Iconography, Identity, and Algorithmic Vision

The name “Jesus Christ” has, for over two thousand years, produced an inexhaustible variety of visual interpretations. From the rigid symmetry of Byzantine Pantocrator icons to the sensuous chiaroscuro of Renaissance oil paintings, from Ethiopian gospel illuminations to contemporary street murals, each image purports to depict the same historical figure while inevitably bearing the marks of its cultural, theological, and aesthetic context.

In Christian art history, this diversity is underpinned by a paradox: the simultaneous belief in a singular, unchanging person and the proliferation of radically divergent faces. Traditions such as the acheiropoieta — “images not made by human hands,” like the Mandylion or the Shroud of Turin — sought to anchor a “true likeness,” yet even these relic-images spawned innumerable stylistic adaptations. The act of picturing Jesus has always been a negotiation between faith, cultural projection, and the limits of visual representation.

One Name, Many Faces brings this ancient dynamic into the era of artificial intelligence. Here, a single personal name is given to an AI image-generation system, which returns a multiplicity of portraits. The algorithm functions as an unorthodox iconographer: instead of working from scripture, oral tradition, or religious mandate, it draws upon statistical correlations within its training data — the accumulated visual biases of the digital age. Just as the medieval painter’s brush was guided by local physiognomy, pigment availability, and prevailing doctrine, the AI’s “hand” is guided by the structures and omissions of its dataset.

This parallel invites reflection on author, authorship, and authenticity. Neither the painted Christ nor the machine-generated portrait can claim objective truth. Yet both become part of a larger visual mythology in which the name — not the likeness — anchors identity. The viewer’s recognition is less an act of verifying reality than of entering a shared symbolic space, where variations do not weaken but rather enrich the aura of the original name.

One Name, Many Faces is not merely a technical experiment but a continuation of an ancient human (and now post-human) impulse: to make the invisible visible, to reconcile the fixed word with the mutable image, and to accept that every “true face” is, inevitably, one among many.

The Speed of Creation
For roughly two thousand years, humanity has created images of Jesus Christ — from the humble frescoes of the early Christian catacombs to Byzantine icons, Renaissance masterpieces, and contemporary reinterpretations. Each work emerged from the slow, cumulative rhythms of human history: years of apprenticeship, the scarcity of materials, the deliberations of faith, and the gradual evolution of style.

In radical contrast, an artificial intelligence produced 19 such representations in a total of only about 100 seconds. If we translate this into human terms — assuming that a single AI “second” corresponds to what a human might achieve in 36 years — the production of this series spans the equivalent of approximately 3,600 years. That is longer than the entire recorded tradition of Christian image-making.

This compression is more than a feat of speed; it is a distortion of time itself. Millennia of human artistry — the slow drying of pigments, the decades of apprenticeship, the generational passing of style and symbol — are folded into the span of a breath. For the AI, centuries arrive and vanish in the blink of an algorithmic eye; for the human witness, each machine-second unfurls into an epoch of labor, devotion, and imagination.
In this strange calculus, time becomes elastic time — a medium in which history can be condensed, accelerated, or stretched at will. The image no longer belongs to a single era, but to all eras at once, suspended in a moment that is both instantaneous and eternal.