The Armenia Pavilion team, led by Electric Architects, scanned hundreds of ancient monuments threatened by conflict, climate, and neglect, training an AI model to capture their essence. The AI generates confabulations—historical fragments of architecture that never existed— carved in stone, remaining true to tradition yet projected into the future.
Mediated by machine intelligence, these artefacts invite visitors to reconsider the role of memory in architecture and the challenge of preserving heritage. Architecture and memory exist in constant dialogue. Architecture embodies collective intelligence, shaping cultural legacies and identity.
AI, as an extension of collective thought, allows us to reinterpret traditional forms and conceive new possibilities. It bridges time, using ancient monuments as a foundation for new memories.
The curator team worked with machine learning algorithms trained on Armenian architecture to generate new, imaginary forms that evoke a familiar yet speculative language of space. These digital models were then translated into physical reality through machine-carved tuff stone, creating a dialogue between ancestral material and computational imagination.
The result is a constellation of microarchitectures that embody both loss and continuity, offering new ways of transmitting identity across borders and generations.
