The goal of this article is to present a first list of ethical concerns that may arise from research and personal use of virtual reality (VR) and related technology, and to offer concrete recommendations for minimizing those risks. Many of the recommendations call for focused research initiatives. In the first part of the article, we discuss the relevant evidence from psychology that motivates our concerns. In Section “Plasticity in the Human Mind,” we cover some of the main results suggesting that (...) one’s environment can influence one’s psychological states, as well as recent work on inducing illusions of embodiment. Then, in Section “Illusions of Embodiment and Their Lasting Effect,” we go on to discuss recent evidence indicating that immersion in VR can have psychological effects that last after leaving the virtual environment. In the second part of the article, we turn to the risks and recommendations. We begin, in Section “The Research Ethics of VR,” with the research ethics of VR, covering six main topics: the limits of experimental environments, informed consent, clinical risks, dual-use, online research, and a general point about the limitations of a code of conduct for research. Then, in Section “Risks for Individuals and Society,” we turn to the risks of VR for the general public, covering four main topics: long-term immersion, neglect of the social and physical environment, risky content, and privacy. We offer concrete recommendations for each of these 10 topics, summarized in Table 1. (shrink)

To create fair and accountable AI and robotics, we need precise regulation and better methods to certify, explain, and audit inscrutable systems.

Research suggests that interventions involving extensive training or counterconditioning can reduce implicit prejudice and stereotyping, and even susceptibility to stereotype threat. This research is widely cited as providing an “existence proof” that certain entrenched social attitudes are capable of change, but is summarily dismissed—by philosophers, psychologists, and activists alike—as lacking direct, practical import for the broader struggle against prejudice, discrimination, and inequality. Criticisms of these “debiasing” procedures fall into three categories: concerns about empirical efficacy, about practical feasibility, and about the (...) failure to appreciate the underlying structural-institutional nature of discrimination. I reply to these criticisms of debiasing, and argue that a comprehensive strategy for combating prejudice and discrimination should include a central role for training our biases away. (shrink)

Luciano Floridi develops the first ethical framework for dealing with the new challenges posed by Information and Communication Technologies. He establishes the conceptual foundations of Information Ethics by exploring important metatheoretical and introductory issues, and answering key theoretical questions of great philosophical interest.

This paper discusses a crisis of accountability that arises when scientific collaborations are massively epistemically distributed. We argue that social models of epistemic collaboration, which are social analogs to what Patrick Suppes called a “model of the experiment,” must play a role in creating accountability in these contexts. We also argue that these social models must accommodate the fact that the various agents in a collaborative project often have ineliminable, messy, and conflicting interests and values; any story about accountability in (...) a massively distributed collaboration must therefore involve models of such interests and values and their methodological and epistemic effects. (shrink)

In this collection of previously published essays, Sally Haslanger draws on insights from feminist and critical race theory and on the resources of contemporary analytic philosophy to develop the idea that gender and race are positions ...

This article argues that an ethical framework will help to harness the potential of AI while keeping humans in control.

Artificial Intelligence (AI) is already having a major impact on society. As a result, many organizations have launched a wide range of initiatives to establish ethical principles for the adoption of socially beneficial AI. Unfortunately, the sheer volume of proposed principles threatens to overwhelm and confuse. How might this problem of ‘principle proliferation’ be solved? In this paper, we report the results of a fine-grained analysis of several of the highest-profile sets of ethical principles for AI. We assess whether these (...) principles converge upon a set of agreed-upon principles, or diverge, with significant disagreement over what constitutes ‘ethical AI.’ Our analysis finds a high degree of overlap among the sets of principles we analyze. We then identify an overarching framework consisting of five core principles for ethical AI. Four of them are core principles commonly used in bioethics: beneficence, non-maleficence, autonomy, and justice. On the basis of our comparative analysis, we argue that a new principle is needed in addition: explicability, understood as incorporating both the epistemological sense of intelligibility (as an answer to the question ‘how does it work?’) and in the ethical sense of accountability (as an answer to the question: ‘who is responsible for the way it works?’). In the ensuing discussion, we note the limitations and assess the implications of this ethical framework for future efforts to create laws, rules, technical standards, and best practices for ethical AI in a wide range of contexts. (shrink)

Should you be targeted by police for a crime that AI predicts you will commit? In this paper, we analyse when, and to what extent, the person-based predictive policing (PP) — using AI technology to identify and handle individuals who are likely to breach the law — could be justifiably employed. We first examine PP’s epistemological limits, and then argue that these defects by no means refrain from its usage; they are worse in humans. Next, based on major AI ethics (...) guidelines (IEEE, EU, and RIKEN, etc.), we refine three basic moral principles specific to person-based PP. We also derive further requirements from case studies, including debates in Chicago, New Orleans, San Francisco, Tokyo, and cities in China. Instead of rejecting PP programs, we analyse what necessary conditions should be met for using the tool to achieve social good. While acknowledging its risks, we conclude that the person-based PP could be beneficial in community policing, especially when merging into a larger governance framework of the social safety net. (shrink)

