexperiencinghistory.hist.uzh.ch soon. An account is necessary to use the app. Registration will open soon.

Project lead:

• Prof. Dr. Felix K. Maier

Technical realisation:

• Dr. Phillip B. Ströbel

Student assistants:

• Michèle Egli
• Nicola A. Steger
• Eva Maria Willi

Student support:

• Zejie Guo (MA thesis)
• Ülkü Karagöz (MA thesis)