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)

Following the publication of numerous ethical principles and guidelines, the concept of 'Trustworthy AI' has become widely used. However, several AI ethicists argue against using this concept, often backing their arguments with decades of conceptual analyses made by scholars who studied the concept of trust. In this paper, I describe the historical-philosophical roots of their objection and the premise that trust entails a human quality that technologies lack. Then, I review existing criticisms about 'Trustworthy AI' and the consequence (...) of ignoring these criticisms: if the concept of 'Trustworthy AI' is kept being used, we risk attributing responsibilities to agents who cannot be held responsible, and consequently, deteriorate social structures which regard accountability and liability. Nevertheless, despite suggestions to shift the paradigm from 'Trustworthy AI' to 'Reliable AI', I argue that, realistically, this concept will be kept being used. I end by arguing that, ultimately, AI ethics is also about power, social justice, and scholarly activism. Therefore, I propose that community-driven and social justice-oriented ethicists of AI and trust scholars further focus on (a) democratic aspects of trust formation; and (b) draw attention to critical social aspects highlighted by phenomena of distrust. This way, it will be possible to further reveal shifts in power relations, challenge unfair status quos, and suggest meaningful ways to keep the interests of citizens. (shrink)

Many policies and ethical guidelines recommend developing “trustworthy AI”. We argue that developing morally trustworthy AI is not only unethical, as it promotes trust in an entity that cannot be trustworthy, but it is also unnecessary for optimal calibration. Instead, we show that reliability, exclusive of moral trust, entails the appropriate normative constraints that enable optimal calibration and mitigate the vulnerability that arises in high-stakes hybrid decision-making environments, without also demanding, as moral trust would, the anthropomorphization of (...) AI and thus epistemically dubious behavior. The normative demands of reliability for inter-agential action are argued to be met by an analogue to procedural metacognitive competence (i.e., the ability to evaluate the quality of one’s own informational states to regulate subsequent action). Drawing on recent empirical findings that suggest providing reliability scores (e.g., F1-scores) to human decision-makers improves calibration in the AI system, we argue that reliability scores provide a good index of competence and enable humans to determine how much they wish to rely on the system. (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)

AI is revolutionizing everyone’s life, and it is crucial that it does so in the right way. AI’s profound and far-reaching potential for transformation concerns the engineering of systems that have some degree of autonomous agency. This is epochal and requires establishing a new, ethical balance between human and artificial autonomy.

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)

Simion and Kelp explain the trustworthiness of an AI based on that AI’s disposition to meet its obligations. Roughly speaking, according to Simion and Kelp, an AI is trustworthy regarding its task if and only if that AI is obliged to complete the task and its disposition to complete the task is strong enough. Furthermore, an AI is obliged to complete a task in the case where the task is the AI’s etiological function or design function. This account has (...) a strength in that it can provide a unificatory rationale for the trustworthy-making properties of AIs. According to this account, being explainable, being safe, and being transparent are the trustworthy-making properties of an AI because an AI can fulfill its etiological or design functions in the case where it is explainable, safe, and transparent. This paper shows that though Simion and Kelp’s account has a strength, this account is not satisfactory for two reasons. The first reason is that an AI’s trustworthiness is not determined just by the AI’s disposition to meet obligations, and the second reason is that it is difficult to explain how an AI’s etiological function and design function have to do with that AI’s obligations. To provide a full-fledged account, Simion and Kelp should dismiss these concerns. (shrink)

The use of AI algorithms in criminal trials has been the subject of very lively ethical and legal debates recently. While there are concerns over the lack of accuracy and the harmful biases that certain algorithms display, new algorithms seem more promising and might lead to more accurate legal decisions. Algorithms seem especially relevant for bail decisions, because such decisions involve statistical data to which human reasoners struggle to give adequate weight. While getting the right legal outcome is a strong (...) desideratum of criminal trials, advocates of the relational theory of procedural justice give us good reason to think that fairness and perceived fairness of legal procedures have a value that is independent from the outcome. According to this literature, one key aspect of fairness is trustworthiness. In this paper, I argue that using certain algorithms to assist bail decisions could increase three diferent aspects of judges’ trustworthiness: (1) actual trustworthiness, (2) rich trustworthiness, and (3) perceived trustworthiness. (shrink)

Recent AI ethics has focused on applying abstract principles downward to practice. This paper moves in the other direction. Ethical insights are generated from the lived experiences of AI-designers working on tangible human problems, and then cycled upward to influence theoretical debates surrounding these questions: 1) Should AI as trustworthy be sought through explainability, or accurate performance? 2) Should AI be considered trustworthy at all, or is reliability a preferable aim? 3) Should AI ethics be oriented toward establishing (...) protections for users, or toward catalyzing innovation? Specific answers are less significant than the larger demonstration that AI ethics is currently unbalanced toward theoretical principles, and will benefit from increased exposure to grounded practices and dilemmas. (shrink)

