AI systems play an increasingly important role in shaping and regulating the lives of millions of human beings across the world. Calls for greater _transparency_ from such systems have been widespread. However, there is considerable ambiguity concerning what “transparency” actually means, and therefore, what greater transparency might entail. While, according to some debates, transparency requires _seeing through_ the artefact or device, widespread calls for transparency imply _seeing into_ different aspects of AI systems. These two notions are in apparent tension with (...) each other, and they are present in two lively but largely disconnected debates. In this paper, we aim to further analyse what these calls for transparency entail, and in so doing, clarify the sorts of transparency that we should want from AI systems. We do so by offering a taxonomy that classifies different notions of transparency. After a careful exploration of the different varieties of transparency, we show how this taxonomy can help us to navigate various domains of human–technology interactions, and more usefully discuss the relationship between technological transparency and human agency. We conclude by arguing that all of these different notions of transparency should be taken into account when designing more ethically adequate AI systems. (shrink)

Value capture occurs when an agent’s values are rich and subtle; they enter a social environment that presents simplified — typically quantified — versions of those values; and those simplified articulations come to dominate their practical reasoning. Examples include becoming motivated by FitBit’s step counts, Twitter Likes and Re-tweets, citation rates, ranked lists of best schools, and Grade Point Averages. We are vulnerable to value capture because of the competitive advantage that such crisp and clear expressions of value have in (...) our private reasoning and our public justification. There is, however, a price. In value capture, we take a central component of our autonomy — our ongoing deliberation over the exact articulation of our values — and we outsource it. And the metrics to which we outsource usually engineered for the interests of some external force, like a large-scale institution’s interest in cross-contextual comprehensibility and quick aggregability. That outsourcing cuts off one of the key benefits to personal deliberation. In value capture, we no longer adjust our values and their articulations in light of own rich experience of the world. Our values should often be carefully tailored to our particular selves or our small-scale communities, but in value capture, we buy our values off the rack. In some cases – like decreasing CO2 emissions – the costs of non-tailored values are outweighed by the benefit of precise collective coordination. In other cases, like in our aesthetic lives, they are not. This suggests that we should want different values suited to different scales. We should want value federalism. Some values are perhaps best pursued at the largest-scale level, others at smaller scales. The problem occurs when we exhibit an excess preference for the largest-scale values – when we consistently let the universal metrics swamp our quieter interests. (shrink)

When thinking about designing social media platforms, we often focus on factors such as usability, functionality, aesthetics, ethics, and so forth. Epistemic considerations have rarely been given the same level of attention in design discussions. This paper aims to rectify this neglect. We begin by arguing that there are epistemic norms that govern environments, including social media environments. Next, we provide a framework for applying these norms to the question of platform design. We then apply this framework to the real-world (...) case of long-form informational content platforms. We argue that many current long-form informational content platforms are epistemically unhealthy. The good news? We provide concrete advice on how to take steps toward improving their health! Specifically, we argue that they should change how they verify and authenticate content creators and how this information is displayed to content consumers. We conclude by connecting this guidance to broader issues about the epistemic health of platforms. (shrink)

A crucial tenet of virtue-based or expertise-based theorizing about politics is that there are ways to identify and select morally and epistemically excellent people to hold office. This paper considers historical challenges to this task that come from within Greek and Confucian thought and political practice. Because of how difficult it is to assess character in ordinary settings, we argue that it is even more difficult to design institutions that select for virtue at the much wider political scale. Specifically, we (...) argue that the vast majority of proposals that purport to select for virtue are either (a) unlikely to be effective, (b) not implementable at the scale of most current nations, or (c) have already been incorporated into most democratic states. Thus defenders of meritocratic institutions should take more seriously the practical barriers that will almost certainly arise when trying to implement these proposals at scale. Historical proponents of meritocratic values were well aware of concerns about the feasibility of their political ideals, and we should be too. (shrink)