In analyzing oppressive systems like racism, social theorists have articulated accounts of the dynamic interaction and mutual dependence between psychological components, such as individuals’ patterns of thought and action, and social components, such as formal institutions and informal interactions. We argue for the further inclusion of physical components, such as material artifacts and spatial environments. Drawing on socially situated and ecologically embedded approaches in the cognitive sciences, we argue that physical components of racism are not only shaped by, but also (...) shape psychological and social components of racism. Indeed, while our initial focus is on racism and racist things, we contend that our framework is also applicable to other oppressive systems, including sexism, classism, and ableism. This is because racist things are part of a broader class of oppressive things, which are material artifacts and spatial environments that are in congruence with an oppressive system. (shrink)

This book deals with the problem of dual-use science research and technology. It first explains the concept of dual use and then offers analyses of collective knowledge and collective ignorance. It goes on to present a theory of collective responsibility, followed by four chapters focusing on a particular scientific field or industry of dual use concern: the chemical industry, the nuclear industry, cyber-technology and the biological sciences. The problem of dual-use science research and technology arises because such research and technology (...) has the potential to be used for great evil as well as for great good. On the one hand, knowledge is a necessary condition, and perhaps a constitutive feature, of technologies that contribute greatly to individual and collective well-being. Consider, for example, nuclear technology that enables the generation of low cost electricity in populations without obvious alternative energy sources. So technological knowledge is a good thing and ignorance of it a bad thing. On the other hand, these same technologies can be extremely harmful to individuals and collectives, as with the atomic bombs dropped on Hiroshima and Nagasaki. So, at least with respect to some technologies evidently knowledge is a bad thing and ignorance a good thing. Accordingly, the question arises as to whether we ought to limit scientific research and/or the development of technology and, if so, which research or technology, in what manner and to what extent. This book examines the answer to that question. (shrink)

One of the most pervasive criticisms of nudges has been the claim that they violate, undermine or decrease people’s autonomy. This claim, however, is seldom backed up by an explicit and detailed conception of autonomy. In this paper, we aim to do three things. First, we want to clear up some conceptual confusion by distinguishing the different conceptions used by Cass Sunstein and his critics in order to get clear on how they conceive of autonomy. Second, we want to add (...) to the existing discussion by distinguishing between ‘autonomy’ as the ability to set your own ends and ‘autocracy’ as the ability to actually realize those ends. This will allow for a more careful ethical evaluation of specific nudge interventions. Third, we will introduce the idea of ‘perimeters of autonomy’ in an attempt to provide a realistic account of personal autonomy and we will argue that it can alleviate most of the worries about nudging being autonomy-undermining. (shrink)

Background Implicit biases are present in the general population and among professionals in various domains, where they can lead to discrimination. Many interventions are used to reduce implicit bias. However, uncertainties remain as to their effectiveness. -/- Methods We conducted a systematic review by searching ERIC, PUBMED and PSYCHINFO for peer-reviewed studies conducted on adults between May 2005 and April 2015, testing interventions designed to reduce implicit bias, with results measured using the Implicit Association Test (IAT) or sufficiently similar methods. (...) -/- Results 30 articles were identified as eligible. Some techniques, such as engaging with others’ perspective, appear unfruitful, at least in short term implicit bias reduction, while other techniques, such as exposure to counterstereotypical exemplars, are more promising. Robust data is lacking for many of these interventions. -/- Conclusions Caution is thus advised when it comes to programs aiming at reducing biases. This does not weaken the case for implementing widespread structural and institutional changes that are multiply justified. (shrink)

As the phenomenon of implicit bias has become increasingly widely known and accepted, a variety of criticisms have similarly gained in prominence. This paper focuses on one particular set of criticisms, generally made from the political left, of what Sally Haslanger calls “implicit bias stories”—a broad term encompassing a wide range of discourses from media discussions to academic papers to implicit bias training. According to this line of thought, implicit bias stories are counterproductive because they serve to distract from the (...) structural and institutional factors that underlie oppression of social groups. This paper argues on the contrary that implicit bias stories, properly told, can help direct attention and concern to structural and institutional factors, and indeed may be especially helpful in motiving action. The key, however, is to tell these stories properly. When implicit bias sto- ries are told in the wrong way, they are indeed counterproductive. This paper looks in detail at several examples of good and bad implicit bias stories, examining what makes some of them counterproductive and others highly effective in motivating action to combat social injustice. (shrink)

“Implicit bias” is a term of art referring to relatively unconscious and relatively automatic features of prejudiced judgment and social behavior. While psychologists in the field of “implicit social cognition” study “implicit attitudes” toward consumer products, self-esteem, food, alcohol, political values, and more, the most striking and well-known research has focused on implicit attitudes toward members of socially stigmatized groups, such as African-Americans, women, and the LGBTQ community.[1] For example, imagine Frank, who explicitly believes that women and men are equally (...) suited for careers outside the home. Despite his explicitly egalitarian belief, Frank might nevertheless implicitly associate women with the home, and this implicit association might lead him to behave in any number of biased ways, from trusting feedback from female co-workers less to hiring equally qualified men over women. Psychological research on implicit bias is relatively recent (§1), but a host of metaphysical (§2), epistemological (§3), and ethical questions (§4) about implicit bias are pressing.[2]. (shrink)

