IAAC Bits 10 – Learning Cities

Description

The digital intelligence that is inevitably starting to penetrate every aspect of our previously analogue systems of living, working or social interacting calls for new models of designing our city and opens new territories of experimentation in the processes related to urban design.

While the idea of intelligent machines that simulate “cognitive functions” such as “learning” or “problem solving” is not new, its extensive use, in recent years, in the urban design discipline opens up a series of new possibilities – as well as plenty of cultural, ethical or even aesthetic hesitations and risks. How do our cities learn? Can machines design and what? Is crowd intelligence appropriately harvested in our evolutionary and generative design processes? And has our current big data analysis approach reached the limit of human and computational intelligence?

Learning Cities explores the “intelligence” applied in the processes and outcomes of designing our urban environments. From a variety of applications of artificial intelligence and machine learning for urban planning to co-creation processes that merge crowd intelligence with digital technologies, Learning Cities highlights that “intelligence” in the built environment should be understood beyond human, object or machinic intelligence alone. Through a variety of contributions from experts in different fields the current IAAC Bits Journal Issue explores novel collective intelligence design processes in which designers, users, the built environment, and digital codes all play a fundamental role in a unique resonance that takes place among them. 

With Contributions of Areti Markopoulou, Manuel Gausa, Jordi Vivaldi, Benjamin Bratton, John Frazer, Molly Wright Steenson, Stanislas Chaillou, Sarah Williams, Theodora Vardouli, Neil Leach, Angelos Chronis, José Sánchez, Mathilde Marengo, Aldo Sollazzo, Aleksandra Sojka, Matias del Campo, Chiara Farinea, Rodrigo Delso, Sandra Maninger, Javier Argota, Cobus Bothma and others.

Sujets

• urban planning
• Architecture

Informations

• Publié par: Actar D
• Date de parution: 20 juin 2023
• Nombre de lectures: 0
• EAN13: 9781638408130
• Langue: English
• Poids de l'ouvrage: 9 Mo

Extrait

Learning
Cities



bits




2




3












Edited by
Areti Markopoulou


Learning
Cities


bits




4


Front page figure: IAAC MaCT, OWNit, F.Ciccone, L.Marcovich, 2017.


IaaC BITS during 2020 has started a new editorial phase that is intended to be more effective, ambitious and intentional both in terms of content and in the layout and configuration of the publication.
These monographic issues – presented with an experimental and proactive foundation and associated to technological and creative innovation – aim at to combining inter-disciplinary and multi-scalar exchanges with a new environmental and socio-cultural sensitivity.
This commitment to advanced culture and knowledge is well-suited to a time of challenges and changes: it conforms the conceptual framework that supports dissemination projects tied to IaaC’s own production, but also to a whole network of exchanges and complicities that frame it and feed into it.
Each issue is meant to be conceived as an articulated system of voices and cross-cutting experiences focused on a central theme, which is understood as a subject for debate and proactive discussion.



5


Index


Learning Cities: Artificial vs. Collective Intelligence in Urban Design
Areti Markopoulou
On Syntethic Intelligence and Design: A conversation
Benjamin Bratton with Areti Markopoulou and Jordi Vivaldi
Brain City
Neil Leach
Machine Estrangement: On Leo Tolstoy, Algorithmic Otherness and Urban Design
Jordi Vivaldi
Learning Machines: A Story of Control, Softness and Agency
in Architectural Computing
Theodora Vardouli
Semantics in Architecture
From Words to Forms and Back Again
Stanislas Chaillou
Architecture and Intelligence: A conversation
Molly Steenson with Areti Markopoulou and Jordi Vivaldi InFraRed: An Intelligent Framework for Resilient Design
Angelos Chronis
Urban Fictions, SPAN
Matias del Campo and Sandra Manninger
Sensing Public Spaces
Aldo Sollazzo
On Evolutionary Digital Design Processes
and Citizen Involvement: A Conversation
John Frazer with Areti Markopoulou and Jordi Vivaldi
Social Realism in the Age of Simulation
Jose Sanchez
Superbarrio
Gamification Tools for Data-driven Participatory Urban Design
Marco Ingrassia, Areti Markopoulou and Chiara Farinea

Data Action
Sarah Williams
A City Like You and Me. Embracing Uncertainty in the Age of Precise Data
Rodrigo Delso and Javier Argota
Relating Human to Machine Aleksandra Sojka, Snoweria Zhang, Deborah Churchill, Cobus Bothma
Learning Cities, Shared Co-Cities, Empathic Cities
Manuel Gausa


37


49


73


121


157


97


139


61


173


187


199


149


109


87


15


09


31



















6




7



Critical Views


What will the city of the future look like?
Can machines design and what?
Which new models of architectural conceptualization arise from algorithmic control?
Which is the novelty and relevance of using artificial intelligence in the design process?
Will the ethical, social, cultural and aesthetic implications of AI affect the quality and form of urban and architectural space?



