An increasing number of regulations propose the notion of ‘AI audits’ as an enforcement mechanism for achieving transparency and accountability for artificial intelligence (AI) systems. Despite some converging norms around various forms of AI auditing, auditing for the purpose of compliance and assurance currently have little to no agreed upon practices, procedures, taxonomies, and standards. We propose the ‘criterion audit’ as an operationalizable compliance and assurance external audit framework. We model elements of this approach after financial auditing practices, and (...) argue that AI audits should similarly provide assurance to their stakeholders about AI organizations’ ability to govern their algorithms in ways that mitigate harms and uphold human values. We discuss the necessary conditions for the criterion audit, and provide a procedural blueprint for performing an audit engagement in practice. We illustrate how this framework can be adapted to current regulations by deriving the criteria on which ‘bias audits’ for hiring algorithms can be performed, as required by the recently effective New York City Local Law 144 of 2021. We conclude by offering critical discussion on the benefits, inherent limitations, and implementation challenges of applying practices of the more mature financial auditing industry to AI auditing where robust guardrails against quality assurance issues are only starting to emerge. Our discussion as informed by experiences in performing these audits in practice highlights the critical role that an audit ecosystem plays in ensuring the effectiveness of such methodology. (shrink)

In this paper, we distinguish between different sorts of assessments of algorithmic systems, describe our process of assessing such systems for ethical risk, and share some key challenges and lessons for future algorithm assessments and audits. Given the distinctive nature and function of a third-party audit, and the uncertain and shifting regulatory landscape, we suggest that second-party assessments are currently the primary mechanisms for analyzing the social impacts of systems that incorporate artificial intelligence. We then discuss two kinds of (...) as-sessments: an ethical risk assessment and a narrower, technical algo-rithmic bias assessment. We explain how the two assessments depend on each other, highlight the importance of situating the algorithm within its particular socio-technical context, and discuss a number of lessons and challenges for algorithm assessments and, potentially, for algorithm audits. The discussion builds on our team’s experience of advising and conducting ethical risk assessments for clients across dif-ferent industries in the last four years. Our main goal is to reflect on the key factors that are potentially ethically relevant in the use of algo-rithms, and draw lessons for the nascent algorithm assessment and audit industry, in the hope of helping all parties minimize the risk of harm from their use. (shrink)

This paper offers an unorthodox appraisal of empirical research bearing on the question of the low representation of women in philosophy. It contends that fashionable views in the profession concerning implicit bias and stereotype threat are weakly supported, that philosophers often fail to report the empirical work responsibly, and that the standards for evidence are set very low—so long as you take a certain viewpoint.

One of the predominant debates in AI Ethics is the worry and necessity to create fair, transparent and accountable algorithms that do not perpetuate current social inequities. I offer a critical analysis of Reuben Binns’s argument in which he suggests using public reason to address the potential bias of the outcomes of machine learning algorithms. In contrast to him, I argue that ultimately what is needed is not public reason per se, but an audit of the implicit moral assumptions (...) of societies within which algorithms are built and applied. Public justification is appealing since it offers us the possibility to align the decision-making outcomes of the algorithm with the core moral values of stakeholders within a constitutional democratic society. My concern is that the common moral principles that form the foundation of public reason are not necessarily neutral, as they still express specific moral ideals and normative standards even if there is moral agreement by society as a whole, or among different stakeholders within society. Appealing to such normative standards may thus still lead to algorithmic outcomes being biased as common moral values may very well still be discriminatory even though they are formed from a consensus, and even if public reason is applied as a kind of filter for potential algorithmic outcomes. Hence, I argue that these implicit moral norms within society that we take as a given in public reasoning, need to be audited from generation to generation in order to effectively mitigate potential algorithmic bias. (shrink)

For more than two decades, police in the United States have used facial recognition to surveil civilians. Local police departments deploy facial recognition technology to identify protestors’ faces while federal law enforcement agencies quietly amass driver’s license and social media photos to build databases containing billions of faces. Yet, despite the widespread use of facial recognition in law enforcement, there are neither federal laws governing the deployment of this technology nor regulations settings standards with respect to its development. To make (...) matters worse, the Fourth Amendment—intended to limit police power and enacted to protect against unreasonable searches—has struggled to rein in new surveillance technologies since its inception. -/- This Article examines the Supreme Court’s Fourth Amendment jurisprudence leading up to Carpenter v. United States and suggests that the Court is reinterpreting the amendment for the digital age. Still, the too-slow expansion of privacy protections raises challenging questions about racial bias, the legitimacy of police power, and ethical issues in artificial intelligence design. This Article proposes the development of an algorithmic auditing and accountability market that not only sets standards for AI development and limitations on governmental use of facial recognition but encourages collaboration between public interest technologists and regulators. Beyond the necessary changes to the technological and legal landscape, the current system of policing must also be reevaluated if hard-won civil liberties are to endure. (shrink)