The moral enhancement of human beings is a constant theme in the history of humanity. Today, faced with the threats of a new, globalised world, concern over this matter is more pressing. For this reason, the use of biotechnology to make human beings more moral has been considered. However, this approach is dangerous and very controversial. The purpose of this article is to argue that the use of another new technology, AI, would be preferable to achieve this goal. Whilst several (...) proposals have been made on how to use AI for moral enhancement, we present an alternative that we argue to be superior to other proposals that have been developed. (shrink)

Sally Haslanger has recently argued that philosophical focus on implicit bias is overly individualist, since social inequalities are best explained in terms of social structures rather than the actions and attitudes of individuals. I argue that questions of individual responsibility and implicit bias, properly understood, do constitute an important part of addressing structural injustice, and I propose an alternative conception of social structure according to which implicit biases are themselves best understood as a special type of structure.

The central contention of The Implicit Mind is that understanding the two faces of spontaneity-its virtues and vices-requires understanding the "implicit mind." In turn, Michael Brownstein maintains that understanding the implicit mind requires the consideration of three sets of questions. First, what are implicit mental states? What kind of cognitive structure do they have? Second, how should we relate to our implicit attitudes? Are we responsible for them? Third, how can we improve the ethics of our implicit minds?

How do cognition and affect interact to produce action? Research in intergroup psychology illuminates this question by investigating the relationship between stereotypes and prejudices about social groups. Yet it is now clear that many social attitudes are implicit. This raises the question: how does the distinction between cognition and affect apply to implicit mental states? An influential view—roughly analogous to a Humean theory of action—is that “implicit stereotypes” and “implicit prejudices” constitute two separate constructs, reflecting different mental processes and neural (...) systems. On this basis, some have also argued that interventions to reduce discrimination should combat implicit stereotypes and prejudices separately. We propose an alternative framework. We argue that all putative implicit stereotypes are affect-laden and all putative implicit prejudices are “semantic,” that is, they stand in co-activating associations with concepts and beliefs. Implicit biases, therefore, consist in “clusters” of semantic-affective associations, which differ in degree, rather than kind. This framework captures the psychological structure of implicit bias, promises to improve the power of indirect measures to predict behavior, and points toward the design of more effective interventions to combat discrimination. (shrink)

I investigate whether tobacco health warnings’ interference with autonomy is ethically justifiable in order to deter people from smoking. I dissociate first the informational role and the persuasive role of tobacco health warnings and show that both roles enable typical addicted smokers to better rule themselves, fostering their autonomy. The fact that some messages address people’s non-deliberative faculties is therefore compensated by a larger positive influence on their autonomy. However, misleading messages are not ethically justified and should be avoided. Tobacco (...) health warnings’ effect on autonomy highlights an important difference between libertarian paternalism and classical paternalism. (shrink)

The received view of implicit bias holds that it is associative and unreflective. Recently, the received view has been challenged. Some argue that implicit bias is not predicated on “any” associative process, but it is unreflective. These arguments rely, in part, on debiasing experiments. They proceed as follows. If implicit bias is associative and unreflective, then certain experimental manipulations cannot change implicitly biased behavior. However, these manipulations can change such behavior. So, implicit bias is not associative and unreflective. This paper (...) finds philosophical and empirical problems with that argument. When the problems are solved, the conclusion is not quite right: implicit bias is not necessarily unreflective, but it seems to be associative. Further, the paper shows that even if legitimate non-associative interventions on implicit bias exist, then both the received view and its recent contender would be false. In their stead would be interactionism or minimalism about implicit bias. (shrink)

Although it has been shown that immersive virtual reality can be used to induce illusions of ownership over a virtual body , information on whether this changes implicit interpersonal attitudes is meager. Here we demonstrate that embodiment of light-skinned participants in a dark-skinned VB significantly reduced implicit racial bias against dark-skinned people, in contrast to embodiment in light-skinned, purple-skinned or with no VB. 60 females participated in this between-groups experiment, with a VB substituting their own, with full-body visuomotor synchrony, reflected (...) also in a virtual mirror. A racial Implicit Association Test was administered at least three days prior to the experiment, and immediately after the IVR exposure. The change from pre- to post-experience IAT scores suggests that the dark-skinned embodied condition decreased implicit racial bias more than the other conditions. Thus, embodiment may change negative interpersonal attitudes and thus represent a powerful tool for exploring such fundamental psychological and societal phenomena. (shrink)

I argue that to have an attitude is, primarily, (1.) to have a dispositional profile that matches, to an appropriate degree and in appropriate respects, a stereotype for that attitude, typically grounded in folk psychology, and secondarily, (2.) in some cases also to meet further stereotypical attitude-specific conditions. To have an attitude, on the account I will recommend here, is mainly a matter of being apt to interact with the world in patterns that ordinary people would regard as characteristic of (...) having that attitude. I contrast this view with alternative views that treat having an attitude as a matter of possessing some particular internally stored representational content or some other "deep" structural feature. (shrink)