Many artificial intelligence (AI) systems currently used for decision-making are opaque, i.e., the internal factors that determine their decisions are not fully known to people due to the systems’ computational complexity. In response to this problem, several researchers have argued that human decision-making is equally opaque and since simplifying, reason-giving explanations (rather than exhaustive causal accounts) of a decision are typically viewed as sufficient in the human case, the same should hold for algorithmic decision-making. Here, I contend that this argument (...) overlooks that human decision-making is sometimes significantly more transparent and trustworthy than algorithmic decision-making. This is because when people explain their decisions by giving reasons for them, this frequently prompts those giving the reasons to govern or regulate themselves so as to think and act in ways that confirm their reason reports. AI explanation systems lack this self-regulative feature. Overlooking it when comparing algorithmic and human decision-making can result in underestimations of the transparency of human decision-making and in the development of explainable AI that may mislead people by activating generally warranted beliefs about the regulative dimension of reason-giving. (shrink)

I discuss an influential argument put forward by Hatherley in theJournal of Medical Ethics. Drawing on influential philosophical accounts of interpersonal trust, Hatherley claims that medical artificial intelligence is capable of being reliable, but not trustworthy. Furthermore, Hatherley argues that trust generates moral obligations on behalf of the trustee. For instance, when a patient trusts a clinician, it generates certain moral obligations on behalf of the clinician for her to do what she is entrusted to do. I make three (...) objections to Hatherley’s claims: (1) At least one philosophical account of interagent trust implies that medical AI is capable of being trustworthy. (2) Even if this account should ultimately be rejected, it does not matter much because what we care mostly about is that medical AI is reliable. (3) It is false that trust in itself generates moral obligations on behalf of the trustee. (shrink)

Australia is a leading AI nation with strong allies and partnerships. Australia has prioritised the development of robotics, AI, and autonomous systems to develop sovereign capability for the military. Australia commits to Article 36 reviews of all new means and method of warfare to ensure weapons and weapons systems are operated within acceptable systems of control. Additionally, Australia has undergone significant reviews of the risks of AI to human rights and within intelligence organisations and has committed to producing ethics guidelines (...) and frameworks in Security and Defence. Australia is committed to OECD’s values-based principles for the responsible stewardship of trustworthy AI as well as adopting a set of National AI ethics principles. While Australia has not adopted an AI governance framework specifically for Defence; Defence Science has published ‘A Method for Ethical AI in Defence’ (MEAID) technical report which includes a framework and pragmatic tools for managing ethical and legal risks for military applications of AI. (shrink)

Artificial intelligence (AI) is expected to revolutionise the practice of medicine. Recent advancements in the field of deep learning have demonstrated success in variety of clinical tasks: detecting diabetic retinopathy from images, predicting hospital readmissions, aiding in the discovery of new drugs, etc. AI’s progress in medicine, however, has led to concerns regarding the potential effects of this technology on relationships of trust in clinical practice. In this paper, I will argue that there is merit to these concerns, since AI (...) systems can be relied on, and are capable of reliability, but cannot be trusted, and are not capable of trustworthiness. Insofar as patients are required to rely on AI systems for their medical decision-making, there is potential for this to produce a deficit of trust in relationships in clinical practice. (shrink)

This paper analyses the phenomenology and epistemology of chatbots such as ChatGPT and Bard. The computational architecture underpinning these chatbots are large language models (LLMs), which are generative artificial intelligence (AI) systems trained on a massive dataset of text extracted from the Web. We conceptualise these LLMs as multifunctional computational cognitive artifacts, used for various cognitive tasks such as translating, summarizing, answering questions, information-seeking, and much more. Phenomenologically, LLMs can be experienced as a “quasi-other”; when that happens, users anthropomorphise them. (...) For most users, current LLMs are black boxes, i.e., for the most part, they lack data transparency and algorithmic transparency. They can, however, be phenomenologically and informationally transparent, in which case there is an interactional flow. Anthropomorphising and interactional flow can, in some users, create an attitude of (unwarranted) trust towards the output LLMs generate. We conclude this paper by drawing on the epistemology of trust and testimony to examine the epistemic implications of these dimensions. Whilst LLMs generally generate accurate responses, we observe two epistemic pitfalls. Ideally, users should be able to match the level of trust that they place in LLMs to the degree that LLMs are trustworthy. However, both their data and algorithmic opacity and their phenomenological and informational transparency can make it difficult for users to calibrate their trust correctly. The effects of these limitations are twofold: users may adopt unwarranted attitudes of trust towards the outputs of LLMs (which is particularly problematic when LLMs hallucinate), and the trustworthiness of LLMs may be undermined. (shrink)

