unit-code
‘If a machine is expected to be infallible, it cannot also be intelligent.’
– Alan Turing, Lecture to the London Mathematical Society, 20 February 1947
As this quote by Turing suggests, designing with artificial intelligence (AI) does not necessarily mean taking the output of a model as an absolute conclusion. It requires interrogation and interaction that are achieved through various algorithmic and computational approaches being integrated into the design process. For this, Research Cluster 11 engages with fundamental mathematical concepts to achieve computational literacy. In addition to these core principles, we strive for the adoption, adjustment and application of advanced machine learning (ML) methods. We argue that the potential of computers in design lies not solely in their pure calculating power but in how algorithms can offer ways of abstraction, after which a multiplicity of solutions can be explored for a wide range of problems. We engage in developing computational methodologies that act within these hidden layers of design.
These principles are explored through speculation on the future of central business districts (CBDs). Today, if we look at the main typology and functional body of CBDs – the skyscraper and standard office building respectively – we cannot help but think that with the advent of computationalism both will undergo major changes.
The replacement of intellectual tasks by extraordinary AI tools forces us to question the very existence of CBDs whose economic rationality is no longer evident. By studying these districts around the world and through a detailed case study of La Défense, the largest purpose-built business district in Europe, the cluster has explored the future of CBDs in the era of AI.
The projects carried out over the year have addressed this discourse across a range of topics and methodologies. Each formulated a proposal to address the contemporary issues that are inherent to the urban typology of CBDs. Through gathering an extensive breadth of data on La Défense collected from multiple sources and by responding to this data through various algorithmic and ML methods, the projects not only speculate on future scenarios of CBDs, but design them.