We are sceptical of concerns over the opacity of algorithmic decision tools. While transparency and explainability are certainly important desiderata in algorithmic governance, we worry that automated decision-making is being held to an unrealistically high standard, possibly owing to an unrealistically high estimate of the degree of transparency attainable from human decision-makers. In this paper, we review evidence demonstrating that much human decision-making is fraught with transparency problems, show in what respects AI fares little worse or better and argue that (...) at least some regulatory proposals for explainable AI could end up setting the bar higher than is necessary or indeed helpful. The demands of practical reason require the justification of action to be pitched at the level of practical reason. Decision tools that support or supplant practical reasoning should not be expected to aim higher than this. We cast this desideratum in terms of Daniel Dennett’s theory of the “intentional stance” and argue that since the justification of action for human purposes takes the form of intentional stance explanation, the justification of algorithmic decisions should take the same form. In practice, this means that the sorts of explanations for algorithmic decisions that are analogous to intentional stance explanations should be preferred over ones that aim at the architectural innards of a decision tool. (shrink)

There is widespread agreement that there should be a principle requiring that artificial intelligence be ‘explicable’. Microsoft, Google, the World Economic Forum, the draft AI ethics guidelines for the EU commission, etc. all include a principle for AI that falls under the umbrella of ‘explicability’. Roughly, the principle states that “for AI to promote and not constrain human autonomy, our ‘decision about who should decide’ must be informed by knowledge of how AI would act instead of us” :689–707, 2018). There (...) is a strong intuition that if an algorithm decides, for example, whether to give someone a loan, then that algorithm should be explicable. I argue here, however, that such a principle is misdirected. The property of requiring explicability should attach to a particular action or decision rather than the entity making that decision. It is the context and the potential harm resulting from decisions that drive the moral need for explicability—not the process by which decisions are reached. Related to this is the fact that AI is used for many low-risk purposes for which it would be unnecessary to require that it be explicable. A principle requiring explicability would prevent us from reaping the benefits of AI used in these situations. Finally, the explanations given by explicable AI are only fruitful if we already know which considerations are acceptable for the decision at hand. If we already have these considerations, then there is no need to use contemporary AI algorithms because standard automation would be available. In other words, a principle of explicability for AI makes the use of AI redundant. (shrink)

The role of values in scientific research has become an important topic of discussion in both scholarly and popular debates. Pundits across the political spectrum worry that research on topics like climate change, evolutionary theory, vaccine safety, and genetically modified foods has become overly politicized. At the same time, it is clear that values play an important role in science by limiting unethical forms of research and by deciding what areas of research have the greatest relevance for society. Deciding how (...) to distinguish legitimate and illegitimate influences of values in scientific research is a matter of vital importance.Recently, philosophers of science have written a great deal on this topic, but most of their work has been directed toward a scholarly audience. This book makes the contemporary philosophical literature on science and values accessible to a wide readership. It examines case studies from a variety of research areas, including climate science, anthropology, chemical risk assessment, ecology, neurobiology, biomedical research, and agriculture. These cases show that values have necessary roles to play in identifying research topics, choosing research questions, determining the aims of inquiry, responding to uncertainty, and deciding how to communicate information. Kevin Elliott focuses not just on describing roles for values but also on determining when their influences are actually appropriate. He emphasizes several conditions for incorporating values in a legitimate fashion, and highlights multiple strategies for fostering engagement between stakeholders so that value influences can be subjected to careful and critical scrutiny. (shrink)

The classic logical positivist account of historical explanation, putting forward what is variously called the "regularity interpretation" (#Gardiner, The Nature of Historical Explanation), the "covering law model" (#Dray, Laws and Explanation in History), or the "deductive model" (Michael #Scriven, "Truisms as Grounds for Historical Explanations"). See also #Danto, Narration and Knowledge, for further criticisms of the model. Hempel formalizes historical explanation as involving (a) statements of determining (initial and boundary) conditions for the event to be explained, and (b) statements of (...) general laws that imply that whenever events of the kind described in (a) occur, an event of the kind to be explained will take place. He admits that in practice historians rarely explain historical events in this way, but instead give only fragmentary "explanation sketches." Such sketches can, however, be "scientifically acceptable," providing it points to where "more specific statements" are to be found (351). An interesting aspect of the article is its "late" positivist, proto‑Kuhnian assertion that "the separation of 'pure description' and 'hypothetical generalization and theory construction' in empirical science is unwarranted" (356). But Hempel does not work out the wider implications of this important conclusion. It is also worth noting that he himself was not much interested in the philosophy of history; other authors took up his model and discussed it within that context. Usefully discussed by #Ricoeur, Time and Narrative, 1:112‑17. (shrink)

The idea that human thinking and machine computing are "radically the same" provides the central theme for this marvelously lucid and witty book on...

Fricker shows that virtue epistemology provides a general epistemological idiom in which these issues can be forcefully discussed.

This article reports the findings of AI4People, an Atomium—EISMD initiative designed to lay the foundations for a “Good AI Society”. We introduce the core opportunities and risks of AI for society; present a synthesis of five ethical principles that should undergird its development and adoption; and offer 20 concrete recommendations—to assess, to develop, to incentivise, and to support good AI—which in some cases may be undertaken directly by national or supranational policy makers, while in others may be led by other (...) stakeholders. If adopted, these recommendations would serve as a firm foundation for the establishment of a Good AI Society. (shrink)