4311
Vakula, Matej "Dark Design: Methodology for Ethics of Generative Modeling Practices in Biodesign Based on Machine, AI and Animal Ethics." Ph. D.
, , 2021
Keywords/Fields of Study : bioart, biodesign, bioethics, ai ethics, machine ethics, animal ethics, generative modeling, biotechnology, organ printing, tissue printing, bioprinting, sciart, nonhuman, posthuman, anthropomorphism, correlationism, transhumanism, neural, embryonic stem
pages, written in English
, copyright Matej Vakula
Abstract: My dissertation, broadly speaking, focuses on the ethics of generative organ design and the ethics of tissue printing. This research critically explores the feasibility and the preconditions for creating machine-learning-based automated ethics for generative design and organ printing. Moreover, by culturally analyzing the development process of these emerging technologies through the process of making simultaneously with the development process of automated ethics and their impact on scientific culture, I argue that specific biological artworks concerned with nonhuman ethics, multi-species interaction, and the relationship between technology, nature, and culture are ideal sources to obtain new ethics that can fill in the gaps caused by the current human-centered approach.
My research intervenes in the debates on algorithmic bias in generative design, particularly what does and what does not get transferred through the reductive processes of algorithmic abstraction into the digital biological model. Simultaneously, my research also intervenes in debates associated with AI ethics by analyzing the automation of bioethics related to generating the biological model. Finally, using bioart as a counter-model that offers better cultural and ethical investigation tools, my dissertation opens a new direction for the empirical and cultural study of non-anthropocentric perspectives within biodesign and bioethics automation.
To prove my argument about the advantage of using art as a source for bioethics training, I have trained the new machine-learning-based ethics on a model bio-artwork I designed and 3D printed using human neural embryonic stem cells put together with animal brain cells. When combined, these tissues construct a reward-based human-animal learning system that metaphorically flips the Descartian and Kantian anthropic hierarchy where the human dominates the "animal-machine." My research targets this anthropic hierarchy because it is the main driving principle behind correlationism and anthropomorphism in contemporary bioethics and AI ethics. Both are biases that prevent a more profound ethical understanding of the relationship between humans and nonhumans and are reflected in the contemporary Transhumanist ideologies embedded in parts of the current scientific culture.
Advisor(s): Kathryn High, Shawn Lawson, Oron Catts, Robert Nideffer
