The ethics of designing artificial agents

In their important paper “Autonomous Agents”, Floridi and Sanders use “levels of abstraction” to argue that computers are or may soon be moral agents. In this paper we use the same levels of abstraction to illuminate differences between human moral agents and computers. In their paper, Floridi and Sanders contributed definitions of autonomy, moral accountability and responsibility, but they have not explored deeply some essential questions that need to be answered by computer scientists who design artificial agents. One such question is, “Can an artificial agent that changes its own programming become so autonomous that the original designer is no longer responsible for the behavior of the artificial agent?” To explore this question, we distinguish between LoA1 (the user view) and LoA2 (the designer view) by exploring the concepts of unmodifiable, modifiable and fully modifiable tables that control artificial agents. We demonstrate that an agent with an unmodifiable table, when viewed at LoA2, distinguishes an artificial agent from a human one. This distinction supports our first counter-claim to Floridi and Sanders, namely, that such an agent is not a moral agent, and the designer bears full responsibility for its behavior. We also demonstrate that even if there is an artificial agent with a fully modifiable table capable of learning* and intentionality* that meets the conditions set by Floridi and Sanders for ascribing moral agency to an artificial agent, the designer retains strong moral responsibility.     

Moral luck and computer ethics: Gauguin in cyberspace

I argue that the problem of ‘moral luck’ is an unjustly neglected topic within Computer Ethics. This is unfortunate given that the very nature of computer technology, its ‘logical malleability’, leads to ever greater levels of complexity, unreliability and uncertainty. The ever widening contexts of application in turn lead to greater scope for the operation of chance and the phenomenon of moral luck. Moral luck bears down most heavily on notions of professional responsibility, the identification and attribution of responsibility. It is immunity from luck that conventionally marks out moral value from other kinds of values such as instrumental, technical, and use value. The paper describes the nature of moral luck and its erosion of the scope of responsibility and agency. Moral luck poses a challenge to the kinds of theoretical approaches often deployed in Computer Ethics when analyzing moral questions arising from the design and implementation of information and communication technologies. The paper considers the impact on consequentialism; virtue ethics; and duty ethics. In addressing cases of moral luck within Computer Ethics, I argue that it is important to recognise the ways in which different types of moral systems are vulnerable, or resistant, to moral luck. Different resolutions are possible depending on the moral framework adopted. Equally, resolution of cases will depend on fundamental moral assumptions. The problem of moral luck in Computer Ethics should prompt us to new ways of looking at risk, accountability and responsibility.     

无情感的机器人需要道德关护吗？——未来AMAs机器人的情感因素分析

目前应用领域的机器人缺乏意识、精神状态和感觉这些情感条件,机器人只是按照人类设定的程序进行遵循一定的规则行为。判定一个机器人能否称得上人工物道德行为体（AMAs）．似乎取决于是否具有情感因素。道德与情感之间有着紧密联系的关系。然而．行为主义和表现主义认为,即使缺乏情感的机器也应当受到道德关护。从机器人的应用实践来看．无论是认知缺陷角色的机器人、奴仆角色机器人还是财产物角色机器人．他们都有相应的道德地位,都应当受到不同方式的伦理关护。随着人工智能的发展,我们认为,未来我们一定能够制造出一种具有情感的AMAs机器人。     

How to be a responsible slave: managing the use of expert information systems

Computer ethicists have for some years been troubled by the issue of how to assign moral responsibility for disastrous events involving erroneous information generated by expert information systems. Recently, Jeroen van den Hoven has argued that agents working with expert information systems satisfy the conditions for what he calls epistemic enslavement. Epistemically enslaved agents do not, he argues, have moral responsibility for accidents for which they bear causal responsibility. In this article, I develop two objections to van den Hoven’s argument for epistemic enslavement of agents working with expert information systems.     