Advancements in machine learning have fuelled the popularity of using AI decision algorithms in procedures such as bail hearings, medical diagnoses and recruitment. Academic articles, policy texts, and popularizing books alike warn that such algorithms tend to be opaque: they do not provide explanations for their outcomes. Building on a causal account of transparency and opacity as well as recent work on the value of causal explanation, I formulate a moral concern for opaque algorithms that is yet to receive a (...) systematic treatment in the literature: when such algorithms are used in life-changing decisions, they can obstruct us from effectively shaping our lives according to our goals and preferences, thus undermining our autonomy. I argue that this concern deserves closer attention as it furnishes the call for transparency in algorithmic decision-making with both new tools and new challenges. (shrink)

Previously proposed strategies for tackling hermeneutical injustices take for granted the interests people have in certain things about them being intelligible to them and/or to others, and seek to enable them to satisfy these interests. Strategies of this sort I call interests-as-given strategies. I propose that some hermeneutical injustices can instead be tackled by doing away with certain of these interests, and so with the possibility of their unfair non-satisfaction. Strategies of this sort I call interests-in-question strategies. As a case (...) study in when such an interests-in-question strategy ought to be pursued, I look at how to tackle hermeneutical injustices arising in the context of gender-affirming healthcare as provided to adults by the National Health Service in the UK. I argue that considerations of trust, privacy, and respect all support pursuing such a strategy. One way to do so, I suggest, would be by replacing the existing gatekeeping model with an informed consent model for the provision of gender-affirming healthcare. Considerations of hermeneutical justice can hence be added to the already-impressive case for undertaking this shift. (shrink)

In this article I argue that often propositional knowledge is acquired and retained by extensive reliance on physical and social scaffolds that create an environment or niche conducive to knowledge. It is incumbent on epistemologists to subject these aids to epistemic assessments. I show that several of the activities involved in the creation of niches within which inquiry can thrive are carried out by whole cultures. New generations benefit from inheriting these niches whilst being able to improve upon them to (...) the advantage of their descendants. Finally, I highlight that the growth of human epistemic achievements is often due to increased outsourcing of cognitive effort and epistemic powers onto impersonal physical and social structures so that human beings can succeed more by contributing less to the solution of problems. (shrink)

This paper explores the socio-epistemic practice of shopping for experts. I argue that expert shopping is particularly likely to occur on what Thi Nguyen calls cognitive islands. To support my argument, I focus on macroeconomics. First, I make a prima-facie case for thinking that macroeconomics is a cognitive island. Then, I argue that ordinary people are particularly likely to engage in expert shopping when it comes to macroeconomic matters. In particular, I distinguish between two kinds of expert shopping, which I (...) call cynical and wishful, and introduce the notion of assisted expert shopping, which occurs when people or organizations shop for experts on behalf of other people. I argue that assisted expert shopping can sometime result in what I call a propagandistic use of expertise. Finally, I critically examine some possible reasons for optimism and find them wanting. I conclude by suggesting that that much of what I said about shopping for macroeconomic experts might also apply mutatis mutandis to other policy-relevant domains of expertise. (shrink)

In the article I argue that intellectual arrogance can be an individual, collective and even corporate vice. I show that arrogance is in all these cases underpinned by defensive positive evaluations of epistemic features of the evaluator in the service of buttressing its illegitimate social dominance. Individual arrogance as superbia or as hubris stems from attitudes biased by the motive of self-enhancement. Collective arrogance is underpinned by positive defensive attitudes to a one’s social identity that seeks to maintain its unwarranted (...) social dominance. Finally, corporations are arrogant when their attitudes are the aggregation of the arrogant dispositions of its managers or when these corporations have inherited structures and policies that are defensive of its illegitimately dominant social status. (shrink)

