Race after technology. Abolitionist tools for the new Jim Code

By S. Merz. 2020

If there is one positive consequence stemming from the election of Donald Trump to the White House, it is the surge of critical investigations examining the relationships between big tech and democ...

Subfields: AI Ethics, Fairness & Society, Social Sciences & Humanities

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In AI Ethics, Fairness & Society

In Social Sciences & Humanities

Citations in the corpus

Measurement and Fairness (2019, critical): “Indeed, by using arrests as a proxy for crimes committed, COMPAS can only exacerbate racist policing practices, rather than transcending them [7, 13, 23, 37, 39].”
Diagnosing Gender Bias in Image Recognition Systems (2020, background): “Input data are typically “found data” from the “real world,” containing the biases and cultural associations of human societies, which then get reproduced as “objective” and “scientific” decisions from algorithms (Benjamin 2019).”
The Gradient of Generative AI Release: Methods and Considerations (2023, background): “AI systems can propagate harms such as exacerbating social inequity [64, 7, 41] and harmful biases [20, 13], which can be further amplified in larger systems as scale increases [6].”
Large image datasets: A pyrrhic win for computer vision? (2020, background): “Decades of work within the ﬁelds of Science and Technology Studies (STS) and the Social Sciences show that there is no single straightforward solution to most of the wider social and ethical challenges that we have discussed [5, 25, 76].”
Bringing the People Back In: Contesting Benchmark Machine Learning Datasets (2020, background): “Sociotechnical systems abound in ways that they have failed people of color (Noble, 2018; Benjamin, 2019), women (Bolukbasi et al., 2016), LGBTQ+ communities (Scheuerman et al., 2019), people with disabilities (Hutchinson et al., 2020; Trewin, 2018), and the working class and those in poverty…”
The Values Encoded in Machine Learning Research (2021, background): “Generality, Robustness • For example, the learned linear combination does not consistently outperform either the uniform combination of base kernels or simply the best single base kernel (see, for example, UCI dataset experiments in [9, 12], see also NIPS 2008 workshop).”
You reap what you sow: On the Challenges of Bias Evaluation Under Multilingual Settings (2022, background): “systems interacting with complex social hierarchies (Winner, 1980; Benjamin, 2019; Costanza-Chock, 2018; Friedler et al., 2021).”
On the genealogy of machine learning datasets: A critical history of ImageNet (2021, background): “Yet, there is a growing abundance of examples of how these systems are failing people of color (Noble, 2018; Benjamin, 2019), women (Bolukbasi et al.”
The Use and Misuse of Counterfactuals in Ethical Machine Learning (2021, background): “Social constructivism hence maintains that a social category – be it racial, gender, or class – was brought into existence or shaped by historical events, social forces, political power, or colonial conquest, all of which could have been very different [6, 12, 20].”
Black Feminist Musings on Algorithmic Oppression (2021, background): “diverse team is an inadequate solution to discriminatory design practices that grow out of the interplay of racism and capitalism” [1].”
Algorithmic Folk Theories and Identity: How TikTok Users Co-Produce Knowledge of Identity and Engage in Algorithmic Resistance (2021, methodology): “Ruha Benjamin’s identication of seemingly objective technology furthering and reproducing societies existing inequities as the “New Jim Code” [11] are connected to our ndings highlighting the perceived suppression of marginalized social identities.”
Should ChatGPT be Biased? Challenges and Risks of Bias in Large Language Models (2023, background): “[35, 36] Policy Decisions Developers might implement policies that prevent (or encourage) a given model behavior.”
The Forgotten Margins of AI Ethics (2022, background): “The treatment of concepts such as fairness and bias in abstract terms is frequently linked to the notion of neutrality and objectivity, the idea that there exists a “purely objective” dataset or a “neutral” representation of people, groups, and the social world independent of the observer/modeller [”
Terms of inclusion: Data, discourse, violence (2020, background): “As Ruha Benjamin (2019) summarizes, inclusion is part “of a larger repertoire of ‘happy talk,’ which involves a willingness to acknowledge and even revel in cultural difference without seriously challenging ongoing structural inequality” (p. 148).”
GPT-4V(ision) System Card (2023, background): “[18, 19] OpenAI has thus added refusals for most instances of sensitive trait requests-you can read more about how in Section 2.4.”
Taxonomizing and Measuring Representational Harms: A Look at Image Tagging (2023, background): “Erasure occurs when image tagging systems fail to correctly apply tags that indicate membership in social groups or tags that refer to attributes, artifacts, or activities that are bound up with the identities of social groups (Keyes 2018; Buolamwini and Gebru 2018; Benjamin 2019).”
Intersectional Bias in Causal Language Models (2021, background): “Sociological studies have pointed to ways in which algorithmic bias can amplify existing real-world racism and sexism [18, 25] and reproduce downstream effects of such prejudice with respect to employability, insurability and credit ratings [8, 12].”