8


Intro




9


Intro


Intro


“…we teach our environments first complex, then self-organizing, intelligence that could eventually become evolutionary”
Warren Brodey, 1967 The application of artificial intelligence, machine learning or statistical models and algorithms for the creation of predictive models in architectural design and urban planning is based on relational and evolutionary logics that promise a more (multi- objective) “optimized” design or a “smarter” city, where “smart” is usually defined based on notions of improvement at infrastructural scales.


Areti Markopoulou


Learning Cities
Collective Intelligence in Urban Design


Most of the discussions around the idea of urban “intelligence” focus on the use of sensory technology in billions of interconnected objects (Internet of Things) or on the processing of billions of images that can be mined from web services to collect vast amounts of data produced by human behaviour and other sources, such as the environment or the material world of our built space. Data from user occupation patterns in housing, for instance, can generate economically viable interior distributions in building development. Similarly, data on how people move in cities can be inserted into computers and generative algorithms with the goal of helping predict traffic and optimizing mobility planning.
In a highly human-centric technocratic world, such notions of optimization might be the gold standard for urban policies and national strategies.


Previous page: IAAC MaCT, Alzette 2.0. S.Subramani, L.Saadi, M.Galdys, I.Reyes, 2020.



10


Intro


Machine Intelligence
While the idea of intelligent machines that simulate “cognitive functions” such as “learning” or “problem solving” has its roots back in the cybernetics of the 1950s, its extensive use, in recent years, in the architectural and urban design disciplines opens up a series of new possibilities – as well as plenty of cultural, ethical or even aesthetic hesitations and risks.
Although it seems to be common thinking that AI is an evolution of cybernetics, one could argue that cybernetics deals with a much more holistic view of intelligence, especially because it focuses on how systems can self-regulate and act in constant feedback with the environment, which contains far more diverse factors (biological, social, mechanical, economic) than the AI computational stored representations of the world. 2,3 Beyond the technical and systemic


But moving beyond an understanding of optimization as selecting the best result from a variety of computed and quantified data-based possibilities (a reading which carries the heavy burden of the last two decade’s mainstream idea of the “smart city”), we need to explore optimization as patterns of co-creation and as a continuous changing state incorporating both human and non-human perspectives. Within a multiverse inhabited by different human, technological and cultural others, purely quantified data sets may be significant, but they fall short, if they are only observed by a cognitive, and always biased, perspective. 1


Superbarrio video-game interface for Nantes.




11


Intro


differences between cybernetics and AI, however, there is a common connotation to AI related to the question of the future, which is of major social (rather than computational) significance. As Steenson highlights in her contribution to this issue, “AI is not only about using the computer, but about having a vision of it. If we talk about AI, we want to talk about a vision of what computation does today and could do in the future.” The digital intelligence that is inevitably starting to penetrate every aspect of our previously analogue systems of living, working or interacting socially becomes the central core of a variety of utopic or dystopic futures. That is the case because such intelligence can be an empowering tool as well as a disempowering one for both the people who inhabit the built environment and those who design and manage it. The potential of recognizing the diverse (microbial, animal, vegetable, machine, human) intelligences in cities and re- orchestrating them for better planning (Bratton pp.15), the ability to create real time digital twins for multi-stakeholder decision making (Chronis pp. 87), or the novel modes of machinic perception generated (Del Campo & Manninger pp. 97), which can enrich the field of urban design (Vivaldi pp. 37) are some aspects of one of the sides of the AI reality that has been embedded in our everyday life and habits, usually without us even noticing it. Other issues such as the anonymous, faceless city designed from the top down, based on statistical simplifications (Vardouli pp. 49), digital exclusion, planning processes with no democratic ends, and mass surveillance followed by the rise of new forms of behavior manipulation (Sollazzo, pp. 109), are critical risks that can be found on the other side of the exact same reality, fully powered by the applications of AI in the design and management of urban environments. Within this context, how do we define “intelligence” in built space? How can we structure the vast data powered by geocoded web services or by millions of interconnected objects (IoT) and turn it into valuable input for informed decision (and design) making? The first two sections of “Learning Cities” promote a number of established and emerging inno