Jeff Behrends

The ongoing explosion of interest in artificial intelligence is fueled in part by recently developed techniques in machine learning. Those techniques allow automated systems to process huge amounts of data, utilizing mathematical methods that depart from traditional statistical approaches, and resulting in impressive advancements in our ability to make predictions and uncover correlations across a host of interesting domains. But as is now widely discussed, the way that those systems arrive at their outputs is often opaque, even to the experts who design and deploy them. Is it morally problematic to make use of opaque automated methods when making high-stakes decisions, like whether to issue a loan to an applicant, or whether to approve a parole request? Many scholars answer in the affirmative. However, there is no widely accepted explanation for why transparent systems are morally preferable to opaque systems. We argue that the use of automated decision-making systems sometimes violates duties of consideration that are owed by decision-makers to decision-subjects, duties that are both epistemic and practical in character. Violations of that kind generate a weighty consideration against the use of opaque decision systems. In the course of defending our approach, we show that it is able to address three major challenges sometimes leveled against attempts to defend the moral import of transparency in automated decision-making.

The particular design of any technology may have profound social implications. Computing technologies are deeply intermeshed with the activities of daily life, playing an ever more central role in how we work, learn, communicate, socialize, and participate in government. Despite the many ways they have improved life, they cannot be regarded as unambiguously beneficial or even value-neutral. Recent experience shows they can lead to unintended but harmful consequences. Some technologies are thought to threaten democracy through the spread of propaganda on online social networks, or to threaten privacy through the aggregation of datasets that include increasingly personal information, or to threaten justice when machine learning is used in such high-stakes, decision-making contexts as loan application reviews, employment procedures, or parole hearings. It is insufficient to ethically assess technology after it has produced negative social impacts, as has happened, for example, with facial recognition software that discriminates against people of color and with self-driving cars that are unable to cope with pedestrians who jay-walk. Developers of new technologies should aim to identify potential harmful consequences early in the design process and take steps to eliminate or mitigate them. This task is not easy. Designers will often have to negotiate among competing values—for instance, between efficiency and accessibility for a diverse user population, or between maximizing benefits and avoiding harm. There is no simple recipe for identifying and solving ethical problems. Computer science education can help meet these challenges by making ethical reasoning about computing technologies a central element in the curriculum. Students can learn to think not only about what technology they could create, but also whether they should create that technology. Learning to reason this way requires courses unlike those currently standard in computer science curricula. A range of university courses on topics in areas of computing, ethics, society and public policy are emerging to meet this need. Some cover computer science broadly, while others focus on specific problems like privacy and security; typically, these classes exist as stand-alone courses in the computer science curriculum. Others have integrated ethics into the teaching of introductory courses on programming, artificial intelligence, and human-computer interaction.

The ongoing explosion of interest in artificial intelligence is fueled in part by recently developed techniques in machine learning. Those techniques allow automated systems to process huge amounts of data, utilizing mathematical methods that depart from traditional statistical approaches, and resulting in impressive advancements in our ability to make predictions and uncover correlations across a host of interesting domains. But as is now widely discussed, the way that those systems arrive at their outputs is often opaque, even to the experts who design and deploy them. Is it morally problematic to make use of opaque automated methods when making high-stakes decisions, like whether to issue a loan to an applicant, or whether to approve a parole request? Many scholars answer in the affirmative. However, there is no widely accepted explanation for why transparent systems are morally preferable to opaque systems. We argue that the use of automated decision-making systems sometimes violates duties of consideration that are owed by decision-makers to decision-subjects, duties that are both epistemic and practical in character. Violations of that kind generate a weighty consideration against the use of opaque decision systems. In the course of defending our approach, we show that it is able to address three major challenges sometimes leveled against attempts to defend the moral import of transparency in automated decision-making.

Internalists about practical reasons maintain that all of an agent’s reasons for action derive their normative force via some relation in which they stand with that agent’s pro-attitudes, or the pro-attitudes that the agent would have in some idealized set of circumstances. One common complaint against internalism is that the view is extensionally inadequate – that it cannot render the correct verdicts about what reasons agents have in a range of important cases. In this paper, I examine that charge of extensional inadequacy, taking as my starting point an argument that Derek Parfit has recently leveled against internalism. Through a close evaluation of that argument and potential replies to it, I attempt to show that internalists cannot accommodate important pre-theoretical intuitions about what reasons we have. However, I also argue that Parfit’s case is importantly overstated; I set out to show that his argument cannot establish, as he thinks it does, that no reasons derive their normative force in the way that internalists believe that all do. In doing so, I draw attention to the possibility of a hybrid position about practical reasons that, surprisingly, receives little attention in the existing literature. If the arguments of the paper succeed, I will have established a modest theoretical advantage for hybridism over internalism – namely, that it is not vulnerable to the charge of extensional inadequacy. My hope is that this goes some way toward establishing hybridism’s credentials as a serious alternative to its “pure” competitors.

Recent calls for reinstituting mandatory home economics education have emphasized its potential to advance gender egalitarian aims. The thought is that, because women’s disproportionate performance of caregiving and household labor is partially caused by gender socialization that better prepares women than men for such work, we can disrupt gender inegalitarian work distributions by preparing everyone for the sort of work in question. The curricula envisioned in these calls are gender-neutral, in the sense that they recommend identical educational interventions for all genders. By exploiting a parallel between gender-neutral educational policies and gender-neutral family leave policies, we argue first that gender-neutral home economics instruction is unlikely to advance gender egalitarian aims, and may in fact reinforce the very outcomes it is meant to disrupt. However, we further argue that a more radical home economics curriculum could avoid these difficulties. To the extent that we value educational interventions as a possible means for advancing gender egalitarian aims, we have good reason to seriously consider adopting a gender non-neutral program in which all and only boys receive mandatory caregiving instruction.

Promotion is the relation between an act and a desire that obtains when the act advances or serves the desire. Under what conditions does an act promote a desire? Probabilistic accounts of promotion, the most prominent accounts, analyze promotion in terms of an increase in the probability of the desire’s satisfaction. In this paper, we clarify the promotion relation and explain why probabilistic accounts are attractive. Then we identify two questions probabilistic accounts must answer: the Baseline Question and the Interpretation Question. We discuss and reject the three answers to the Baseline Question found in the literature, and explain the challenge future attempts at answering this question will face. Proponents of probabilistic accounts have not adequately addressed the Interpretation Question. We survey three answers to this question, finding each unsatisfactory. We conclude that no satisfactory probabilistic account has yet been offered, and that there are significant hurdles to providing one in the future.

In this article, I articulate and respond to an epistemological challenge to meta-normative realism. The challenge has it that, if realism about the normative is correct, and if evolutionary forces have significantly influenced our normative judgments, then it would be a remarkable coincidence if the content of the normative facts and our normative judgments were aligned. I criticize David Enoch's recent attempt to meet this challenge, but provide an alternative response that is structurally similar. I argue that if realism is correct, then it would be remarkable if the content of our normative judgments and the normative facts were not significantly aligned.

Source Hybridism about practical reasons is the position that facts that constitute reasons sometimes derive their normative force from external metaphysical grounds, and sometimes from internal. Although historically less popular than either Source Internalism or Source Externalism, hybridism has lately begun to garner more attention. Here, I further the hybridist's cause by defending Source Hybridism from three objections. I argue that we are not warranted in rejecting hybridism for any of the following reasons: that hybridists cannot provide an account of normative weight, that hybridists are committed to implausible results concerning practical deliberation, or that Source Hybridism is objectionably unparsimonious