EASST/4S conference on "Locating and Timing Matters: Significance and agency of STS in emerging worlds"
Sprache des Tagungstitel:
Modern European Science and Technology has been obsessed from the beginning with elaborating categories of differentiation, discrimination and hierarchization, especially regarding sex, gender, sexuality and race. Although already challenged in the twentieth century by feminist, decolonial and queer STS, these categories have entered computing often unconsciously as patterns of re-cognition.
So, contrary to the promises of high-tech advocates, the mathematical models constituting algorithms and ?artificial intelligence? (AI) do not necessarily provide more objectivity when pervasively installed as automated assistance or even sole method for decision making and discernment in everyday life. Instead, they are accused to reinforce discrimination concerning e.g. schooling, loans, health insurance and even threaten democracy as ?weapons of math destruction? (O?Neil 2012). Automation is analyzed as encoded inequity and amplifying racial hierarchies while appearing neutral (Benjamin 2019). Actually, algorithmic profiling is criticized of discriminating job seekers through privileging (some) men by predicting lower chances for other groups differentiated via demographic data (Allhutter et al. 2020).
The paper asks how feminist and decolonial epistemology can contribute to critical computing and change these problematic patterns of re-cognition. The paper explores and discusses participatory methods for confronting and eventually overcoming tendencies in programming algorithms and systems of AI which perpetuate and amplify biases present in big data sets. The goal is to contribute to a conceptualization of becoming an interdisciplinary high-tech anti-discrimination activist collective.