Over the years, companies have adopted hiring algorithms because they promise wider job candidate pools, lower recruitment costs and less human bias. Despite these promises, they also bring perils. Using them can inflict unintentional harms on individual human rights. These include the five human rights to work, equality and nondiscrimination, privacy, free expression and free association. Despite the human rights harms of hiring algorithms, the AI ethics literature has predominantly focused on abstract ethical principles. This is problematic for two (...) reasons. First, AI principles have been criticized for being vague and not actionable. Second, the use of vague ethical principles to discuss algorithmic risks does not provide any accountability. This lack of accountability creates an algorithmic accountability gap. Closing this gap is crucial because, without accountability, the use of hiring algorithms can lead to discrimination and unequal access to employment opportunities. This paper makes two contributions to the AI ethics literature. First, it frames the ethical risks of hiring algorithms using international human rights law as a universal standard for determining algorithmic accountability. Second, it evaluates four types of algorithmic impact assessments in terms of how effectively they address the five human rights of job applicants implicated in hiring algorithms. It determines which of the assessments can help companies audit their hiring algorithms and close the algorithmic accountability gap. (shrink)

AI systems that demonstrate significant bias or lower than claimed accuracy, and resulting in individual and societal harms, continue to be reported. Such reports beg the question as to why such systems continue to be funded, developed and deployed despite the many published ethical AI principles. This paper focusses on the funding processes for AI research grants which we have identified as a gap in the current range of ethical AI solutions such as AI procurement guidelines, AI impact assessments (...) and AI audit frameworks. We highlight the responsibilities of funding bodies to ensure investment is channelled towards trustworthy and safe AI systems and provides case studies as to how other ethical funding principles are managed. We offer a first sight of two proposals for funding bodies to consider regarding procedures they can employ. The first proposal is for the inclusion of a Trustworthy AI Statement’ section in the grant application form and offers an example of the associated guidance. The second proposal outlines the wider management requirements of a funding body for the ethical review and monitoring of funded projects to ensure adherence to the proposed ethical strategies in the applicants Trustworthy AI Statement. The anticipated outcome for such proposals being employed would be to create a ‘stop and think’ section during the project planning and application procedure requiring applicants to implement the methods for the ethically aligned design of AI. In essence it asks funders to send the message “if you want the money, then build trustworthy AI!”. (shrink)

Unreproducibility stemming from a loss of data integrity can be prevented with hash functions, secure sketches, and Benford's Law when combined with the historical practice of a Pli Cacheté where scientific discoveries were archived with a 3rd party to later prove the date of discovery. Including the distinct systems of preregistation and data provenance tracking becomes the starting point for the creation of a complete ontology of scientific documentation. The ultimate goals in such a system--ideally mandated--would rule out several forms (...) of dishonesty, catch computational and database errors, catch honest mistakes, and allow for automated data audits of large collaborative open science projects. (shrink)

This article sheds light on managing and utilizing scientific and technological funds (Sci-Tech funds) in Vietnam.

What are the sensory individuals of audition? What are the entities our auditory system attributes properties to? We examine various proposals about the nature of the sensory individuals of audition, and show that while each can account for some aspects of auditory perception, each also faces certain difficulties. We then put forward a new conception of sensory individuals according to which auditory sensory individuals are composite individuals. A feature shared by all existing accounts of sounds and sources is that they (...) postulate sensory individuals that are non-composite. They identify the sensory individual of sound hearing or source hearing with one type of entity in the environment, be they sound waves, vibrations, or interactions. We question this assumption and argue that our perceptual systems represent two or more aspects of the environment as a single sensory individual. Finally, we show that taking auditory individuals to be composite sensory individuals allows for an account of audition that is less problematic than its existing alternatives. (shrink)

Abstract: This research paper presents a novel approach to predict audit risks using a neural network model. The dataset used for this study was obtained from Kaggle and comprises 774 samples with 18 features, including Sector_score, PARA_A, SCORE_A, PARA_B, SCORE_B, TOTAL, numbers, marks, Money_Value, District, Loss, Loss_SCORE, History, History_score, score, and Risk. The proposed neural network architecture consists of three layers, including one input layer, one hidden layer, and one output layer. The neural network model was trained and validated, achieving (...) an impressive accuracy of 100% and an average error of 0.000015, indicating its robust predictive capability. Moreover, we conducted feature importance analysis to identify the most influential features for predicting audit risk. The key features found to be critical for classifying fraudulent activities in audit risk prediction are Sector_score, PARA_A, SCORE_A, PARA_B, SCORE_B, TOTAL, numbers, marks, Money_Value, District, Loss, Loss_SCORE, History, and History_score. This research contributes to the field of audit risk prediction by demonstrating the effectiveness of a neural network-based approach and highlighting the importance of specific features in detecting fraudulent activities. The findings have significant implications for auditors and organizations seeking to enhance their audit risk assessment processes, ultimately leading to improved financial transparency and fraud detection. (shrink)

Are individuals morally responsible for their implicit biases? One reason to think not is that implicit biases are often advertised as unconscious, ‘introspectively inaccessible’ attitudes. However, recent empirical evidence consistently suggests that individuals are aware of their implicit biases, although often in partial and inarticulate ways. Here I explore the implications of this evidence of partial awareness for individuals’ moral responsibility. First, I argue that responsibility comes in degrees. Second, I argue that individuals’ partial awareness of their implicit biases makes (...) them (partially) morally responsible for them. I argue by analogy to a close relative of implicit bias: moods. (shrink)

