AI in the Cathedral

KONSERT
NIDAROSDOMEN 28. juli, kl. 19:00
Billetter: Olavsfest.no
Kunsterisk leder: Øyvind Brandtsegg
Kuratorer: Zane Cerpina, Espen Gangvik
Partner: Olavsfest, Nidarosdomen

KI i Katedralen (AI in the Cathedral) [2026]

Øyvind Brandtsegg / Petra Bjørkhaug / Anna Thu Schmidt / Daniel Buner Formo / Andreas Bergsland / Dag Aakre / Pekka Stokke
Kostymer: Vilja Øyen Brandtsegg

The Cathedral

In a cathedral, a particular form of listening arises. The space carries centuries of cultural history, human longing, consolation, and the sense of the divine. It is a place where people have sought answers when language falls short, and where music and ritual have held experiences that resist verbal articulation. In this performance, we explore what happens when this space for existential support encounters the algorithms that increasingly shape our lives. What happens when comfort, interpretation, and guidance—once the domain of human beings, perhaps even the sacred—are taken over by machines? Can algorithmic pastoral care offer solace? Will we believe it—and more importantly: will we be able to?

Transcendence

This transcendence applies to the music as well. In traditional musical performance, expression is bounded by the limits of the human body: how fast fingers can move, how long a musician can endure, how many details a person can manage at once. When these limitations are lifted, new spaces open. How many tones can sound simultaneously when no hand needs to strike them? How long can a musical phrase be sustained when no lungs require air? Can we reach other, deeper artistic expressions when technique no longer defines the boundaries? Perhaps it is precisely in this tension—between the human and the unbounded—that new forms and artistic expressions can emerge.

Musical experience can be seen as a dance between recognition and surprise. We listen for patterns, and we are moved when the patterns break. But what happens when memory—human as well as machine—also shifts? Large language models, which underlie much of contemporary artificial intelligence, represent the world through text. Yet in dance, music, and visual art there are forms of knowing that cannot be captured in words. A machine may assemble words that make sense, but it has no experience of what those words mean, how they resonate, or how they move within a room. 

Precision

Modern algorithms often work by finding the most probable connections within large datasets, and by generalizing from them. But what happens to art if averages are allowed to dominate? If an approximate solution is good enough, because the system needs to move on? For art to strike its mark, it needs precision even where no concrete answers exist. An artwork carries traces of how it was made: is it handcrafted or mass‑produced? Shaped from finite, sustainable materials, or drawn from an endless digital flow? Both the method and the materiality become part of the aesthetic object—and in this meeting between aesthetics and ethics, new questions arise.

Power

For all the good the church holds, it is also an institution that has exercised power in many ways, not always positively. A cathedral contains all these layers, distilled and amplified. Nidaros Cathedral also carries the role of national shrine, and is therefore a bearer of the national, the sacred, and our shared cultural heritage.

Cultural heritage and knowledge are, in principle, the common property of humanity. Power and ownership are also inseparable from artificial intelligence. When machine learning is trained on enormous amounts of data, fragments of everyone’s property become part of a shared data‑body. Is this a new form of collective knowledge, or a resource for a few actors to capitalize on?

Historically, the power structures of the church have at times overshadowed its underlying message. In a similar way, the economic interests of technology companies may overshadow the genuine benefits that AI could offer people. Where the church promised salvation, technology companies promise artificial general intelligence. But the promise of a future universal intelligence may also obscure a sober assessment of what today’s algorithms are actually capable of. Technological centralization furthermore creates dependencies on providers in particular geopolitical regions. What happens when those power balances shift? And amid all of this stands the existential core—the part that is universal and belongs to everyone.

Interfaces

In technology, an interface is a point of connection between worlds—how access is granted to resources beyond a threshold. The church can also be understood as an interface, a place where people seek contact with something beyond the everyday. The organ is itself such an interface: between air and sound, between human intention and the resonance of the room, between the earthly and the divine.

What happens when the priest’s speech is channeled through the organ, when the words take on the instrument’s force and timbre? In this performance, we explore the interface between the human body, machine learning, and acoustic space. Our instruments use algorithms that connect the collective and the popular—such as a database of karaoke recordings—with the individual expression of a dancer moving through a virtual latent space. When the dancer’s movements call forth voices from this data‑space, and these voices manifest physically through the organ pipes, a circle is formed: from human movement in the room, through digitalization, then machine learning, then spectral analysis resulting in key-presses on the organ, and back into physical sound in the room. It is an exploration of how technology can extend, but also mirror, human expression.

Interplay

This year marks 70 years since the term artificial intelligence was first introduced. Since then, we have attempted to simulate ever more aspects of human intelligence: mathematics, language, emotion, social intelligence, creativity, musicality. Each time we (partly) succeed at simulating one domain, it becomes clearer what is still missing. This gradual unveiling has also become a way of understanding ourselves by understanding the technology and its simulations.

In this performance, we deliberately use simpler, more interpretable algorithms—not out of nostalgia, but because they give us an intuitive grip on the creative process. They require fewer resources, are more sustainable, and allow us to see more clearly how the digital and the human can collaborate.

Throughout the performance, the computer aesthetic shines through in varying degrees. The work is an interplay between old and new, between classical cybernetics and modern machine learning, between human experience and algorithmic computation. From this meeting point, a space opens where we can ask the big questions again.

—

Photo credits. Photo representing the project by Juliane Schütz. Portrait photo of Øyvind: Juliane Schütz. Portrait photo of Petra: Juliane Schütz. Portrait photo of Andreas: Berre. Portrait photo of Anna: Juliane Schütz. Portrait photo of Dag: from private archives. KONSERT