AI Seminar

Inference and the “Racial Drama” of AI

Dr. Ramon AmaroLecturer (Assistant Professor equiv.) in Art & Visual Cultures of the Global SouthUniversity College London (UCL)


Zoom (Zoom link; Password if needed: MichiganAI)


This talk discusses logical inference and the assessment of racial discrimination in artificial intelligence research. Because discriminatory behaviours can rarely be directly observed, researchers are faced with the challenge of identifying the presence or absence of racial and ethnic bias in data sets, and therefore rely on the observation of the human face to infer race and ethnicity based on skin color and other assumed equivalent characteristics. While inference rules are important in the development of artificial intelligence research, in that they provide general templates from which applied proofs take shape, I argue that this approach compounds statements of racial rigidity by prioritising logical judgement, thereby reinforcing the superficiality of equivalency between race, racism, and phenotypical characteristics. In order to shift the base assumptions of equivalency, this talk discusses the role of inference in AI research, along with what Frantz Fanon calls the “racial drama” of identity staging in inferred scientific analysis. 


Dr. Ramon Amaro’s writing, research and practice emerge at the intersections of Black Study, psychopathology, digital culture, and the critique of computation reason. He draws on Frantz Fanon’s theory of sociogenic alienation to problematize the de-localisation of the Black psyché in contemporary computational systems, such as machine learning and artificial intelligence. Ramon’s research pulls away from notions of psychic negation, as set forth by the Fanonian model of representation, to investigate alternative modes of relation between humans, race and technology. His ultimate aim is to develop new methodologies for the study of human centered technology. Dr. Amaro holds a PhD. in Philosophy from Goldsmiths (University of London), an MA in Sociological Research Methods from the University of Essex, and a BSe. in Mechanical Engineering from the University of Michigan, Ann Arbor. Dr. Amaro’s forthcoming monograph “The Black Technical Object: On Machine Learning and the Aspiration of Black Being” (Sternberg, 2022) is his first major work on the abstruse nature of machine learning, mathematics, and the deep incursion of racial hierarchy.



AI Lab