Future-bias is the preference, all else being equal, for positive states of affairs to be located in the future not the past, and for negative states of affairs to be located in the past not the future. Three explanations for future-bias have been posited: the temporal metaphysics explanation, the practical irrelevance explanation, and the three mechanisms explanation. Understanding what explains future-bias is important not only for better understanding the phenomenon itself, but also because many philosophers think that (...) which explanation is the correct one has implications for our evaluation of the normative status of future-bias. In this paper we focus on the connection between future-bias and the temporal value asymmetry. The latter refers to the fact that people tend to assign more value to a state of affairs when it is located in the future, as opposed to being equidistant in the past. All three candidate explanations offered for future-bias proceed by explaining the temporal value asymmetry, and then supposing that future-bias is simply a manifestation of that asymmetry. Recently, however, it has been argued that the conditions under which people display the temporal value asymmetry and future-bias are different, and that this is reason to doubt that the latter is a manifestation of the former. In this paper we empirically test one response to this argument against the manifestation thesis—the simulation response—according to which once we control for the degree to which people can simulate the object of their preference, we find that people display future-bias and the temporal value asymmetry in the same conditions. We find no evidence in favour of the simulation response. Moreover, our results support some previous research that suggests that the temporal value asymmetry is less robust than previously thought, and this in itself casts doubt on the idea that future bias is just a manifestation of this asymmetry. (shrink)

Present-bias is the preference, all else being equal, for positive events to be located in the present rather than the non-present, and for negative events to be located in the non-present rather than the present. Very little attention has been given to present-bias in the contemporary literature on time biases. This may be because it is often assumed that present-bias is not a distinctive psychological kind; that what explains people’s being present-biased is just what explains them displaying (...) various other time-biases. According to this view, there is no need to investigate present-bias independently of investigating these other biases, since present-bias is really just a manifestation of these other biases. We call this the manifestation thesis. We take up the question of whether the manifestation thesis is true, and argue that it is not. Thus, by failing to investigate present-bias in its own right we are failing fully to understand the spectrum of ways in which people display time biases. In turn, we suggest, this may have implications for the ways we evaluate whether present-bias is rationally permissible. (shrink)

A series of recent developments points towards auditing as a promising mechanism to bridge the gap between principles and practice in AI ethics. Building on ongoing discussions concerning ethics-based auditing, we offer three contributions. First, we argue that ethics-based auditing can improve the quality of decision making, increase user satisfaction, unlock growth potential, enable law-making, and relieve human suffering. Second, we highlight current best practices to support the design and implementation of ethics-based auditing: To be feasible and effective, ethics-based auditing (...) should take the form of a continuous and constructive process, approach ethical alignment from a system perspective, and be aligned with public policies and incentives for ethically desirable behaviour. Third, we identify and discuss the constraints associated with ethics-based auditing. Only by understanding and accounting for these constraints can ethics-based auditing facilitate ethical alignment of AI, while enabling society to reap the full economic and social benefits of automation. (shrink)

For synergistic interactions between humans and artificial intelligence (AI) systems, AI outputs often need to be explainable to people. Explainable AI (XAI) systems are commonly tested in human user studies. However, whether XAI researchers consider potential cultural differences in human explanatory needs remains unexplored. We highlight psychological research that found significant differences in human explanations between many people from Western, commonly individualist countries and people from non-Western, often collectivist countries. We argue that XAI research currently overlooks these variations and that (...) many popular XAI designs implicitly and problematically assume that Western explanatory needs are shared cross-culturally. Additionally, we systematically reviewed over 200 XAI user studies and found that most studies did not consider relevant cultural variations, sampled only Western populations, but drew conclusions about human-XAI interactions more generally. We also analyzed over 30 literature reviews of XAI studies. Most reviews did not mention cultural differences in explanatory needs or flag overly broad cross-cultural extrapolations of XAI user study results. Combined, our analyses provide evidence of a cultural bias toward Western populations in XAI research, highlighting an important knowledge gap regarding how culturally diverse users may respond to widely used XAI systems that future work can and should address. (shrink)

This paper examines the role of prestige bias in shaping academic philosophy, with a focus on its demographics. I argue that prestige bias exacerbates the structural underrepresentation of minorities in philosophy. It works as a filter against (among others) philosophers of color, women philosophers, and philosophers of low socio-economic status. As a consequence of prestige bias our judgments of philosophical quality become distorted. I outline ways in which prestige bias in philosophy can be mitigated.