Ethics for things

This paper considers the ways that Information Ethics (IE) treats things. A number of critics have focused on IE’s move away from anthropocentrism to include non-humans on an equal basis in moral thinking. I enlist Actor Network Theory, Dennett’s views on ?as if’ intentionality and Magnani’s characterization of ?moral mediators’. Although they demonstrate different philosophical pedigrees, I argue that these three theories can be pressed into service in defence of IE’s treatment of things. Indeed the support they lend to the extension of moral status to non-human objects can be seen as part of a trend towards the accommodation of non-humans into our moral and social networks. A number of parallels are drawn between philosophical arguments over artificial intelligence and information ethics.     

Moral Responsibility for Harm Caused by Computer System Failures

When software is written and then utilized in complex computer systems, problems often occur. Sometimes these problems cause a system to malfunction, and in some instances such malfunctions cause harm. Should any of the persons involved in creating the software be blamed and punished when a computer system failure leads to persons being harmed? In order to decide whether such blame and punishment are appropriate, we need to first consider if the people are “morally responsible”. Should any of the people involved in creating the software be held morally responsible, as individuals, for the harm caused by a computer system failure?This article provides one view of moral responsibility and then discusses some barriers to holding people morally responsible. Next, it provides information about the Therac-25, a computer-controlled medical linear accelerator, and its computer systems failures that led to deaths and injuries. Finally it investigates whether two key people involved in the Therac-25 case could reasonably be considered to have some degree of moral responsibility for the deaths and injuries. The conclusions about whether or not these people were morally responsible necessarily rest upon a certain amount of speculation about what they knew and what they did. These limitations, however, should not cause us to conclude that discussions of moral responsibility are fruitless. In some cases, determinations of moral responsibility may be made and in others the investigation is still worthwhile, as the article demonstrates.     

当代青年提升自我思想道德修养方法探究

人类社会是多层次的有机的集合体,个人作为社会最基本的分子,每一个人的行为选择都会对社会和谐与发展产生一定的影响,表现为个人道德修养水平越高,整个社会的文明程度也越高。当代青年是我国社会主义建设事业的接班人和建设者,担负着实现我国伟大复兴的重大责任,他们更应该提高综合素质,提升自身思想道德修养。个人修养中包括礼仪、道德、思想、文化等内容,本文将侧重从提升青年的自我思想品德方面作一些阐述。     

Un-making artificial moral agents

Floridi and Sanders, seminal work, “On the morality of artificial agents” has catalyzed attention around the moral status of computer systems that perform tasks for humans, effectively acting as “artificial agents.” Floridi and Sanders argue that the class of entities considered moral agents can be expanded to include computers if we adopt the appropriate level of abstraction. In this paper we argue that the move to distinguish levels of abstraction is far from decisive on this issue. We also argue that adopting certain levels of abstraction out of context can be dangerous when the level of abstraction obscures the humans who constitute computer systems. We arrive at this critique of Floridi and Sanders by examining the debate over the moral status of computer systems using the notion of interpretive flexibility. We frame the debate as a struggle over the meaning and significance of computer systems that behave independently, and not as a debate about the ‘true’ status of autonomous systems. Our analysis leads to the conclusion that while levels of abstraction are useful for particular purposes, when it comes to agency and responsibility, computer systems should be conceptualized and identified in ways that keep them tethered to the humans who create and deploy them.     