While artificial intelligence (AI) is increasingly applied for decision-making processes, ethical decisions pose challenges for AI applications. Given that humans cannot always agree on the right thing to do, how would ethical decision-making by AI systems be perceived and how would responsibility be ascribed in human-AI collaboration? In this study, we investigate how the expert type (human vs. AI) and level of expert autonomy (adviser vs. decider) influence trust, perceived responsibility, and reliance. We find that participants consider humans to be (...) more morally trustworthy but less capable than their AI equivalent. This shows in participants’ reliance on AI: AI recommendations and decisions are accepted more often than the human expert’s. However, AI team experts are perceived to be less responsible than humans, while programmers and sellers of AI systems are deemed partially responsible instead. (shrink)

As systems based on opaque Artificial Intelligence (AI) continue to flourish in diverse real-world applications, understanding these black box models has become paramount. In response, Explainable AI (XAI) has emerged as a field of research with practical and ethical benefits across various domains. This paper not only highlights the advancements in XAI and its application in real-world scenarios but also addresses the ongoing challenges within XAI, emphasizing the need for broader perspectives and collaborative efforts. We bring together experts from diverse (...) fields to identify open problems, striving to synchronize research agendas and accelerate XAI in practical applications. By fostering collaborative discussion and interdisciplinary cooperation, we aim to propel XAI forward, contributing to its continued success. Our goal is to put forward a comprehensive proposal for advancing XAI. To achieve this goal, we present a manifesto of 27 open problems categorized into nine categories. These challenges encapsulate the complexities and nuances of XAI and offer a road map for future research. For each problem, we provide promising research directions in the hope of harnessing the collective intelligence of interested stakeholders. (shrink)

This research explores how a person with whom information has been shared and, importantly, an artificial intelligence (AI) system used to deduce information from the shared data contribute to making the disclosure context private. The study posits that private contexts are constituted by the interactions of individuals in the social context of intersubjectivity based on trust. Hence, to make the context private, the person who is the trustee (i.e., with whom information has been shared) must fulfil trust norms. According to (...) the commitment account of trustworthiness, a person is trustworthy only if they satisfy the norm of competence. It is argued that a person using an AI system to answer a question is competent only if they are ex post justified in believing what has been delivered by the AI system. A person’s belief is justified in the doxastic sense only if the AI system is accurate. This feature of AI’s performance affects a person’s competence and, as a result, trustworthiness. The effect of AI on trust as an essential component of making the context private, and thus on privacy, means an AI system also impacts privacy. Therefore, a private context is constituted when the individual with whom the information is shared fulfils the competence norm and the AI system used for analysing the information is sufficiently accurate to adhere to this norm. The result of this research emphasises the significance of the relationship between individuals involved in information-sharing and how an AI system used for analysing that information impacts the relationship regarding making the context private, as well as how it impacts privacy. The findings of this research have significant implications for improving or ameliorating privacy regulations in light of trust. (shrink)

The field of AI has undergone a series of transformations, each marking a new phase of development. The initial phase emphasized curation of symbolic models which excelled in capturing reasoning but were fragile and not scalable. The next phase was characterized by machine learning models—most recently large language models (LLMs)—which were more robust and easier to scale but struggled with reasoning. Now, we are witnessing a return to symbolic models as complementing machine learning. Successes of LLMs contrast with their inscrutability, (...) inaccuracy, and hallucinations, which underwrite concerns over the reliability and trustworthiness of these systems, motivating investigations into commonsense reasoning, AI explainability, and formal verification techniques. Moreover, proper assessments of hybrid machine learning/symbolic systems require novel strategies to facilitate comparisons of performance and guide future AI progress. The EDGeS AAAI 2023 Spring symposium brought together researchers focusing on novel assessments and benchmarks for evaluating hybrid and artificial general intelligence (AGI) systems. This symposium revealed what was already suspected: research concerning evaluation of machine learning/symbolic hybrids and AGI is, unfortunately, lacking. Even so, the discussion was fruitful. (shrink)

It has been argued that our currently most satisfactory social epistemology of science can’t account for science that is based on artificial intelligence (AI) because this social epistemology requires trust between scientists that can take full responsibility for the research tools they use, and scientists can’t take full responsibility for the AI tools they use since these systems are epistemically opaque. I think this argument overlooks that much AI-based science can be done without opaque models, and that agents can take (...) full responsibility for the systems they use even if these systems are opaque. Requiring that an agent fully understand how a system works is an untenably strong condition for that agent to take full responsibility for the system and risks absolving AI developers from responsibility for their products. AI-based science need not create trust-related social epistemological problems if we keep in mind that what makes both individual scientists and their use of AI systems trustworthy isn’t full transparency of the internal processing but their adherence to social and institutional norms that ensure that scientific claims can be trusted. (shrink)

This paper addresses the cultural sustainability of artificial intelligence use through one of its most widely discussed instances: autonomous driving. The introduction of self-driving cars places us in a radically novel moral situation, requiring advance, reflectively endorsed, forced, and iterable choices, with yet uncharted forms of risk imposition. The argument is meant to explore the necessity and possibility of maintaining one of our most fundamental moral-cultural principles in this new context, that of the equal treatment of persons. It is claimed (...) that the implementation of this principle requires central and uniform regulation in autonomous mobility. (shrink)