In recent years, the ethical impact of AI has been increasingly scrutinized, with public scandals emerging over biased outcomes, lack of transparency, and the misuse of data. This has led to a growing mistrust of AI and increased calls for mandated ethical audits of algorithms. Current proposals for ethical assessment of algorithms are either too high level to be put into practice without further guidance, or they focus on very specific and technical notions of fairness or transparency that do (...) not consider multiple stakeholders or the broader social context. In this article, we present an auditing framework to guide the ethical assessment of an algorithm. The audit instrument itself is comprised of three elements: a list of possible interests of stakeholders affected by the algorithm, an assessment of metrics that describe key ethically salient features of the algorithm, and a relevancy matrix that connects the assessed metrics to stakeholder interests. The proposed audit instrument yields an ethical evaluation of an algorithm that could be used by regulators and others interested in doing due diligence, while paying careful attention to the complex societal context within which the algorithm is deployed. (shrink)

Abstract: This research focuses on utilizing Artificial Neural Networks (ANNs) to predict Audit Risk accurately, a critical aspect of ensuring financial system integrity and preventing fraud. Our dataset, gathered from Kaggle, comprises 18 diverse features, including financial and historical parameters, offering a comprehensive view of audit-related factors. These features encompass 'Sector_score,' 'PARA_A,' 'SCORE_A,' 'PARA_B,' 'SCORE_B,' 'TOTAL,' 'numbers,' 'marks,' 'Money_Value,' 'District,' 'Loss,' 'Loss_SCORE,' 'History,' 'History_score,' 'score,' and 'Risk,' with a total of 774 samples. Our proposed neural network architecture, consisting of three (...) layers (1 input, 1 hidden, and 1 output), forms the core of this study. While it may seem simple, the power of ANNs lies not in complexity but in their ability to uncover intricate data patterns. The model underwent rigorous training and validation, resulting in remarkable outcomes—an accuracy of 100% and an average error rate of 0.000015. In addition to performance metrics, our research investigates feature importance, revealing the key contributors to Audit Risk prediction. Notably, 'Sector_score,' 'PARA_A,' 'SCORE_A,' 'PARA_B,' 'SCORE_B,' 'TOTAL,' 'numbers,' 'marks,' 'Money_Value,' 'District,' 'Loss,' 'Loss_SCORE,' 'History,' and 'History_score' emerged as the most influential factors, shedding light on the elements crucial for precise risk assessment. This study advances Audit Risk prediction models and highlights the potential of ANNs in strengthening financial system integrity and fraud prevention. The insights gained from our findings offer practical guidance for stakeholders in the audit and financial sectors. (shrink)

What is the mental representation that is responsible for implicit bias? What is this representation that mediates between the trigger and the biased behavior? My claim is that this representation is neither a propositional attitude nor a mere association. Rather, it is mental imagery: perceptual processing that is not directly triggered by sensory input. I argue that this view captures the advantages of the two standard accounts without inheriting their disadvantages. Further, this view also explains why manipulating mental imagery (...) is among the most efficient ways of counteracting implicit bias. (shrink)

chapter on perception and bias including implicit bias.

Future-bias is preferring some lesser future good to a greater past good because it is in the future, or preferring some greater past pain to some lesser future pain because it is in the past. Most of us think that this bias is rational. I argue that no agents have future-biased preferences that are rationally evaluable—that is, evaluable as rational or irrational. Given certain plausible assumptions about rational evaluability, either we must find a new conception of future-bias (...) that avoids the difficulties I raise, or we must conclude that future-biased preferences are not subject to rational evaluation. (shrink)

All else being equal, most of us typically prefer to have positive experiences in the future rather than the past and negative experiences in the past rather than the future. Recent empirical evidence tends not only to support the idea that people have these preferences, but further, that people tend to prefer more painful experiences in their past rather than fewer in their future (and mutatis mutandis for pleasant experiences). Are such preferences rationally permissible, or are they, as time-neutralists contend, (...) rationally impermissible? And what is it that grounds their having the normative status that they do have? We consider two sorts of arguments regarding the normative status of future-biased preferences. The first appeals to the supposed arbitrariness of these preferences, and the second appeals to their upshot. We evaluate these arguments in light of the recent empirical research on future-bias. (shrink)

If you care about securing knowledge, what is wrong with being biased? Often it is said that we are less accurate and reliable knowers due to implicit biases. Likewise, many people think that biases reflect inaccurate claims about groups, are based on limited experience, and are insensitive to evidence. Chapter 3 investigates objections such as these with the help of two popular metaphors: bias as fog and bias as shortcut. Guiding readers through these metaphors, I argue that they (...) clarify the range of knowledge-related objections to implicit bias. They also suggest that there will be no unifying problem with bias from the perspective of knowledge. That is, they tell us that implicit biases can be wrong in different ways for different reasons. Finally, and perhaps most importantly, the metaphors reveal a deep—though perhaps not intractable—disagreement among theorists about whether implicit biases can be good in some cases when it comes to knowledge. (shrink)

Philosophers have long noted, and empirical psychology has lately confirmed, that most people are “biased toward the future”: we prefer to have positive experiences in the future, and negative experiences in the past. At least two explanations have been offered for this bias: belief in temporal passage and the practical irrelevance of the past resulting from our inability to influence past events. We set out to test the latter explanation. In a large survey, we find that participants exhibit significantly (...) less future bias when asked to consider scenarios where they can affect their own past experiences. This supports the “practical irrelevance” explanation of future bias. It also suggests that future bias is not an inflexible preference hardwired by evolution, but results from a more general disposition to “accept the things we cannot change”. However, participants still exhibited substantial future bias in scenarios in which they could affect the past, leaving room for complementary explanations. Beyond the main finding, our results also indicate that future bias is stake-sensitive and that participants endorse the normative correctness of their future-biased preferences and choices. In combination, these results shed light on philosophical debates over the rationality of future bias, suggesting that it may be a rational response to empirical realities rather than a brute, arational disposition. (shrink)

