Is explainable artificial intelligence intrinsically valuable?

Colaner, Nathan

doi:10.1007/s00146-021-01184-2

Cited by 29 publications

(11 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Studies have also shown that AI's predictive capabilities may be exaggerated (Dressel and Farid, 2018;Jung et al, 2017;Salganik et al, 2020). AI's opacity makes algorithmic systems difficult to interrogate and hold accountable (Colaner, 2021;Innerarity, 2021;Loi et al, 2020). Algorithmic systems also incent surveillance and data collection because they need large information sets for model training and analysis.…”

Section: The Functions and Risks Of Algorithmic Decision-makingmentioning

confidence: 99%

“…In a similar vein, scholars suggest that humans in the loop can fix errors and place guard rails around absurd, unethical, or inappropriate results (Henderson, 2018; Jones, 2017; Rahwan, 2018). Studies have measured the relative importance of transparency in using algorithms (König et al, 2022) and find explainability may not be intrinsically valuable (Colaner, 2021). Other scholars focus on impact statements (Katyal, 2019; Metcalf et al, 2021; Reisman et al, 2018) modeled after environmental or privacy impact assessments, or ex ante transparency as to goals and metrics (Loi et al (2020) to document and assess a system's fairness.…”

Section: Theorymentioning

confidence: 99%

See 1 more Smart Citation

Governing algorithmic decisions: The role of decision importance and governance on perceived legitimacy of algorithmic decisions

Waldman

Martin

2022

Big Data & Society

View full text Add to dashboard Cite

The algorithmic accountability literature to date has primarily focused on procedural tools to govern automated decision-making systems. That prescriptive literature elides a fundamentally empirical question: whether and under what circumstances, if any, is the use of algorithmic systems to make public policy decisions perceived as legitimate? The present study begins to answer this question. Using factorial vignette survey methodology, we explore the relative importance of the type of decision, the procedural governance, the input data used, and outcome errors on perceptions of the legitimacy of algorithmic public policy decisions as compared to similar human decisions. Among other findings, we find that the type of decision—low importance versus high importance—impacts the perceived legitimacy of automated decisions. We find that human governance of algorithmic systems (aka human-in-the-loop) increases perceptions of the legitimacy of algorithmic decision-making systems, even when those decisions are likely to result in significant errors. Notably, we also find the penalty to perceived legitimacy is greater when human decision-makers make mistakes than when algorithmic systems make the same errors. The positive impact on perceived legitimacy from governance—such as human-in-the-loop—is greatest for highly pivotal decisions such as parole, policing, and healthcare. After discussing the study’s limitations, we outline avenues for future research.

show abstract

Section: The Functions and Risks Of Algorithmic Decision-makingmentioning

confidence: 99%

Section: Theorymentioning

confidence: 99%

Governing algorithmic decisions: The role of decision importance and governance on perceived legitimacy of algorithmic decisions

Waldman

Martin

2022

Big Data & Society

View full text Add to dashboard Cite

show abstract

“…XAI is important, like informed consent, because it provides data subjects with a sense of awareness about how their data are being used. Providing a meaningful explanation of how one’s data are being used can be humanising because it gives data subjects a greater sense of control, with respect to their information (Colaner 2021 ). XAI promises data users, ‘fairness, trust, and governability’ (Colaner 2021 ).…”

Section: Big Data and Informed Consentmentioning

confidence: 99%

“…Providing a meaningful explanation of how one’s data are being used can be humanising because it gives data subjects a greater sense of control, with respect to their information (Colaner 2021 ). XAI promises data users, ‘fairness, trust, and governability’ (Colaner 2021 ).…”

Section: Big Data and Informed Consentmentioning

confidence: 99%

AI, big data, and the future of consent

2021

View full text Add to dashboard Cite

In this paper, we discuss several problems with current Big data practices which, we claim, seriously erode the role of informed consent as it pertains to the use of personal information. To illustrate these problems, we consider how the notion of informed consent has been understood and operationalised in the ethical regulation of biomedical research (and medical practices, more broadly) and compare this with current Big data practices. We do so by first discussing three types of problems that can impede informed consent with respect to Big data use. First, we discuss the transparency (or explanation) problem. Second, we discuss the re-repurposed data problem. Third, we discuss the meaningful alternatives problem. In the final section of the paper, we suggest some solutions to these problems. In particular, we propose that the use of personal data for commercial and administrative objectives could be subject to a ‘soft governance’ ethical regulation, akin to the way that all projects involving human participants (e.g., social science projects, human medical data and tissue use) are regulated in Australia through the Human Research Ethics Committees (HRECs). We also consider alternatives to the standard consent forms, and privacy policies, that could make use of some of the latest research focussed on the usability of pictorial legal contracts.

show abstract

“…Having that said, it can even be questioned whether explainability is intrinsically valuable at all [53]. An AI system that is explainable in human terms can be seen as mere means to an end, namely, to foster fairness, trust, accountability, or individual control over decision-making processes.…”

Section: Example 1: Explainable Ai Systemsmentioning

confidence: 99%

Blind spots in AI ethics

Hagendorff

2021

AI Ethics

View full text Add to dashboard Cite

This paper critically discusses blind spots in AI ethics. AI ethics discourses typically stick to a certain set of topics concerning principles evolving mainly around explainability, fairness, and privacy. All these principles can be framed in a way that enables their operationalization by technical means. However, this requires stripping down the multidimensionality of very complex social constructs to something that is idealized, measurable, and calculable. Consequently, rather conservative, mainstream notions of the mentioned principles are conveyed, whereas critical research, alternative perspectives, and non-ideal approaches are largely neglected. Hence, one part of the paper considers specific blind spots regarding the very topics AI ethics focusses on. The other part, then, critically discusses blind spots regarding to topics that hold significant ethical importance but are hardly or not discussed at all in AI ethics. Here, the paper focuses on negative externalities of AI systems, exemplarily discussing the casualization of clickwork, AI ethics’ strict anthropocentrism, and AI’s environmental impact. Ultimately, the paper is intended to be a critical commentary on the ongoing development of the field of AI ethics. It makes the case for a rediscovery of the strength of ethics in the AI field, namely its sensitivity to suffering and harms that are caused by and connected to AI technologies.

show abstract

Is explainable artificial intelligence intrinsically valuable?

Cited by 29 publications

References 20 publications

Governing algorithmic decisions: The role of decision importance and governance on perceived legitimacy of algorithmic decisions

Governing algorithmic decisions: The role of decision importance and governance on perceived legitimacy of algorithmic decisions

AI, big data, and the future of consent

Blind spots in AI ethics

Contact Info

Product

Resources

About