To place epistemic trust in someone is to take their word for something. Much of the existing literature on epistemic trust concerns epistemic authorities. But as important as authority is to epistemic trust, it pales in comparison to the epistemic importance of conflicts of interests. In economics, we say that buyers shouldn’t take the word of sellers. Caveat emptor: let the buyer beware. I argue for a similar principle in epistemology. Caveat auditor: let the hearer beware. Others often have incentives (...) to testify in ways that are odds with our epistemic goals. Given this, our epistemic trust in others should be calibrated to reflect the epistemically virtuous and perverse incentives of ourselves and others. This basic principle explains the need for epistemic caution in a wide range of applied topics from politics to product reviews to fake news. (shrink)

Patient data is used to drive an ecosystem of advanced digital tools in health care, like predictive models or artificial intelligence‐based decision support. Patients themselves, however, receive little information about these technologies or how they affect their care. This raises important questions about patient trust and continued engagement in a health care system that extracts their data but does not treat them as key stakeholders. This essay explores these tensions and provides steps forward for health systems as they design advanced (...) health information‐technology (IT) policies and practices. It centers patients, their concerns, and the ways they perceive trustworthiness to reframe advanced health IT in service of patient interests. (shrink)

Katherine Hawley was skeptical about group trust. Her main reason for this skepticism was that the distinction between trust and reliance, central to many theories of interpersonal trust, does not apply to trust in groups. Hawley’s skeptical arguments successfully shift the burden of proof to those who wish to continue with a concept of group trust. Nonetheless, I argue that a commitments account of the trust/reliance distinction can shoulder that burden. According to that commitments account, trust is a distinctive kind (...) of reliance grounded in a positive appraisal of features of the trustee’s practical rationality, foremost their commitments and their capacity to act on those commitments. This is one way we can make sense of the difference between trusting and relying on individual people. I argue it is also a way we can make sense of the difference between trusting and relying on groups. (shrink)

1. The current political scene in many countries is populated by polarised groups with sharply contrasting loyalties and beliefs implying that there are fundamental schisms between opposing groups....

The purpose of this paper is to advance understanding of intellectually virtuous honesty, by examining the relationship between a recent account of intellectual honesty and a recent account of intellectual transparency. The account of intellectual honesty comes from Nathan King, who adapts the work of Christian Miller on moral honesty, while the account of intellectual transparency comes from T. Ryan Byerly. After introducing the respective accounts, I identify four potential differences between intellectual honesty and intellectual transparency as understood by these (...) accounts. I then turn to the question of how to think about the relationship between these traits in light of these potential differences. I make the case that intellectual transparency can either be regarded as an exceptionally strong or ideal variety of intellectual honesty, or it can be regarded as a distinct virtue from intellectual honesty which is a more cardinal virtue than the latter. Along the way, I also note some places where a case can be made that Miller's and King's accounts of honesty and intellectual honesty are in need of refinement or clarification. (shrink)

It is widely held that transparency incentivizes good behavior. Though that may be, sometimes, there are tradeoffs here: transparency incentivizes people to conceal genuine reasons for action and instead manufacture insincere reasons for public consumption. The evidence for this comes from moral psychology and economics: when people are observed, they acquire an incentive to make more deontological and intuitive moral judgments than they would otherwise. In contrast, transparency incentivizes politicians and leaders to make more consequentialist and calculated moral judgments than (...) they would otherwise. Transparency incentivizes people to foster distinct (inauthentic) moral identities—one personal and one public—or to make moral judgments based on reputational reasons that sometimes diverge from the moral facts. In either case, transparency can be morally and politically corrupting. (shrink)