Our focus here is on whether, when influenced by implicit biases, those behavioural dispositions should be understood as being a part of that person’s character: whether they are part of the agent that can be morally evaluated.[4] We frame this issue in terms of control. If a state, process, or behaviour is not something that the agent can, in the relevant sense, control, then it is not something that counts as part of her character. A number of theorists have argued (...) that individuals do not have control, in the relevant sense, over the operation of implicit bias. We will argue that this claim is mistaken. We articulate and develop a notion of control that individuals have with respect to implicit bias, and argue that this kind of control can ground character-based evaluation of such behavioural dispositions. (shrink)

It has been argued that implicit biases are operative in philosophy and lead to significant epistemic costs in the field. Philosophers working on this issue have focussed mainly on implicit gender and race biases. They have overlooked ideological bias, which targets political orientations. Psychologists have found ideological bias in their field and have argued that it has negative epistemic effects on scientific research. I relate this debate to the field of philosophy and argue that if, as some studies (...) suggest, the same bias also exists in philosophy then it will lead to hitherto unrecognised epistemic hazards in the field. Furthermore, the bias is epistemically different from the more familiar biases in respects that are important for epistemology, ethics, and metaphilosophy. (shrink)

What is the status of research on implicit bias? In light of meta‐analyses revealing ostensibly low average correlations between implicit measures and behavior, as well as various other psychometric concerns, criticism has become ubiquitous. We argue that while there are significant challenges and ample room for improvement, research on the causes, psychological properties, and behavioral effects of implicit bias continues to deserve a role in the sciences of the mind as well as in efforts to understand, and ultimately (...) combat, discrimination and inequality. (shrink)

Many scientists routinely generalize from study samples to larger populations. It is commonly assumed that this cognitive process of scientific induction is a voluntary inference in which researchers assess the generalizability of their data and then draw conclusions accordingly. We challenge this view and argue for a novel account. The account describes scientific induction as involving by default a generalization bias that operates automatically and frequently leads researchers to unintentionally generalize their findings without sufficient evidence. The result is unwarranted, (...) overgeneralized conclusions. We support this account of scientific induction by integrating a range of disparate findings from across the cognitive sciences that have until now not been connected to research on the nature of scientific induction. The view that scientific induction involves by default a generalization bias calls for a revision of the current thinking about scientific induction and highlights an overlooked cause of the replication crisis in the sciences. Commonly proposed interventions to tackle scientific overgeneralizations that may feed into this crisis need to be supplemented with cognitive debiasing strategies against generalization bias to most effectively improve science. (shrink)

The Value-free ideal of science (VFI) is a view that claims that scientists should not use non-epistemic values when they are justifying their hypotheses, and is widely considered to be obsolete in the philosophy of science. I will defend the ideal by demonstrating that acceptance of non-epistemic values, prohibited by VFI, necessitates legitimizing certain problematic scientific practices. Such practices, including biased methodological decisions or Questionable Research Practices (QRP), significantly contribute to the Replication Crisis. I will argue that the realizability of (...) VFI is not a necessary condition for its validity. Then, I will show how some of the prominent proposals of value-laden science legitimize problematic scientific practices, provide real-world examples, and generalize the argument. Finally, I will show how value-laden methodological decisions contribute to the Replicability Crisis and discuss two strategies for realizing VFI. (shrink)

In recent years, a disagreement has erupted between two camps of philosophers about the rationality of bias toward the near and bias toward the future. According to the traditional hybrid view, near bias is rationally impermissible, while future bias is either rationally permissible or obligatory. Time neutralists, meanwhile, argue that the hybrid view is untenable. They claim that those who reject near bias should reject both biases and embrace time neutrality. To date, experimental work has (...) focused on future-directed near bias. The primary aim of this paper is to shed light on the debate by investigating past-directed near bias. If people treat the past and future differently with respect to near bias, by being future-directed but not past-directed near biased, then this supports a particular version of the hybrid view: temporal metaphysic hybridism. If people treat the past and future the same with respect to near bias, then this supports a simple version of time neutralism, which explains both future bias and near bias in terms of the functioning of a single mechanism: the anticipatory/retrospectory mechanism. Our results undermine the claim that people are future-directed, but not past-directed, near biased, and hence do not support temporal metaphysic hybridism. They also fail to support simple time-neutralism; instead, they suggest that there are multiple mechanisms that differently shape future- and past-directed preferences. (shrink)

Written by a diverse range of scholars, this accessible introductory volume asks: What is implicit bias? How does implicit bias compromise our knowledge of others and social reality? How does implicit bias affect us, as individuals and participants in larger social and political institutions, and what can we do to combat biases? An interdisciplinary enterprise, the volume brings together the philosophical perspective of the humanities with the perspective of the social sciences to develop rich lines of inquiry. (...) It is written in a non-technical style, using relatable examples that help readers understand what implicit bias is, its significance, and the controversies surrounding it. Each chapter includes discussion questions and additional reading suggestions. A companion webpage contains teaching resources. The volume will be an invaluable resource for students—and researchers—seeking to understand criticisms surrounding implicit bias, as well as how one might answer them by adopting a more nuanced understanding of bias and its role in maintaining social injustice. (shrink)

