From the 8th to the 9th of December 2023, the Oxford-Berlin Early Career Colloquium was held. Various promising early career researchers presented their projects concerning ethics in AI, discussing its multifaceted aspects such as normativity in the digital community, the normative relationship between humans and AI, systemic injustice perpetuated through the AI capabilities, the valorization of AI-driven practices, and gender issues in data processing. At this colloquium, numerous scholars, including Shannon Brick, Atay Kozlovski, Megan Hyska, Mike Barnes, Kathleen Creel, Stefano Calboli, Jared Parmer, and Marion Boulicault, presented their projects. Additionally, Ted Lechterman, Linda Eggert, Rob Reich, Max Kiener, Inken Titz, Kate Vredenburgh, and Thomas Grote participated as commentators.
On the first day, six scholars presented their compelling research projects: In her project titled Digital Manipulation and Domination, Brick argues for de-centralizing the value of autonomy to theorize digital manipulation and points out the inadequacy of the republican concept of freedom and account of domination in the theorization. Kozlovski investigates in The Structure of Normativity and Hard Choices in AI Decision-making the structure of normativity and examines the current and future roles of AI in decision-making, discussing both explicatory and moral implications for AI. Hyska and Barnes, in their paper Interrogating Authenticity as a Norm for Online Speech, observe online platforms where (in)authenticity is used as a norm for social media, and attempt to identify two levels of authenticity—collective and individual—and interrogate the asymmetrical attribution of significance these levels. In Algorithmic Monoculture and Systematic Exclusion, Creel identifies elements that standardize algorithmic decision resulting in homogenization of outcomes and investigates when such homogenization becomes problematic while suggesting normative actions to address it. Calboli emphasizes the importance of paying attention to the characterization of social robot nudges as tools rather than their behavioral objectives in his project Ethical Design of Social Robot Nudgers. In addition to these presentations, Vincent C. Müller delivered a keynote speech on Normative Metacognition in People and AI, in which he discussed the role that normative metacognition – that is, the idea that one can reflect one’s goals and values, and e. g. decide on whether and when to change them – plays in human and AI cognition. This question bears on a variety of normative problems in AI, for example value alignment, frame problems, and what it means to have a goal.
On the second day, two appealing projects were presented, respectively by Parmer and Boulicault, followed by a panel discussion on challenges of generative AI. Parmer’s research project, Automation and the Aftermath of Meaningful Work, attempts to assess the conditions that constitute a practice as meaningful work. He unfolds his account of why value-learning AI cannot engage in meaningful practices. Boulicault critically examines in her project, Precision Medicine, Alzheimer’s Diseases and the Ossification of a Sex Binary, how the approach of precision medicine in Alzheimer’s disease reinforces a sex binary. This can lead, according to her account, to essentializing sex differences and erroneously constructing Alzheimer’s in males and females as different diseases. The workshop ended with a panel discussion about the ethical challenges of generative AI as they are perceived from practitioners from art, industry, and culture. The aim of the concluding panel discussion was to amplify the transdisciplinary nature of much of philosophy of AI, and to draw academic philosophers’ attention to exploring the real-world challenges of generative AI with experts from a range of non-academic fields. Expertise was drawn from various fields outside of academia: panelists included visual artist Roman Lipski, tech educator Ruth Starkman, tech entrepreneur Fabian Stelzer, theatre director Martha Kunicki, and film editor Sebastian Wild. All of them use and employ generative AI in their fields, and reported on their experiences of the transformative power of generative AI.
This event was made possible through the support of the McGovern Fund and the Oxford-Berlin Research Partnership of the Berlin University Alliance.
Written by Imyeon Han and Joy Lehrmann, Freie Universität Berlin