Entscheidungen verweisen in einem begrifflichen Sinne auf Gründe. Entscheidungssysteme bieten eine probabilistische Verlässlichkeit als Rechtfertigung von Empfehlungen an. Doch nicht für alle Situationen mögen Verlässlichkeitsgründe auch angemessene Gründe sein. Damit eröffnet sich die Idee, die Güte von Gründen von ihrer Angemessenheit zu unterscheiden. Der Aufsatz betrachtet an einem Beispiel, einem KI-Lügendetektor, die Frage, ob eine (zumindest aktuell nicht gegebene) hohe Verlässlichkeit den Einsatz rechtfertigen kann. Gleicht er nicht einem Richter, der anhand einer Statistik Urteile fällen würde?

The Salivaomics Knowledge Base (SKB) is designed to serve as a computational infrastructure that can permit global exploration and utilization of data and information relevant to salivaomics. SKB is created by aligning (1) the saliva biomarker discovery and validation resources at UCLA with (2) the ontology resources developed by the OBO (Open Biomedical Ontologies) Foundry, including a new Saliva Ontology (SALO). We define the Saliva Ontology (SALO; http://www.skb.ucla.edu/SALO/) as a consensus-based controlled vocabulary of terms and relations dedicated to the salivaomics (...) domain and to saliva-related diagnostics following the principles of the OBO (Open Biomedical Ontologies) Foundry. The Saliva Ontology is an ongoing exploratory initiative. The ontology will be used to facilitate salivaomics data retrieval and integration across multiple fields of research together with data analysis and data mining. The ontology will be tested through its ability to serve the annotation ('tagging') of a representative corpus of salivaomics research literature that is to be incorporated into the SKB. Background Saliva (oral fluid) is an emerging biofluid for non-invasive diagnostics used in the detection of human diseases. The need to advance saliva research is strongly emphasized by the National Institute of Dental and Craniofacial Research (NIDCR), and is included in the NIDCR's 2004- 2009 expert panel long-term research agenda [1]. The ability to monitor health status, disease onset, progression, recurrence and treatment outcome through noninvasive means is highly important to advancing health care management. Saliva is a perfect medium to be explored for personalized individual medicine including diagnostics, offering a non-invasive, easy to obtain means for detecting and monitoring diseases. Saliva testing potentially allows the patient to collect their own saliva samples at home, yielding convenience for the patient and savings in health costs, and facilitating multiple sampling. Specimen collection is less objectionable to patients and easier in children and elderly individuals. Due to these advantages. (shrink)

We describe the rationale for an application ontology covering the domain of human body fluids that is designed to facilitate representation, reuse, sharing and integration of diagnostic, physiological, and biochemical data, We briefly review the Blood Ontology (BLO), Saliva Ontology (SALO) and Kidney and Urinary Pathway Ontology (KUPO) initiatives. We discuss the methods employed in each, and address the project of using them as starting point for a unified body fluids ontology resource. We conclude with a description of how the (...) body fluids ontology initiative may provide support to basic and translational science. (shrink)

There is a need recognized by the National Institute of Dental & Craniofacial Research and the National Cancer Institute to advance basic, translational and clinical saliva research. The goal of the Salivaomics Knowledge Base (SKB) is to create a data management system and web resource constructed to support human salivaomics research. To maximize the utility of the SKB for retrieval, integration and analysis of data, we have developed the Saliva Ontology and SDxMart. This article reviews the informatics advances in saliva (...) diagnostics made possible by the Saliva Ontology and SDxMart. (shrink)

Hiện nay, chế độ sở hữu ở Việt Nam đã có những đổi mới cơ bản, nhưng vẫn còn những khác biệt rất lớn so với chế độ sở hữu ở các nền kinh tế thị trường hiện đại. Trong cơ cấu của chế độ sở hữu ở Việt Nam, tỷ trọng của sở hữu nhà nước còn quá lớn; kinh tế nhà nước giữ vai trò chủ đạo… Chính những khác biệt này đã làm cho nền kinh tế thị (...) trường Việt Nam bị méo mó và giảm lợi thế cạnh tranh quốc gia. Do vậy, công cuộc đổi mới và hoàn thiện chế độ sở hữu ở Việt Nam phải hướng tới việc khắc phục những hạn chế trên và đạt tới sức cạnh tranh ngang hàng với các nền kinh tế thị trường hiện đại khác. (shrink)

Hiện nay, chế độ sở hữu ở Việt Nam đã có những đổi mới cơ bản, nhưng vẫn còn những khác biệt rất lớn so với chế độ sở hữu ở các nền kinh tế thị trường hiện đại. Trong cơ cấu của chế độ sở hữu ở Việt Nam, tỷ trọng của sở hữu nhà nước còn quá lớn; kinh tế nhà nước giữ vai trò chủ đạo… Chính những khác biệt này đã làm cho nền kinh tế thị (...) trường Việt Nam bị méo mó và giảm lợi thế cạnh tranh quốc gia. Do vậy, công cuộc đổi mới và hoàn thiện chế độ sở hữu ở Việt Nam phải hướng tới việc khắc phục những hạn chế trên và đạt tới sức cạnh tranh ngang hàng với các nền kinh tế thị trường hiện đại khác. (shrink)

ĐỀ CƯƠNG HỌC PHẦN VĂN HÓA KINH DOANH 1. Tên học phần: VĂN HÓA KINH DOANH (BUSINESS CULTURE) 2. Mã học phần: BMGM1221 3. Số tín chỉ: 2 (24,6) (để học được học phần này, người học phải dành ít nhất 60 giờ chuẩn bị cá nhân).

