Beyond Imitation: Generative AI and the Future of Learning and Making

Artificial intelligence is now designing, drawing, and describing architecture—mirroring, augmenting, and occasionally replacing our ways of seeing, sorting, making, and imagining. A groundbreaking imitation machine, Generative AI creates new output drawn from curated collections of precedents, which in turn inevitably reflect biases and implicit values.  

Held on November 14–15, 2025, at Ca’ Giustinian, the headquarters of the Venice Architecture Biennale, the international symposium Generative AI and the Future of Learning and Making: Imitation, Bias, and Estrangement brings together architects, scientists, and theorists to explore intelligence as a shared condition between humans, machines, and the environments they co-construct. As one of the final symposia of the Biennale, as part of the Gens Public Programme, it looks toward the future of architecture—beyond imitation, beyond automation, and beyond the idea of technology as merely instrumental.

Emerging from the broader Beyond Imitation research platform, a collaboration between the Media x Design Lab, École Polytechnique Fédérale de Lausanne (EPFL), and The Bartlett School of Architecture, University College London (UCL), under the leadership of Mario Carpo, Lidia Gasperoni, and Jeffrey Huang, the symposium examines how AI reshapes creativity and knowledge across three intersecting trajectories: Generative AI in Architecture, Politics of Datasets, and Forms of Intelligence: Learning and Making from Neuroscience to Robotics.

The first trajectory, Generative AI in Architecture: State of the Art, investigates how algorithms reconfigure authorship and imitation, producing novelty through the assimilation, repurposing, and transformation of existing models. The second, Politics of Datasets: Invisible Influences and Alternative Epistemologies, uncovers the aesthetic and ideological hierarchies embedded in data infrastructures and the epistemic frameworks they sustain. The third, Forms of Intelligence: Learning and Making from Neuroscience to Robotics, envisions the emergence of AI systems that learn from the world itself, where cognition arises from material engagement and embodied experience.

Each trajectory addresses a distinct facet of the same condition: the tension between imitation and creation, convention and invention, automation and understanding, representation, experimentation, and experience. These inquiries form a network of perspectives on the contemporary entanglement of architecture and computation. They share a central concern: how intelligence—whether human or artificial—can engage with precedent while responding to unprecedented challenges.

The symposium convenes leading contributors including Refik Anadol, Matias del Campo, Immanuel Koh, Kyle Steinfeld, Mark Wilson-Jones, Noam Elcott, Catherine Griffiths, Orit Halpern, Matteo Pasquinelli, Daniela Rus, Olaf Blanke, Philip Yuan, and Fabio Gramazio, among others.

The event is curated by Mario Carpo (The Bartlett, UCL/Italian Academy at Columbia University in New York), Lidia Gasperoni (The Bartlett, UCL), and Jeffrey Huang (Media x Design Lab, EPFL).

For details, schedule, and updates, visit beyondimitation.ai. This event is free and open to the public.