This paper investigates the value of trust and the proper attitude lay people ought to have towards scientific experts. Trust in expertise is usually considered to be valuable, while distrust is often analyzed in cases where it is harmful. I will draw on accounts from political philosophy and argue that it is not only public trust that is valuable when it comes to scientific expertise – but also public vigilance. Expertise may be distorted in different ways, which cannot be remedied (...) by internal control mechanisms alone. This reveals the importance of some forms of democratic oversight. The proper attitude is vigilant trust in expertise. However, vigilant trust seems to be a contradictory notion: How can one be trusting and watchful at the same time? I will show that it is not, and that trust and vigilance can be compatible to a certain extent. I will do so by distinguishing between different levels of both trust and vigilance. Interestingly, this argument requires acknowledging the value of some forms of distrust in scientific expertise, even if that distrust targets trustworthy experts. (shrink)

Open science is good for both epistemic and social reasons, but in nonobvious ways, it can have detrimental epistemic side effects. Drawing on case studies and the social epistemology of science, I show how practices intended to increase transparency, communication, and information sharing in science can backfire. We should not reject Open Science, just implement it carefully. I argue that we can do so by treating openness as a governing value in science, and thus, that our pursuit of openness needs (...) to be balanced against our pursuit of the whole scheme of values that govern science. (shrink)

With focus on the development and use of artificial intelligence (AI) systems in the digital health context, we consider the following questions: How does the European Union (EU) seek to facilitate the development and uptake of trustworthy AI systems through the AI Act? What does trustworthiness and trust mean in the AI Act, and how are they linked to some of the ongoing discussions of these terms in bioethics, law, and philosophy? What are the normative components of trustworthiness? And how (...) do the requirements of the AI Act relate to these components? We first explain how the EU seeks to create an epistemic environment of trust through the AI Act to facilitate the development and uptake of trustworthy AI systems. The legislation establishes a governance regime that operates as a socio-epistemological infrastructure of trust which enables a performative framing of trust and trustworthiness. The degree of success that performative acts of trust and trustworthiness have achieved in realising the legislative goals may then be assessed in terms of statutorily defined proxies of trustworthiness. We show that to be trustworthy, these performative acts should be consistent with the ethical principles endorsed by the legislation; these principles are also manifested in at least four key features of the governance regime. However, specified proxies of trustworthiness are not expected to be adequate for applications of AI systems within a regulatory sandbox or in real-world testing. We explain why different proxies of trustworthiness for these applications may be regarded as ‘special’ trust domains and why the nature of trust should be understood as participatory. (shrink)

Digital Democratic Innovations (DDIs) have largely been conceived of, by the academic community, as a possible solution to the crisis of representative democracy. DDIs can be defined as initiatives or institutions designed with the goal of deepening citizens’ participation and influence on political decisions through the use of digital tools and platforms. There is a hope that DDIs (as well as usual, non-digital DIs) could help nurture political trust in governing institutions. Yet the vast majority of research on trust and (...) DDIs/DIs focuses only on one side of political trust – the trust of citizens vis-à-vis government. What has largely been ignored, however, is the trust of decision-makers towards citizen participation and its outcomes. We contend that political trust should be seen as a two-sided relationship that implies reciprocity. Therefore, we claim that it is necessary to design DDIs in such a way that they generate political trust from both common citizens and decision-makers. Citizens must trust institutions and processes implemented by the ones in power, and decision-makers must trust the input of citizens’ participatory processes. We argue that transparency is crucial when it comes to fostering reciprocal trust, and we identify the necessary elements to achieve it via DDIs. (shrink)

Philosophical reflection on transparency, including group transparency, is beginning to gain steam. This paper contributes to this work by developing a conceptualization of transparency as an intellectual character trait that groups can possess, and by presenting a novel argument for thinking that such transparency should be understood along non-summativist lines. According to the account offered, a group’s being intellectually transparent consists in the group’s tending to attend well to its perspective and to share its perspective faithfully with others in order (...) to promote their epistemic goods. It is argued that this kind of group intellectual transparency, pace summativism, does not always consist merely in group members possessing intellectual transparency. The argument given for this conclusion works differently from existing arguments for non-summativism about group character traits, and it retains persuasive power even if summativist views of most group phenomena, including other group character traits, are correct. (shrink)