(Mặt trận) - Chế độ sở hữu trong nền kinh tế thị trường định hướng xã hội chủ nghĩa Việt Nam trước hết phải tuân theo các nguyên tắc của nền kinh tế thị trường hiện đại. Trong các nguyên tắc của nền kinh tế thị trường hiện đại, nguyên tắc sở hữu tư nhân là nền tảng của nền kinh tế thị trường - là nguyên tắc quan trọng. Xa rời nguyên tắc này, dù chúng ta cố gắng xây (...) dựng nền kinh tế thị trường thì đó vẫn là một nền kinh tế thị trường khó được quốc tế công nhận. Sở hữu tư nhân hiện đại là chế độ sở hữu được cổ phần hóa, được xã hội hóa rất cao, công khai, minh bạch trên các thị trường chứng khoán. Tính xã hội cao của chế độ cổ phần hóa đã được C.Mác đánh giá, xem như là chế độ sở hữu quá độ sang hình thái kinh tế - xã hội mới. (shrink)

Despite the importance of human blood to clinical practice and research, hematology and blood transfusion data remain scattered throughout a range of disparate sources. This lack of systematization concerning the use and definition of terms poses problems for physicians and biomedical professionals. We are introducing here the Blood Ontology, an ongoing initiative designed to serve as a controlled vocabulary for use in organizing information about blood. The paper describes the scope of the Blood Ontology, its stage of development and some (...) of its anticipated uses. (shrink)

Since anonymous agents can spread misinformation with impunity, many people advocate for greater accountability for internet speech. This paper provides a veritistic argument that accountability mechanisms can cause significant epistemic problems for internet encyclopedias and social media communities. I show that accountability mechanisms can undermine both the dissemination of true beliefs and the detection of error. Drawing on social psychology and behavioral economics, I suggest alternative mechanisms for increasing the trustworthiness of internet communication.

Situationists such as John Doris, Gilbert Harman, and Maria Merritt suppose that appeal to reliable behavioral dispositions can be dispensed with without radical revision to morality as we know it. This paper challenges this supposition, arguing that abandoning hope in reliable dispositions rules out genuine trust and forces us to suspend core reactive attitudes of gratitude and resentment, esteem and indignation. By examining situationism through the lens of trust we learn something about situationism (in particular, the radically revisionary moral implications (...) of its adoption) as well as something about trust (in particular, that the conditions necessary for genuine trust include a belief in a capacity for robust dispositions). (shrink)

Overconfidence is typically damaging to one’s reputation as a trustworthy source of information. Previous research shows that the reputational cost associated with conveying a piece of false information is higher for confident than unconfident speakers. When judging speaker trustworthiness, individuals do not exclusively rely on past accuracy but consider the extent to which speakers expressed a degree of confidence that matched the accuracy of their claims (their “confidence-accuracy calibration”). The present study experimentally examines the interplay between confidence, accuracy and (...) a third factor, namely evidence, in the assessment of speaker trustworthiness. Experiment 1 probes the hypothesis that overconfidence does not backfire when a confident but inaccurate claim is justified: the trustworthiness of a confident speaker who turns out to be wrong is restored if the confidence expressed is based on strong evidence (good confidence-evidence calibration). Experiment 2 investigates the hypothesis that confidence can backfire if a confident and accurate claim is not justified: the trustworthiness of a confident speaker who turns out to be right is damaged if the confidence expressed is based on weak evidence (bad confidence-evidence calibration). Our results support both hypotheses and thus suggest that “confidence-evidence calibration” plays a crucial role in the assessment of speaker trustworthiness. (shrink)

Artificial intelligence (AI) research enjoyed an initial period of enthusiasm in the 1970s and 80s. But this enthusiasm was tempered by a long interlude of frustration when genuinely useful AI applications failed to be forthcoming. Today, we are experiencing once again a period of enthusiasm, fired above all by the successes of the technology of deep neural networks or deep machine learning. In this paper we draw attention to what we take to be serious problems underlying current views of artificial (...) intelligence encouraged by these successes, especially in the domain of language processing. We then show an alternative approach to language-centric AI, in which we identify a role for philosophy. (shrink)