Randomization is an integral part of well-designed statistical trials, and is also a required procedure in legal systems. Implementation of honest, unbiased, understandable, secure, traceable, auditable and collusion resistant randomization procedures is a mater of great legal, social and political importance. Given the juridical and social importance of randomization, it is important to develop procedures in full compliance with the following desiderata: (a) Statistical soundness and computational efficiency; (b) Procedural, cryptographical and computational security; (c) Complete auditability and traceability; (d) Any (...) attempt by participating parties or coalitions to spuriously influence the procedure should be either unsuccessful or be detected; (e) Open-source programming; (f) Multiple hardware platform and operating system implementation; (g) User friendliness and transparency; (h) Flexibility and adaptability for the needs and requirements of multiple application areas (like, for example, clinical trials, selection of jury or judges in legal proceedings, and draft lotteries). This paper presents a simple and easy to implement randomization protocol that assures, in a formal mathematical setting, full compliance to the aforementioned desiderata for randomization procedures. (shrink)

Having a confirmation bias sometimes leads us to hold inaccurate beliefs. So, the puzzle goes: why do we have it? According to the influential argumentative theory of reasoning, confirmation bias emerges because the primary function of reason is not to form accurate beliefs, but to convince others that we’re right. A crucial prediction of the theory, then, is that confirmation bias should be found only in the reasoning domain. In this article, we argue that there is evidence (...) that confirmation bias does exist outside the reasoning domain. This undermines the main evidential basis for the argumentative theory of reasoning. In presenting the relevant evidence, we explore why having such confirmation bias may not be maladaptive. (shrink)

Some recent work in philosophy of religion addresses what can be called the “axiological question,” i.e., regardless of whether God exists, would it be good or bad if God exists? Would the existence of God make the world a better or a worse place? Call the view that the existence of God would make the world a better place “Pro-Theism.” We argue that Pro-Theism is not implausible, and moreover, many Theists, at least, (often implicitly) think that it is true. That (...) is, many Theists think that various good outcomes would arise if Theism is true. We then discuss work in cognitive science concerning human cognitive bias, before discussing two noteworthy attempts to show that at least some religious beliefs arise because of cognitive bias: Hume’s, and Draper’s and Nichols’s. We then argue that, as a result of certain cognitive biases that result when good outcomes might be at stake, Pro-Theism causes many Theists to inflate the epistemic probability that God exists, and as a result, Theists should lower the probability they assign to God’s existence. Finally, based our arguments, we develop a novel objection to Pascal’s wager. (shrink)

Humans typically display hindsight bias. They are more confident that the evidence available beforehand made some outcome probable when they know the outcome occurred than when they don't. There is broad consensus that hindsight bias is irrational, but this consensus is wrong. Hindsight bias is generally rationally permissible and sometimes rationally required. The fact that a given outcome occurred provides both evidence about what the total evidence available ex ante was, and also evidence about what that evidence (...) supports. Even if you in fact evaluate the ex ante evidence correctly, you should not be certain of this. Then, learning the outcome provides evidence that if you erred, you are more likely to have erred low rather than high in estimating the degree to which the ex ante evidence supported the hypothesis that that outcome would occur. (shrink)

Addressing biases in natural-language processing (NLP) systems presents an underappreciated ethical dilemma, which we think underlies recent debates about bias in NLP models. In brief, even if we could eliminate bias from language models or their outputs, we would thereby often withhold descriptively or ethically useful information, despite avoiding perpetuating or amplifying bias. Yet if we do not debias, we can perpetuate or amplify bias, even if we retain relevant descriptively or ethically useful information. Understanding this (...) dilemma provides for a useful way of rethinking the ethics of algorithmic bias in NLP. (shrink)

The rationality of future bias figures crucially in various metaphysical and ethical arguments (Prior 1959; Parfit 1984; Fischer 2019). Recently, however, philosophers have raised several arguments to the effect that future bias is irrational (Dougherty 2011; Suhler and Callender 2012; Greene and Sullivan 2015). Particularly, Greene and Sullivan (2015) claim that future bias is irrational because future bias leads to two kinds of irrational planning behaviors in agents who also seek to avoid regret. In this paper, (...) I join others who have objected to Greene and Sullivan’s argument (Dorsey 2016; Tarsney 2017), but for significantly different reasons that put into question an unchallenged premise in the debate. These reasons are based on the relationship between the alleged irrational planning behaviors and certain features of regret that regret shares with future bias. First, regret is dynamic, involving preferences that change over time and in inconsistent ways. Second, regret comes in degrees, meaning that we can rank our potential regrets. Because regret has these features, I show that the future-biased agents in Greene and Sullivan’s cases do not need to act in irrational ways to rationally avoid regret. (shrink)

