Bias In, Bias Out

How researchers see this paper

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

A Causal Bayesian Networks Viewpoint on Fairness (2018, neutral): “Multiple studies [22,34,37,46] have argued that the core premise of RAIs, to assess the likelihood a defendant reoffends, is impossible to measure and that the empirical proxy used (e.”
It's COMPASlicated: The Messy Relationship between RAI Datasets and Algorithmic Fairness Benchmarks (2021, background): “Critics have raised concerns that RAIs themselves are not racially neutral and will contribute to racist decision-making, lack mechanisms of transparency for defendants or accountability for decision-makers, and overstate their applicability in the CJ context [15, 16, 17, 18].”
Out of One, Many: Using Language Models to Simulate Human Samples (2022, background): “Our conception of algorithmic fidelity goes beyond prior observations that language models reflect human-likebiasespresent in the text corporaused to create them (Barocas andSelbst 2016; Caliskan, Bryson, and Narayanan 2017; Garg et al. 2018; Mayson 2018; Panch et al. 2019).”
The ethics of algorithms: key problems and solutions (2020, unknown): “For these reasons, the use of algorithms in some settings is questioned altogether (Selbst et al. 2019; Mayson 2019; Katell et al. 2020; Abebe et al. 2020).”