This essay analyses the Confucian conception of xin, an attribute that broadly resembles what we would ordinarily call trustworthiness. More specifically, it provides an analysis of the psychology of someone who is xin and highlights a feature of the Confucian conception of trustworthiness: the trustworthy person has to ensure that there is a match between her self-presentation and the way she is. My goal is not to argue against any of the existing accounts of trustworthiness but to draw on (...) Confucian insights so as to shed light on features of trustworthiness that are overlooked in current discussions. I hope to show that the Confucian conception of trustworthiness puts more emphasis on the way a trustworthy person actively tries to make sure another’s dependency on her is not unwarranted than on how the trustworthy person responds to the one who gives trust. (shrink)

Abstract: We examine how trustworthy behaviour can be achieved in the financial sector. The task is to ensure that firms are motivated to pursue long-term interests of customers rather than pursuing short-term profits. Firms’ self-interested pursuit of reputation, combined with regulation, is often not sufficient to ensure that this happens. We argue that trustworthy behaviour requires that at least some actors show a concern for the wellbeing of clients, or a respect for imposed standards, and that the behaviour (...) of these actors is copied in such a way that it becomes a behavioural norm. We briefly suggest what such behavioural norms might need to be if trustworthy behaviour is to be achieved, and consider how they might be supported; we describe the research that is necessary in order to understand these norms in more detail. We argue that the norms of traders are different from the norms of those engaged in other activities, since they are inevitably self-interested, and we consider the risk that traders’ norms might undermine those of other actors. We analyse the task for governance in dealing with this problem, and the role which leadership by a corporate board and management might play in doing this. We describe the need for further research in to describe how this might be done. (shrink)

Under the Superstition Mountains in central Arizona toil those who would rob humankind o f its humanity. These gray, soulless monsters methodically tear away at our meaning, our subjectivity, our essence as transcendent beings. With each advance, they steal our freedom and dignity. Who are these denizens of darkness, these usurpers of all that is good and holy? None other than humanity’s arch-foe: The Cognitive Scientists -- AI researchers, fallen philosophers, psychologists, and other benighted lovers of computers. Unless they are (...) stopped, humanity -- you and I -- will soon be nothing but numbers and algorithms locked away on magnetic tape. (shrink)

An advanced artificial intelligence could pose a significant existential risk to humanity. Several research institutes have been set-up to address those risks. And there is an increasing number of academic publications analysing and evaluating their seriousness. Nick Bostrom’s superintelligence: paths, dangers, strategies represents the apotheosis of this trend. In this article, I argue that in defending the credibility of AI risk, Bostrom makes an epistemic move that is analogous to one made by so-called sceptical theists in the debate about the (...) existence of God. And while this analogy is interesting in its own right, what is more interesting are its potential implications. It has been repeatedly argued that sceptical theism has devastating effects on our beliefs and practices. Could it be that AI-doomsaying has similar effects? I argue that it could. Specifically, and somewhat paradoxically, I argue that it could amount to either a reductio of the doomsayers position, or an important and additional reason to join their cause. I use this paradox to suggest that the modal standards for argument in the superintelligence debate need to be addressed. (shrink)

Trust can be thought of as a three place relation: A trusts B to do X. Trustworthiness has two components: competence (does the trustee have the relevant skills, knowledge and abilities to do X?) and willingness (is the trustee intending or aiming to do X?). This chapter is about the willingness component, and the different motivations that a trustee may have for fulfilling trust. The standard assumption in economics is that agents are self-regarding, maximizing their own consumption of goods and (...) services. This is too restrictive. In particular, people may be concerned with the outcomes of others, and they may be concerned to follow ethical principles. I distinguish weak trustworthiness, which places no restrictions on B’s motivation for doing X, from strong trustworthiness, where the behaviour must have a particular non-selfish motivation, in finance the fiduciary commitment to promote the interests of the truster. I discuss why strong trustworthiness may be more efficient and also normatively preferable to weak. In finance, there is asymmetric information between buyer and seller, which creates a need for trustworthy assessment of products. It also creates an ambiguity about whether the relationship is one of buyer and seller, governed by caveat emptor, or a fiduciary relationship of advisor and client. This means that there are two possible reasons why trust may be breached: because the trustee didn’t realise that the truster framed the relationship as a fiduciary one, or because the trustee did realise but actively sought to take advantage of the trust. Correspondingly, there are two possible types of agent: normal people who are not always self-regarding and who are trust responsive (if they believe that they are being trusted then they are likely to fulfill that trust), and knaves, after Hume’s character who is always motivated by his own private interest. We can increase the trustworthiness of normal people by getting them to re-frame the situation as one of trust, so they will be strongly trustworthy (i.e. change of institutional culture), and by providing non-monetary incentives (the correct choice of incentive will depend on exactly what their non-selfish motivation is). Knaves need sanctions, which can make them weakly trustworthy. However, this is a delicate balance because sanctions can crowd out normative frames. We can also increase the trustworthiness of financiers by making finance less attractive to knaves; changing the mix of types in finance could help support the necessary cultural change. (shrink)