Future-biased agents care not only about what experiences they have, but also when they have them. Many believe that A-theories of time justify future bias. Although presentism is an A-theory of time, some argue that it nevertheless negates the justification for future bias. Here, I claim that the alleged discrepancy between presentism and future bias is a special case of the cross-time relations problem. To resolve the discrepancy, I propose an account of future bias as a (...) preference for certain tensed truths properly relativized to the present. (shrink)

It has widely been assumed, by philosophers, that our first-person preferences regarding pleasurable and painful experiences exhibit a bias toward the future (positive and negative hedonic future-bias), and that our preferences regarding non-hedonic events (both positive and negative) exhibit no such bias (non-hedonic time-neutrality). Further, it has been assumed that our third-person preferences are always time-neutral. Some have attempted to use these (presumed) differential patterns of future-bias—different across kinds of events and perspectives—to argue for the irrationality (...) of hedonic future-bias. This paper experimentally tests these descriptive hypotheses. While as predicted we found first-person hedonic future-bias, we did not find that participants were time-neutral in all other conditions. Hence, the presumed asymmetry of hedonic/non-hedonic and first/third-person preferences cannot be used to argue for the irrationality of future-bias, since no such asymmetries exist. Instead, we develop a more fine-grained approach, according to which three factors—positive/negative valence, first/third-person, and hedonic/non-hedonic—each independently influence, but do not determine, whether an event is treated in a future-biased or time-neutral way. We discuss the upshots of these results for the debate over the rationality of future-bias. (shrink)

Philosophical literature has focused on the concept of trust, but often considers distrust merely as an afterthought. Distrust however, because of its pervasive role in our everyday lives, can be quite damaging. Thus, understanding the rationality of distrust is crucial for understanding our testimonial practices. In this paper I analyze whether it is rational or irrational to distrust an informant on the basis of identity bias. My aim is to show that distrust is irrational when based on negative identity (...) bias. First, I adopt Annalisa Coliva’s account of social hinge epistemology where it is rationally required to assume certain basic presuppositions, such that people are generally reliable informants, to obtain propositional justification and participate in the testimonial practice. Secondly, I show how distrust based on negative identity bias can spread across other domains of interaction and jeopardize the testimonial practice as we fail to assume that people are generally reliable informants. Thirdly, I argue that considering bias as a defeater is beneficial to maintain the claim that bias-based distrust is irrational, as it prevents the acquisition of propositional justification. Finally, I show that distrust is rational when based on evidence and reason that the speaker is not reliable or sincere. (shrink)

Important decisions that impact humans lives, livelihoods, and the natural environment are increasingly being automated. Delegating tasks to so-called automated decision-making systems can improve efficiency and enable new solutions. However, these benefits are coupled with ethical challenges. For example, ADMS may produce discriminatory outcomes, violate individual privacy, and undermine human self-determination. New governance mechanisms are thus needed that help organisations design and deploy ADMS in ways that are ethical, while enabling society to reap the full economic and social benefits of (...) automation. In this article, we consider the feasibility and efficacy of ethics-based auditing as a governance mechanism that allows organisations to validate claims made about their ADMS. Building on previous work, we define EBA as a structured process whereby an entity’s present or past behaviour is assessed for consistency with relevant principles or norms. We then offer three contributions to the existing literature. First, we provide a theoretical explanation of how EBA can contribute to good governance by promoting procedural regularity and transparency. Second, we propose seven criteria for how to design and implement EBA procedures successfully. Third, we identify and discuss the conceptual, technical, social, economic, organisational, and institutional constraints associated with EBA. We conclude that EBA should be considered an integral component of multifaced approaches to managing the ethical risks posed by ADMS. (shrink)

Recent empirical research has substantiated the finding that very many of us harbour implicit biases: fast, automatic, and difficult to control processes that encode stereotypes and evaluative content, and influence how we think and behave. Since it is difficult to be aware of these processes - they have sometimes been referred to as operating 'unconsciously' - we may not know that we harbour them, nor be alert to their influence on our cognition and action. And since they are difficult to (...) control, considerable work is required to prevent their influence. We here focus on the implications of these findings for epistemology. We first look at ways in which implicit biases thwart our knowledge seeking practices (sections 1 & 2). Then we set out putative epistemic benefits of implicit bias, before considering ways in which epistemic practices might be improved (section 3). Finally, we consider the distinctive challenges that the findings about implicit bias pose to us as philosophers, in the context of feminist philosophy in particular (section 4). (shrink)

Clinical Legal Education programmes now take place in most law schools in England and Wales. However, legal education continues to be predominantly focused on the analysis and application of rules, doctrines, and theories to hypothetical scenarios or essay questions. This form of pedagogy either minimises or ignores the role of the client in terms of supplying lawyers with knowledge pertinent to their case. In other words, it overlooks the fact that the lawyer’s acquisition of knowledge is not confined to technical (...) rationality. This article seeks to achieve three broad aims. First, to contribute to the debate concerning the epistemology of reflective practice. Second, to develop a theoretical reflective cycle, informed by Kant’s transcendental idealism, which seeks to maximise knowledge acquisition in legal education, namely knowledge supplied by the client. Third, to address implicit bias using the proposed reflective cycle. An optimal pedagogy for using this cycle is Clinical Legal Education, namely live client clinics. (shrink)

