Science and Engineering Ethics最新文献

Engineering ethics is a required aspect of accredited ABET programs, but there is widespread variation in how ethics is taught, to what ends, and how those ends are assessed. This variation makes it challenging to identify practices for teaching ethics to engineers aligned with extant practices in the field. In this study, we revise a recent coding framework by reviewing exemplary engineering ethics programs recognized by the National Academy of Engineering in 2016, or what we refer to as "exemplars." We pursue two primary objectives: (1) To apply and revise a prior coding framework to codify ethics learning objectives, instructional strategies, and assessment strategies in engineering education; and (2) To use the revised coding framework to identify trends in learning objectives, instructional strategies, and assessment strategies of NAE exemplars. We employ systemic review procedures to update the coding framework using 24 of 25 exemplars as a data source. The updated framework includes four primary categories associated with learning objectives, instructional strategies, assessment data collection strategies, and assessment design characteristics. Results indicate that ethical sensitivity or awareness was present in every exemplar as a learning objective, often alongside ethical reasoning-based learning objectives and the formation of professional skills. Exemplars employed numerous instructional strategies in tandem, as we coded eight out of 18 instructional strategies among at least half of the exemplars. Assignments/homework and summative reflections were the most oft-used sources of assessment data. Due to our challenges in coding assessment approaches, we offer practical suggestions for assessing engineering ethics instruction which are based on many of our coding discussions. We hope that this coding framework, the results classifying exemplary features of the NAE programs, and our practical suggestions can guide future instructors as they design, classify, assess, and report their approaches to engineering ethics education.

There is an ongoing debate about genetic engineering (GE) in food production. Supporters argue that it makes crops more resilient to stresses, such as drought or pests, and should be considered by researchers as a technology to address issues of global food security, whereas opponents put forward that GE crops serve only the economic interests of transnational agrifood-firms and have not yet delivered on their promises to address food shortage and nutrient supply. To address discourse failure regarding the GE debate, research needs to understand better what drives the divergent positions and which moral attitudes fuel the mental models of GE supporters and opponents. Hence, this study investigates moral attitudes regarding GE opposition and support in Germany. Results show that GE opponents are significantly more absolutist than supporters and significantly less likely to hold outcome-based views. Furthermore, GE opponents are more willing to donate for preventing GE admission than supporters are willing to donate for promoting GE admission. Our results shed light on why the divide between opponents and supporters in the German GE debate could remain stark and stable for so long.

It may soon become possible not just to replace, but to re-grow healthy tissues after injury or disease, because of innovations in the field of Regenerative Medicine. One particularly promising innovation is a regenerative valve implant to treat people with heart valve disease. These implants are fabricated from so-called 'smart', 'lifelike' materials. Implanted inside a heart, these implants stimulate re-growth of a healthy, living heart valve. While the technological development advances, the ethical implications of this new technology are still unclear and a clear conceptual understanding of the notions 'smart' and 'lifelike' is currently lacking. In this paper, we explore the conceptual and ethical implications of the development of smart lifelike materials for the design of regenerative implants, by analysing heart valve implants as a showcase. In our conceptual analysis, we show that the materials are considered 'smart' because they can communicate with human tissues, and 'lifelike' because they are structurally similar to these tissues. This shows that regenerative valve implants become intimately integrated in the living tissues of the human body. As such, they manifest the ontological entanglement of body and technology. In our ethical analysis, we argue this is ethically significant in at least two ways: It exacerbates the irreversibility of the implantation procedure, and it might affect the embodied experience of the implant recipient. With our conceptual and ethical analysis, we aim to contribute to responsible development of smart lifelike materials and regenerative implants.

In this paper I argue that Artificial Intelligence and the many data science methods associated with it, such as machine learning and large language models, are first and foremost epistemic technologies. In order to establish this claim, I first argue that epistemic technologies can be conceptually and practically distinguished from other technologies in virtue of what they are designed for, what they do and how they do it. I then proceed to show that unlike other kinds of technology (including other epistemic technologies) AI can be uniquely positioned as an epistemic technology in that it is primarily designed, developed and deployed to be used in epistemic contexts such as inquiry, it is specifically designed, developed and deployed to manipulate epistemic content such as data, and it is designed, developed and deployed to do so particularly through epistemic operations such as prediction and analysis. As has been shown in recent work in the philosophy and ethics of AI (Alvarado, AI and Ethics, 2022a), understanding AI as an epistemic technology will also have significant implications for important debates regarding our relationship to AI technologies. This paper includes a brief overview of such implications, particularly those pertaining to explainability, opacity, trust and even epistemic harms related to AI technologies.

