Towards critical entanglements of machine learning aesthetics and architecture

Machine learning algorithms are increasingly used in diverse fields of creative production, from music to painting, and from theater to architecture. Investigating the relationship between digital authorship and creativity has been an issue of exploration from as early as Turing’s inquiry on whether computers can think. These preoccupations are especially complex since neither creativity nor intelligence have been defined. As many machine learning researchers have shown, the
definition of intelligence has been constantly changing over the last years, posing conceptual difficulties to the field of artificial intelligence in general [1].

Building on previous work connecting technology and creative practices in general [2] and to architectural production in particular [3], [4], this paper presents a design-based investigation on using machine learning algorithms to create conceptual architecture. It presents the design methods employed together with the results of these explorations and discusses them in the broader context of technology and creativity building on Kyle Steinfeld’s [5] categorization of machine learning in art
and architecture as (a) actor, (b) as material and (c) as provocateur. This might help to uncover implicit biases in training models and will contribute to current discussions on (creative) authorship in a (post)-digital age [6].

Traditionally, theory is elaborated at a slower pace compared to practice, and this has been especially true over the last decades dominated by technology-driven development. Across design fields, many have called for more theory in design research [7]. Unpacking technology is important if we are to understand how new tools are changing the way we conceptualize, design, build, and operate architecture, and on the ways in which we educate future generations of architects. In a technology-saturated landscape, we should ensure both intimate technology practices and a critical stance towards these same technologies. Both the algorithms themselves and the data used to train them need to be dissected to contribute to useful theories of the digital.

We ask: how should we ensure that we, architects, designers and artists maintain a critical stance towards technology in our current entanglements? How is the work of the creative practitioner changing in the (post)-digital age? What is gained and, importantly, what is lost in this retooling? These questions are important because, in Donna Harraway’s words: by the late twentieth century, our time, a mythic time, we are all chimeras, theorized and fabricated hybrids of machine and organism; in
short, we are cyborgs. This cyborg is our ontology; it gives us our politics. [8]