Tica, Kristina "COMPUTATIONAL AESTH-ETHICS: Understanding visual computation processes between the image and its context." Master's
, University of Art and Design (Kunstuniversität), 2023
Keywords/Fields of Study : machinic vision, image-making, image-reading, database, automated creativity, aesth-ethics, generative ideology, algorithmic thoughtlessness, algorithmic culture, demystification
Abstract: This master thesis is developed as an artistic practical and theoretical research, established at an intersection of computational image-making and image-reading processes, and their aesth-ethics in a broader socio-technical environment. The presentation of my artistic research, along with relevant projects in the same field, will be followed by a theoretical discussion on the ontology of a digital, computer-generated image, its aesthetics and ethics within underlying ideologies in the discourse of their representation - between their content and context. The discussion is located on the axis between the image analysis and synthesis processes elaborated through the implementation of machine learning-based tools.
The analysis in this research goes beyond the representational surface of the image, articulating the history of artificial intelligence and computer-generated images, correlations and oppositions between the affordance of the medium, tools and their syntax, and their limits of creativity within the context of critical data and image semiotics. Through a selected series of three projects: Digital Prayer (2020), FUTUREFALSEPOSITIVE (2021), and PROMPT: WAR STORIES (2022), the concept and the approach in using different machine learning-based tools will be elaborated, defining the threshold between the human agency and algorithmic automation.
The epistemology of computed, statistical images is disseminated through their aesthetics of representation, carrying an ideology of 'thoughtlessness by automation’, examining the balances of co-creation between the human and the machine - determining agency, against metaphors, fantasy and mystification; probability and prediction. Between optimisation and accuracy, quantity and quality, universality and uniqueness, this research underline artistic examples and theoretical concepts that address the importance of human agency in the age of automation. Paradigms of development and integration of machine learning systems on a political, social and ethical scale can be understood through the artistic appropriation of these tools.
From the database to the combinatorial infinity and statistical transcendence, the critical points between the spectacle and backstage resources of statistical art will be extracted, observing the image behind the image, and the datasets beyond the computation. The semiotic interplay between the content and context becomes necessary to address responsibility and visibility, so as to better understand not only the technologies but the systemic structures, along with the ideologies, that lie behind them.
Department: Interface Cultures
, University of Art and Design (Kunstuniversität)
Advisor(s): Manuela Naveau