The debate about the ethical implications of Artificial Intelligence dates from the 1960s :741–742, 1960; Wiener in Cybernetics: or control and communication in the animal and the machine, MIT Press, New York, 1961). However, in recent years symbolic AI has been complemented and sometimes replaced by Neural Networks and Machine Learning techniques. This has vastly increased its potential utility and impact on society, with the consequence that the ethical debate has gone mainstream. Such a debate has primarily focused on principles—the (...) ‘what’ of AI ethics —rather than on practices, the ‘how.’ Awareness of the potential issues is increasing at a fast rate, but the AI community’s ability to take action to mitigate the associated risks is still at its infancy. Our intention in presenting this research is to contribute to closing the gap between principles and practices by constructing a typology that may help practically-minded developers apply ethics at each stage of the Machine Learning development pipeline, and to signal to researchers where further work is needed. The focus is exclusively on Machine Learning, but it is hoped that the results of this research may be easily applicable to other branches of AI. The article outlines the research method for creating this typology, the initial findings, and provides a summary of future research needs. (shrink)

As the range of potential uses for Artificial Intelligence, in particular machine learning, has increased, so has awareness of the associated ethical issues. This increased awareness has led to the realisation that existing legislation and regulation provides insufficient protection to individuals, groups, society, and the environment from AI harms. In response to this realisation, there has been a proliferation of principle-based ethics codes, guidelines and frameworks. However, it has become increasingly clear that a significant gap exists between the theory of (...) AI ethics principles and the practical design of AI systems. In previous work, we analysed whether it is possible to close this gap between the ‘what’ and the ‘how’ of AI ethics through the use of tools and methods designed to help AI developers, engineers, and designers translate principles into practice. We concluded that this method of closure is currently ineffective as almost all existing translational tools and methods are either too flexible or too strict. This raised the question: if, even with technical guidance, AI ethics is challenging to embed in the process of algorithmic design, is the entire pro-ethical design endeavour rendered futile? And, if no, then how can AI ethics be made useful for AI practitioners? This is the question we seek to address here by exploring why principles and technical translational tools are still needed even if they are limited, and how these limitations can be potentially overcome by providing theoretical grounding of a concept that has been termed ‘Ethics as a Service.’. (shrink)

Common-sense allows that talk about moral truths makes perfect sense. If you object to the United States’ Declaration of Independence’s assertion that it is a truth that ‘all men’ are ‘endowed by their Creator with certain unalienable Rights’, you are more likely to object that these rights are not unalienable or that they are not endowed by the Creator, or even that its wording ignores the fact that women have rights too, than that this is not the sort of thing (...) which could be a truth. Whether it is a truth or not seems beside the point, anyway; the writers of the Declaration could just have well written, ‘We hold it to be self-evident that all men are created equal, and also that it is self-evident that all men are endowed by their Creator with certain unalienable Rights,’ save only that its cadence would lack some of the poetic resonance of the version which garnered Hancock’s signature. Yet famously, ethical noncognitivists have proclaimed that moral sentences can’t be true or false – that, like ‘Hooray!’ or ‘dammit!’, they are not even the kinds of things to be true or false. Noncognitivism is sometimes even defined as the view that this is so, but even philosophers who define ‘noncognitivism’ more broadly, as consistent with the idea that moral sentences may be true or false, have believed that they needed to do important philosophical spadework in order to make sense of how moral sentences could be true or false. In this article we’ll look at the puzzle about moral truth as it is faced by early noncognitivists and by metaethical expressivists, the early noncognitivists’ contemporary cousins. We’ll look at what it would take for expressivists to ‘earn the right’ to talk about moral truths at all, and in particular at what it would take for them to earn the right to claim that moral truths behave in the ways that we should expect – including that meaningful moral sentences which lack presuppositions are true or false, and that classically valid arguments are truth-preserving.. (shrink)

What does it mean to say that music perception is nonconceptual? As the passages from Meyer and Budd illustrate, one frequently encounters claims of this kind: it is often suggested that there is a level of perceptual contact with, or understanding or enjoyment of, music—one in which listeners typically engage—that does not require conceptualization. But just what does a claim of this sort amount to, and what arguments may be adduced for it? And is all musical hearing nonconceptual, or are (...) there ways of enjoying fugues and sonatas that do involve conceptualizing cadences and contrapuntal devices? If the latter, how are we to conceive of the relationship between conceptual and nonconceptual ways of hearing? (shrink)

Les recherches scientifiques sur la COVID-19 sont à la fois menées et diffusées à une cadence effrénée. Bien qu’il soit inspirant de voir la communauté de la recherche répondre avec autant de vigueur à la crise causée par la pandémie, toute cette activité a par ailleurs engendré un chaos de mauvaises données, de résultats contradictoires et de manchettes exagérées. Alors que la polarisation, la déformation et la médiatisation des résultats scientifiques s’intensifient chaque jour, les inquiétudes se font de plus (...) en plus sentir quant à la perspective que la science pertinente soit présentée au public d’une manière qui puisse causer de la confusion, créer de fausses attentes et éroder la confiance du public. Dans cette note, nous explorons les principaux enjeux associés à la présentation de la science dans le contexte de la pandémie de la COVID-19. Plusieurs de ces enjeux ne sont pas nouveaux. Mais la pandémie de la COVID-19 a braqué les projecteurs sur le processus de la recherche biomédicale et a amplifié les ramifications néfastes des problèmes de communication publique. Nous devons faire mieux. À ce titre, nous conclurons ce rapport en formulant dix recommandations qui s’adressent aux acteurs clés qui interviennent dans la communication de la science sur la COVID-19, notamment les gouvernements, les bailleurs de fonds, les universités, les éditeurs, les médias et les communautés de recherche. (shrink)