A widespread assumption in debates about trust and trustworthiness is that the evaluative norms of principal interest on the trustor’s side of a cooperative exchange regulate trusting attitudes and performances whereas those on the trustee’s side regulate dispositions to respond to trust. The aim here will be to highlight some unnoticed problems with this asymmetrical picture – and in particular, how it elides certain key evaluative norms on both the trustor’s and trustee’s side the satisfaction of which are critical to (...) successful cooperative exchanges – and to show that a symmetrical, ‘achievement-first’ approach to theorising about trust and trustworthiness (and their relation to each other) has important advantages by comparison. The view I develop is guided by a structural analogy with practical reasoning. Just as practical reasoning is working as it should only when there is realisation (knowledge and action) of states (belief and intention) with reverse directions of fits (mind-to-world and world-to-mind), likewise, cooperation between trustor and trustee is functioning as it should only when there is an analogous kind of realisation on both sides of the cooperative exchange – viz., when the trustor ‘matches’ her achievement in trusting (an achievement in fitting reliance to reciprocity) with the trustee’s achievement in responding to trust (an achievement in fitting reciprocity to reliance). An upshot of viewing cooperation between trustor and trustee as exhibiting achievement-theoretic structure is that we will be better positioned to subsume trustworthiness (and its cognates on the trustee’s side), like trust, under a wider suite of evaluative norms that regulate attempts, dispositions, and achievements symmetrically on both sides of a cooperative exchange, with ‘matching achievements’ as the gold standard. (shrink)

Cyberspace relies on information technologies to mediate relations between different people, across different communication networks and is reliant on the supporting technology. These interactions typically occur without physical proximity and those working depending on cybersystems must be able to trust the overall human–technical systems that support cyberspace. As such, detailed discussion of cybersecurity policy would be improved by including trust as a key value to help guide policy discussions. Moreover, effective cybersystems must have resilience designed into them. This paper argues (...) that trustworthy cybersystems are a key element to resilient systems, and thus are core to cybersecurity policy. The paper highlights the importance of trustworthiness for resilient cybersystems. The importance of trustworthiness is shown through a discussion of three events where trustworthiness was the target or casualty of cyberattacks: Stuxnet, hacking of communications and the Edward Snowden revelations. The impact of losing trust is highlighted, to underpin the argument that a resilient cybersystem ought to design in trustworthiness. The paper closes off by presenting a general set of policy implications arising from recognition of the interplay between trust, trustworthiness and resilience for effective cybersecurity. (shrink)

The expanding integration of artificial intelligence (AI) in various aspects of society makes the infosphere around us increasingly complex. Humanity already faces many obstacles trying to have a better understanding of our own minds, but now we have to continue finding ways to make sense of the minds of AI. The issue of AI’s capability to have independent thinking is of special attention. When dealing with such an unfamiliar concept, people may rely on existing human properties, such as survival desire, (...) to make assessments. Employing information-processing-based Bayesian Mindsponge Framework (BMF) analytics on a dataset of 266 residents in the United States, we found that the more people believe that an AI agent seeks continued functioning, the more they believe in that AI agent’s capability of having a mind of its own. Moreover, we also found that the above association becomes stronger if a person is more familiar with personally interacting with AI. This suggests a directional pattern of value reinforcement in perceptions of AI. As the information processing of AI becomes even more sophisticated in the future, it will be much harder to set clear boundaries about what it means to have an autonomous mind. (shrink)

Implicit stochastic models, including both ‘deep neural networks’ (dNNs) and the more recent unsupervised foundational models, cannot be explained. That is, it cannot be determined how they work, because the interactions of the millions or billions of terms that are contained in their equations cannot be captured in the form of a causal model. Because users of stochastic AI systems would like to understand how they operate in order to be able to use them safely and reliably, there has emerged (...) a new field called ‘explainable AI’ (XAI). When we examine the XAI literature, however, it becomes apparent that its protagonists have redefined the term ‘explanation’ to mean something else, namely: ‘interpretation’. Interpretations are indeed sometimes possible, but we show that they give at best only a subjective understanding of how a model works. We propose an alternative to XAI, namely certified AI (CAI), and describe how an AI can be specified, realized, and tested in order to become certified. The resulting approach combines ontologies and formal logic with statistical learning to obtain reliable AI systems which can be safely used in technical applications. (shrink)

The hypergrowth and hyperconnectivity of networks of artificial intelligence (AI) systems and algorithms increasingly cause our interactions with the world, socially and environmentally, more technologically mediated. AI systems start interfering with our choices or making decisions on our behalf: what we see, what we buy, which contents or foods we consume, where we travel to, who we hire, etc. It is imperative to understand the dynamics of human-computer interaction in the age of progressively more competent AI. This essay presents five (...) premises to serve as starting points for building our understanding of the “interactions’ in human-computer interactions. (shrink)

