Diversity Reading List

Comment: This paper offers a clear overview of the literature on the right to explanation and counters the mainstream view that, in the context of automated decision-making technology, that we hold such a right. It would therefore offer a useful introduction to ideas about explanability in relation to the ethics of AI and automated technologies, and could be used in a reading group context as well as in upper undergraduate and graduate level courses.

Comment: This paper asserts a right to explanation grounded in an interest in informed self-advocacy, the term the author uses to describe a cluster of abilities to represent one's interests and values to decision-makers and to further those interests and values within an institution. Vredenburgh also argues that such form of self-advocacy are necessary for hierarchical, non-voluntary institutions to be legitimate and fair - and it is on these grounds that a person may reasonably reject insitutional set-ups that prevent them from engaging in these abilities. In this sense, Vredenburgh's argument applies to a broader set of problems then simply algorithmic opacity - they may feasibly be applied to cases in which systems (such as bureacratic ones) deny an individual this right to explanation. Therefore, this paper presents an argument which would be useful as further or specialised reading in a variety of classroom contexts, including courses or reading groups addressing technological and algorithmic ethics, basic political rights, bureacratic ethics, as well as more general social and political philosophical courses. It might be interesting, for example, to use it to in an introductory social/political course to discuss with students some of the ethical questions that are particular to a 21st century context. As systems become more complex and individuals become further removed from the institutional decision-making that guides/rules/directs their lives, what right do we have to understand the processes that condition our experience? In what other situations might these rights become challenged?

Comment: This paper offers a novel approach to the exploration of alienation at work (i.e., what makes work bad) from an algorithmic ethics perspective. It relies on the noninstrumental value of explanation to make its central argument, and grounds this value in the role that explanation plays in our ability to form a practical orientation towards our scoial world. In this sense, it examines an interesting, and somewhat underexplored, connection between algorithmic ethics, justice, the future of work, and social capabilities. As such, it could be useful in a wide range of course contexts. This being said, the central argument is fairly complex, and relies on some previous understanding of analytic political philosophy and philosophy of AI. It also employs technical language from these domains, and therefore would be best utilised for masters-level or other advanced philosophical courses and study.

Comment (from this Blueprint): In this 2017 paper, historian Joanna Radin explores how reusing big data can contribute to the continued subjugation of Akimel O’odham, who live in the southewestern region of the US, otherwise known as the "Pima". This reading also illustrates how data can, over time, become used for what it was never intended or collected for. Radin emphasizes the dangers of forgetting that data represent human beings.

Comment (from this Blueprint): This is a written transcript of the James Baldwin lecture, delivered by the computer scientist Arvind Narayanan, at Princeton in 2022. Narayanan's prior research has examined algorithmic bias and standards of fairness with respect to algorithmic decision making. Here, he engages critically with his own discipline, suggesting that there are serious limits to the sorts of quantitative methods that computer scientists recruit to investigate the potential biases in their own tools. Narayanan acknowledges that in voicing this critique, he is echoing claims by feminist researchers from fields beyond computer science. However, his own arguments, centered as they are on the details of the quantitative methods he is at home with, home in on exactly why these prior criticisms hold up in a way that seeks to speak more persuasively to Narayanan's own peers in computer science and other quantitative fields.

Comment (from this Blueprint): In this paper, Martin and Pease look at how mathematics happens online, emphasising how this embodies the picture of mathematics given by Polya and Lakatos, two central figures in philosophy of mathematical practice. They look at multiple venues of online mathematics, including the polymath projects of collaborative problem-solving, and mathoverflow, which is a question-and-answer forum. By looking at the discussions that take place when people are doing maths online, they argue that you can get rich new kinds of data about the processes of mathematical discovery and understanding. They discuss how online mathematics can become a “social machine”, and how this can open up new ways of doing mathematics.

Comment (from this Blueprint): Dick traces the history of the AURA automated reasoning assistant in the 1970s and 80s, arguing that the introduction of the computer system led to novel contributions to mathematics by unconventional means. Dick’s emphasis is on the AURA system as changing the material culture of mathematics, and thereby leading to collaboration and even negotiations between the mathematicians and the computer system.

Comment: This is a paper that gives an account of enhancement and disability in terms of one's relative position on a harmed and benefitted continuum, and defends enhancement on completely general moral grounds according to which the pro tanto duty to enhance is the same as the pro tanto duty not to disable. It pairs well with criticisms of the 'new eugenics', such as Robert Sparrow's 'A Not-So-New Eugenics' (2011) and more generally with consequentialist or specifically harm-based accounts of moral obligation.

Comment: Argues that the automatism inherent in the production of a photograph has made it possible for artists to extend the tradition of self-portraiture in a way that is radically different from previous visual arts. Demonstrates that automatism need not stand in competition or conflict with artistic agency.

Artworks to use with this text:

Ilse Bing, Self-portrait with Leica (1931)

It is usual for portraits to show a person's head either in profile or in a frontal position, but this self-portrait shows both alternatives simultaneously. It also depicts the presence of two mirrors in such a way that we are in a position to judge that the camera has recorded its own reflection. Thus, we see both the face of the artist and the "face" of the camera: it is a double self-portrait. Argues that the automatism inherent in the production of a photograph has made it possible for artists to extend the tradition of self-portraiture in a way that is radically different from previous visual arts. Demonstrates that automatism need not stand in competition or conflict with artistic agency.

Artworks to use with this text:

Ilse Bing, Self-portrait with Leica (1931)

It is usual for portraits to show a person's head either in profile or in a frontal position, but this self-portrait shows both alternatives simultaneously. It also depicts the presence of two mirrors in such a way that we are in a position to judge that the camera has recorded its own reflection. Thus, we see both the face of the artist and the "face" of the camera: it is a double self-portrait.

Comment: Good as further reading for philosophy of science courses or as introductory reading for courses specialized in philosophy of technology. It is an easy paper but the topic is very specific, so in this last sense it is more suitable for postgraduates.