IN-PERSON London Heritages. Section A

Abstract

In the realm of architectural heritage studies, the integration of nonconscious cognition, facilitated by advancements in artificial intelligence (AI) and machine vision systems, presents a transformative opportunity for collaborative intelligence. Drawing upon insights from data philosophers Luciana Parisi and Matteo Pasquinellie and architectural historian and critic Mario Carpo this study explores synthetic cognition’s non-optical vision as an opportunity to “unlearn” biases ingrained in traditional approaches to architectural analysis. Emphasising the centrality of data in design processes, this paradigm shift underscores the necessity for architects to comprehend the pivotal role of data curation AI-driven design workflows by becoming machine educators. Departing from the anthropocentric model of perception, architects are urged to embrace non-human agents to augment precedent analysis and challenge biases ingrained in the reading of architectural history. Contemporary visual paradigm is no longer optical. By training machine vision algorithms to differentiate architectural styles and employing techniques such as Visualising Activations or DeepDream this research reveals the distinctive characteristics and biases inherent in multitude of notational formats. This transition from analogue into the digital is framed as an evolution of Alberti’s notational principles. Case studies such as the point cloud scan of Christ Church Spitalfields and and Our Lady Help of Christians Church examin the tension between data compression and spatial accuracy under the machine gaze. In response, educational strategies must evolve to encompass both traditional architectural teachings and methodologies tailored to non-conscious learners; a curriculum that fosters human-machine collaboration and equips architects with the skills to navigate AI-driven design processes.