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

Does AI conform to humans, or will we conform to AI? An ethical evaluation of AI-intensive companies will allow investors to knowledgeably participate in the decision. The evaluation is built from nine performance indicators that can be analyzed and scored to reflect a technology’s human-centering. When summed, the scores convert into objective investment guidance. The strategy of incorporating ethics into financial decisions will be recognizable to participants in environmental, social, and governance investing, however, this paper argues that conventional ESG frameworks (...) are inadequate for AI-intensive companies. To fully account for contemporary technology, the following categories of evaluation will be developed and featured as vital investing criteria: autonomy, dignity, privacy, performance. With these priorities established, the larger goal is a model for humanitarian investing in AI-intensive companies that is intellectually robust, manageable for analysts, useful for portfolio managers, and credible for investors. (shrink)

This paper examines how science advice can provide policy recommendations in a trustworthy manner. Despite their major political importance, expert recommendations are understudied in the philosophy of science and social epistemology. Matthew Bennett has recently developed a notion of what he calls recommendation trust, according to which well-placed trust in experts’ policy recommendations requires that recommendations are aligned with the interests of the trust-giver. While interest alignment might be central to some cases of public trust, this paper argues against (...) the significance of interest-alignment to meritorious public trust. First, political bodies and citizens can have a basic kind of well-placed recommendation trust in science advice based on an all-things-considered judgement regarding the possession of relevant competencies, responsible conduct, and a proper institutional design. Moreover, scientists’ policy recommendations can be seen as open-ended and as dynamic proposals that enable inter-institutional reasoning and political deliberation. Finally, by providing conditional recommendations, scientists can expand the scope of feasible policy options from which political bodies can choose, thus making the condition of interest alignment even less significant to the trustworthy provision of recommendations. (shrink)

The increasing demand for transparency in AI has recently come under scrutiny. The question is often posted in terms of “epistemic double standards”, and whether the standards for transparency in AI ought to be higher than, or equivalent to, our standards for ordinary human reasoners. I agree that the push for increased transparency in AI deserves closer examination, and that comparing these standards to our standards of transparency for other opaque systems is an appropriate starting point. I suggest that a (...) more fruitful exploration of this question will involve a different comparison class. We routinely treat judgments made by highly trained experts in specialized fields as fair or well-grounded even though—by the nature of expert/layperson division of epistemic labor—an expert will not be able to provide an explanation of the reasoning behind these judgments that makes sense to most other people. Regardless, laypeople are thought to be acting reasonably—and ethically—in deferring to the judgments of experts that concern their areas of specialization. I suggest that we reframe our question regarding the appropriate standards of transparency in AI as one that asks when, why, and to what degree it would be ethical to accept opacity in AI. I argue that our epistemic relation to certain opaque AI technology may be relevantly similar to the layperson’s epistemic relation to the expert in certain respects, such that the successful expert/layperson division of epistemic labor can serve as a blueprint for the ethical use of opaque AI. (shrink)

Trustworthiness is typically regarded as a desirable feature of national identification systems (NISs); but the variegated nature of the trustor communities associated with such systems makes it difficult to see how a single system could be equally trustworthy to all actual and potential trustors. This worry is accentuated by common theoretical accounts of trustworthiness. According to such accounts, trustworthiness is relativized to particular individuals and particular areas of activity, such that one can be trustworthy with regard to some (...) individuals in respect of certain matters, but not trustworthy with regard to all trustors in respect of every matter. The present article challenges this relativistic approach to trustworthiness by outlining a new account of trustworthiness, dubbed the expectation-oriented account. This account allows for the possibility of an absolutist (or one-place) approach to trustworthiness. Such an account, we suggest, is the approach that best supports the effort to develop NISs. To be trustworthy, we suggest, is to minimize the error associated with trustor expectations in situations of social dependency (commonly referred to as trust situations), and to be trustworthy in an absolute sense is to assign equal value to all expectation-related errors in all trust situations. In addition to outlining the features of the expectation-oriented account, we describe some of the implications of this account for the design, development, and management of trustworthy NISs. (shrink)

Since the release of OpenAI's ChatGPT, the world has entered a race to develop more capable and powerful AI, including artificial general intelligence (AGI). The development is constrained by the dependency of AI on the model, quality, and quantity of training data, making the AI training process highly costly in terms of resources and environmental consequences. Thus, improving the effectiveness and efficiency of the AI training process is essential, especially when the Earth is approaching the climate tipping points and planetary (...) boundaries. It is evident that AI can function better when trained to mimic certain human mental processes. Based on insights from quantum mechanics and the informational entropy-based notion of value, we suggest AI developers can enhance the AI training processes' effectiveness and efficiency by creating and implementing parameter systems that are capable of effectively assigning probabilities to informational quanta. Such systems can help reduce entropy within the system, which lowers the energy required for data storage and processing while decreasing the likelihood of information loss during training. Successfully applying the concept of information entropy-based value in AI development will be crucial for advancing generative AI and achieving AGI in a sustainable manner. (shrink)