Machine Learning and Architectural Pedagogy: Unlearning Precedent, Confronting Historical Bias, and Proposing Futures Architecture
This project applies machine learning (ML) to architectural heritage analysis in an effort to identify previously overlooked cultural and social biases implicit in design pedagogy. This research reconsiders the role of AI technologies within architectural pedagogy and asks how this might inform future design. In doing so it proposes to assemble a novel 3D digital dataset of historical styles of London-based baroque and brutalist architecture from which an intelligent system can learn and generate propositions that highlight overlooked properties of historical data. This process challenges the traditional reading of precedents in architectural education and positions ML as a method for unlearning in architectural history.
More about Malgorzata
Malgorzata is an architect, an educator and a current researcher at the Royal College of Art in London. In her research practice Malgorzata explores the transdisciplinary advancements in AI and architectural applications of machine vision to urban and heritage studies. Her work has been awarded, published and presented internationally. In her research and teaching practice Malgorzata positions ML as an emerging semiotic system, tackling issues such as data bias and algorithmic governance and as means for unlearning in architecture.
Postgraduate Diploma in Architectural Practice, RIBA pt. III
University of Greenwich (UoG) 2017, Merit
Master of Architecture, RIBA pt. II
UoG 2014, Merit
Bachelor of Arts in Architecture, RIBA p. I
London South Bank University (LSBU) 2011, First Class Honours
Visual Artist (Visual Advertising)
School of Fine Arts, Wroclaw, Poland 2006, Distinction
Research Associate, Laboratory for Design & Machine Learning, RCA, 2020-present
Associate Lecturer, Architect Degree Apprenticeship (Level 7), Oxford Brookes University, 2019-present
Project Architect (ARB), PRP Architects, 2016 - present
Architect (ARB), Design Consultancy, 2014-present
Student Prize for Recognition of Excellence, RIBA South London Society of Architecture 2009
Tempietti of the Large Hadron Collider’ at The Gold Mine Group Exhibition, LonCon, ExCel Exhibition Centre, 2014
Starzynska, Malgorzata. 2019. “Machine Vision and Algorithmic Creativity.” In: Dylan Dixon (ed.) RIBA Research Awards Book of Abstracts: 60.
Machine Learning and Architectural Pedagogy: Unlearning Precedent, Confronting Historical Bias, and Proposing Futures Architecture presented at ‘ARCH+ features SToA: Threads and Hinges – The Global Flow of Architectural Elements’, IGmA University of Stuttgart, 15 April 2021, Stuttgart, Germany.
Application of Image Recognition Algorithms to Architectural Styles Analysis presented at ‘Connections: Exploring Heritage, Architecture, Cities, Art, Media’ at University of Kent, 3-4 Sep 2020, Canterbury, UK.
Machine Vision and Algorithmic Creativity presented at the ‘Pixels, Vectors and Algorithms' at Architekturmuseum der TU München, 11 Oct 2019, Germany.
Machine Vision and Algorithmic Creativity presented at the ‘RIBA Research Matters 2019 - Nottingham’, 17-18 Oct 2019, Nottingham,UK.
The Machine, EAP Pre-sessional Programme, 26 Aug 2020, RCA, London, UK.
The Architecture of Learning Algorithms, Perspectives Lecture Series, 6 Nov 2019, The Bartlett School of Architecture, UCL, London
Machine Vision and Algorithmic Creativity, Critical Historical Studies Lecture Series, 29 Nov 2019, RCA, London