The concept of bias is pervasive in both popular discourse and empirical theorizing within philosophy, cognitive science, and artificial intelligence. This widespread application threatens to render the concept too heterogeneous and unwieldy for systematic investigation. This article explores recent philosophical literature attempting to identify a single theoretical category—termed ‘bias’—that could be unified across different contexts. To achieve this aim, the article provides a comprehensive review of theories of bias that are significant in the fields of philosophy of (...) mind, cognitive science, machine learning, and epistemology. It focuses on key examples such as perceptual bias, implicit bias, explicit bias, and algorithmic bias, scrutinizing their similarities and differences. Although these explorations may not conclusively establish the existence of a natural theoretical kind, pursuing the possibility offers valuable insights into how bias is conceptualized and deployed across diverse domains, thus deepening our understanding of its complexities across a wide range of cognitive and computational processes. (shrink)

The underrepresentation of women, people of color, and especially women of color—and the corresponding overrepresentation of white men—is more pronounced in philosophy than in many of the sciences. I suggest that part of the explanation for this lies in the role played by the idealized rational self, a concept that is relatively influential in philosophy but rarely employed in the sciences. The idealized rational self models the mind as consistent, unified, rationally transcendent, and introspectively transparent. I hypothesize that acceptance of (...) the idealized rational self leads philosophers to underestimate the influence of implicit bias on their own judgments and prevents them from enacting the reforms necessary to minimize the effects of implicit bias on institutional decision-making procedures. I consider recent experiments in social psychology that suggest that an increased sense of one’s own objectivity leads to greater reliance on bias in hiring scenarios, and I hypothesize how these results might be applied to philosophers’ evaluative judgments. I discuss ways that the idealized rational self is susceptible to broader critiques of ideal theory, and I consider some of the ways that the picture functions as a tool of active ignorance and color-evasive racism. (shrink)

In the philosophy of science, it is a common proposal that values are illegitimate in science and should be counteracted whenever they drive inquiry to the confirmation of predetermined conclusions. Drawing on recent cognitive scientific research on human reasoning and confirmation bias, I argue that this view should be rejected. Advocates of it have overlooked that values that drive inquiry to the confirmation of predetermined conclusions can contribute to the reliability of scientific inquiry at the group level even when (...) they negatively affect an individual’s cognition. This casts doubt on the proposal that such values should always be illegitimate in science. It also suggests that advocates of that proposal assume a narrow, individualistic account of science that threatens to undermine their own project of ensuring reliable belief formation in science. (shrink)

In analyzing implicit bias, one key issue is to clarify its metaphysical nature. In this paper, I develop a novel account of implicit bias by highlighting a particular kind of belief-like state that is partly constituted by phenomenal experiences. I call these states ‘qualiefs’ for three reasons: qualiefs draw upon qualitative experiences of what an object seems like to attribute a property to this very object, they share some of the distinctive features of proper beliefs, and they also (...) share some characteristics of what Gendler calls ‘aliefs’. I proceed as follows: First, I develop a general theory of qualiefs. Second, I argue that implicit bias involves generic qualiefs that involve experiences that have been shaped by stereotypes. Elaborating on the particular content of a generic qualief, I explain why we are unaware of the bias even though it involves an experience. Third, I demonstrate that the qualief-model best explains the key features of implicit bias: it accounts for the biases´ implicitness and automaticity. Moreover, it elucidates how implicit bias can be insensitive to logical form and evidence, but at the same time it can serve as propositional input to further mental states. (shrink)

Nearly everyone prefers pain to be in the past rather than the future. This seems like a rationally permissible preference. But I argue that appearances are misleading, and that future-biased preferences are in fact irrational. My argument appeals to trade-offs between hedonic experiences and other goods. I argue that we are rationally required to adopt an exchange rate between a hedonic experience and another type of good that stays fixed, regardless of whether the hedonic experience is in the past or (...) future. (shrink)

There is almost a consensus among conditional experts that indicative conditionals are not material. Their thought hinges on the idea that if indicative conditionals were material, A → B could be vacuously true when A is false, even if B would be false in a context where A is true. But since this consequence is implausible, the material account is usually regarded as false. It is argued that this point of view is motivated by the grammatical form of conditional sentences (...) and the symbols used to represent their logical form, which misleadingly suggest a one-way inferential direction from A to B. That conditional sentences mislead us into a directionality bias is a phenomenon that is well-documented in the literature about conditional reasoning. It is argued that this directional appearance is deceptive and does not reflect the underlying truth conditions of conditional sentences. This directional bias is responsible for both the unpopularity of the material account of conditionals and some of the main alternative principles and themes in conditional theory, including the Ramsey’s test, the Equation, Adams’ thesis, conditional-assertion and possible world theories. The directional mindset forgets a hard- earned lesson that made classical logic possible in the first place, namely, that grammatical form of sentences can mislead us about its truth conditions. There is a case to be made for a material account of indicative conditionals when we break the domination of words over the human mind. (shrink)