If existing or future robots appear to have some capacity, state or property, how can we determine whether they truly have it or whether we are deceived into believing so? John Danaher addresses this question by formulating his approach to what he refers to as superficial state deception (SSD) from the perspective of his theory termed ethical behaviourism (EB), which was initially designed to determine the moral status of robots. In summary, Danaher believes that focusing on behaviour is sufficient to determine whether SSD occurs. My general claim is that Danaher's approach to SSD based on EB is implausible since it results in the impossibility of conceptualizing SSD, e.g., it does not enable determining whether or not SSD occurs in a particular case. Moreover, I show how Danaher's approach to SSD needs to be transformed to become plausible. To make my point, I (1) examine the main features of EB and distinguish its two versions by showing how Danaher revised the original EB in response to criticism; (2) discuss Danaher's approach to the problem of deception from the perspective of EB; (3) criticize that approach by showing that it requires revisions analogous to those that have already been recommended in reference to EB, and (4) propose an alternative method for determining the presence of SSD that covers diverse, plausible approaches to SSD.

This article suggests several design principles intended to assist in the development of ethical algorithms exemplified by the task of fighting fake news. Although numerous algorithmic solutions have been proposed, fake news still remains a wicked socio-technical problem that begs not only engineering but also ethical considerations. We suggest employing insights from ethics of care while maintaining its speculative stance to ask how algorithms and design processes would be different if they generated care and fight fake news. After reviewing the major characteristics of ethics of care and the phases of care, we offer four algorithmic design principles. The first principle highlights the need to develop a strategy to deal with fake news on the part of the software designers. The second principle calls for the involvement of various stakeholders in the design processes in order to increase the chances of successfully fighting fake news. The third principle suggests allowing end-users to report on fake news. Finally, the last principle proposes keeping the end-user updated on the treatment in the suspected news items. Implementing these principles as care practices can render the developmental process more ethically oriented as well as improve the ability to fight fake news.

This paper clarifies why bias cannot be completely mitigated in Machine Learning (ML) and proposes an end-to-end methodology to translate the ethical principle of justice and fairness into the practice of ML development as an ongoing agreement with stakeholders. The pro-ethical iterative process presented in the paper aims to challenge asymmetric power dynamics in the fairness decision making within ML design and support ML development teams to identify, mitigate and monitor bias at each step of ML systems development. The process also provides guidance on how to explain the always imperfect trade-offs in terms of bias to users.

To foster research integrity (RI), research institutions should develop a continuous RI education approach, addressing various target groups. To support institutions to achieve this, we developed RI education guidelines together with RI experts and research administrators, exploring similarities and differences in recommendations across target groups, as well as recommendations about RI education using approaches other than formal RI training. We used an iterative co-creative process. We conducted four half-day online co-creation workshops with 16 participants in total, which were informed by the RI education evidence-base. In the first two workshops, participants generated ideas for guidelines' content, focusing on different target groups and various approaches to RI education. Based on this content we developed first drafts of the guidelines. Participants in the third and fourth workshop refined those drafts. We next organized a working group which further prioritized, reorganized, and optimized the content of the guidelines. We developed four guidelines on RI education focusing on (a) bachelor, master and PhD students; (b) post-doctorate and senior researchers; (c) other RI stakeholders; as well as (d) continuous RI education. Across guidelines, we recommend mandatory RI training; follow-up refresher training; informal discussions about RI; appropriate rewards and incentives for active participation in RI education; and evaluation of RI educational events. Our work provides experience-based co-created guidance to research institutions on what to consider when developing a successful RI education strategy. Each guideline is offered as a distinct, publicly available tool in our toolbox ( www.sops4ri.eu/toolbox ) which institutions can access, adapt and implement to meet their institution-specific RI education needs.Trial registration https://osf.io/zej5b .

As artificial intelligence becomes more sophisticated and robots approach autonomous decision-making, debates about how to assign moral responsibility have gained importance, urgency, and sophistication. Answering Stenseke's (2022a) call for scaffolds that can help us classify views and commitments, we think the current debate space can be represented hierarchically, as answers to key questions. We use the resulting taxonomy of five stances to differentiate-and defend-what is known as the "blank check" proposal. According to this proposal, a person activating a robot could willingly make themselves answerable for whatever events ensue, even if those events stem from the robot's autonomous decision(s). This blank check solution was originally proposed in the context of automated warfare (Champagne & Tonkens, 2015), but we extend it to cover all robots. We argue that, because moral answerability in the blank check is accepted voluntarily and before bad outcomes are known, it proves superior to alternative ways of assigning blame. We end by highlighting how, in addition to being just, this self-initiated and prospective moral answerability for robot harm provides deterrence that the four other stances cannot match.