Uncanny Patterns: Visual Common Sense and the ‚Life‘ of Algorithms
Prof. Carolin Behrmann
Wednesday, 23 February 2022, 5:30 pm
ONLINE Lecture, hosted by the Warburg Institute.
Free and open to all via Zoom; please book in advance for this event via the Warburg Institute website.
The spectre of computer vision is haunting the idea of ‚common sense‘ today: Big-data and smart algorithms that permeate daily life are linked to image recognition technologies and immeasurable image archives of the web. Visual algorithms try to predetermine reality, life and action, but the expectation that machines would learn through visualization what is and how to react according to a ‚common sense‘ is still a hard nut to crack for artificial intelligence. The talk discusses the ‚uncanny‘ effect of automated visual pattern recognition on the production of (imaginative) ‚common sense‘ that is shared by many — once defined as the ground of historicity (Giambattista Vico) and democratic politics (Hannah Arendt). It questions the commonsensical „power“ of tertiary memory (or non lived-experience) and biased algorithmsthat undermine subject positions and also aesthetic experience. It will look at artworks intersecting at well-known iconographies and investigations in, or work with AI technology (e.g. Refik Anadol, Davide Quayola, Beeple crap) creating own algorithms and digital architectures. Pattern recognition seems here a cooperative work of artist and machine that transformed collective representations and our ways of seeing. If this „dissensual common sense“ (Jacques Rancière) is a tension that produces commonality in the future, how does it mediate our relations to the past?
The Bilderfahrzeuge Lecture series 2021/22: “Pattern Recognition”
The term “pattern recognition” owes its ubiquity across a wide range of sciences to advances in fields like computer programming, statistics and cybernetics after the Second World War. Yet the basic process of ordering raw data or visual images into identifiable structures has existed for much longer, also within traditionally humanistic contexts. Aby Warburg’s multidirectional and nonlinear science of images (Bildwissenschaft) began with the research of patterns and ‘pathos formulas’ in the historical record of European and Near Eastern art. Today, as our interactions with images are increasingly shaped by nonhuman intelligences, the practice of visual pattern recognition offers new ways of historicizing the changing relationships between psychology and technology, intuition and automation, cognition and control. This lecture series explores how pattern recognition has functioned and continues to function as a distinct modality of image-based knowledge production, from the study of global ornaments in the nineteenth century to digital surveillance and profiling in the twenty-first.