On the intrinsic value of information objects and the infosphere

What is the most general common set ofattributes that characterises something asintrinsically valuableand hence as subject to some moral respect, andwithout which something would rightly beconsidered intrinsically worthless or even positivelyunworthy and therefore rightly to bedisrespected in itself? Thispaper develops and supports the thesis that theminimal condition of possibility of an entity'sleast intrinsic value is to be identified with itsontological status as an information object.All entities, even when interpreted as only clusters ofinformation, still have a minimal moral worthqua information objects and so may deserve to be respected. Thepaper is organised into four main sections.Section 1 models moral action as an information systemusing the object-oriented programmingmethodology (OOP). Section 2 addresses the question of whatrole the several components constituting themoral system can have in an ethical analysis. If theycan play only an instrumental role, thenComputer Ethics (CE) is probably bound to remain at most apractical, field-dependent, applied orprofessional ethics. However, Computer Ethics can give rise to amacroethical approach, namely InformationEthics (IE), if one can show that ethical concern should beextended to include not only human, animal orbiological entities, but also information objects. Thefollowing two sections show how this minimalistlevel of analysis can be achieved. Section 3 provides anaxiological analysis of information objects. Itcriticises the Kantian approach to the concept ofintrinsic value and shows that it can beimproved by using the methodology introduced in the first section.The solution of the Kantian problem prompts thereformulation of the key question concerningthe moral worth of an entity: what is theintrinsic value of x qua an object constituted by itsinherited attributes? In answering thisquestion, it is argued that entitiescan share different observable propertiesdepending on the level of abstraction adopted,and that it is still possible to speak of moral value even at thehighest level of ontological abstractionrepresented by the informational analysis. Section 4 develops aminimalist axiology based on the concept ofinformation object. It further supports IE's position byaddressing five objections that may undermineits acceptability.     

IT artifact bias: How exogenous predilections influence organizational information system paradigms

Efforts in IS research have long sought to bridge the gap between the information technology (IT) function and strategic business interests. People perceive affordances (possibilities for action) in information technology artifacts differently as cognitive structures (schema) which bias individual focus. This study explores how an individual's tendency to perceive the ‘trees’ in an IT ‘forest’ (artifact preference) affects their assessment of efforts to achieve more effective IT outcomes. The effect is demonstrated using a relatively simple IT success model. Further, in a sample of 120 survey responses supported by ten semi-structured interviews, we demonstrate that job role and organizational IT complexity systematically impact artifact perception. A better understanding of IT artifact bias promises to help organizations better assess information systems.     

Artificial evil and the foundation of computer ethics

Moral reasoning traditionally distinguishes two types of evil:moral (ME) and natural (NE). The standard view is that ME is theproduct of human agency and so includes phenomena such as war,torture and psychological cruelty; that NE is the product ofnonhuman agency, and so includes natural disasters such asearthquakes, floods, disease and famine; and finally, that morecomplex cases are appropriately analysed as a combination of MEand NE. Recently, as a result of developments in autonomousagents in cyberspace, a new class of interesting and importantexamples of hybrid evil has come to light. In this paper, it iscalled artificial evil (AE) and a case is made for considering itto complement ME and NE to produce a more adequate taxonomy. Byisolating the features that have led to the appearance of AE,cyberspace is characterised as a self-contained environment thatforms the essential component in any foundation of the emergingfield of Computer Ethics (CE). It is argued that this goes someway towards providing a methodological explanation of whycyberspace is central to so many of CE's concerns; and it isshown how notions of good and evil can be formulated incyberspace. Of considerable interest is how the propensity for anagent's action to be morally good or evil can be determined evenin the absence of biologically sentient participants and thusallows artificial agents not only to perpetrate evil (and forthat matter good) but conversely to `receive' or `suffer from'it. The thesis defended is that the notion of entropy structure,which encapsulates human value judgement concerning cyberspace ina formal mathematical definition, is sufficient to achieve thispurpose and, moreover, that the concept of AE can be determinedformally, by mathematical methods. A consequence of this approachis that the debate on whether CE should be considered unique, andhence developed as a Macroethics, may be viewed, constructively,in an alternative manner. The case is made that whilst CE issuesare not uncontroversially unique, they are sufficiently novel torender inadequate the approach of standard Macroethics such asUtilitarianism and Deontologism and hence to prompt the searchfor a robust ethical theory that can deal with them successfully.The name Information Ethics (IE) is proposed for that theory. Itis argued that the uniqueness of IE is justified by its beingnon-biologically biased and patient-oriented: IE is anEnvironmental Macroethics based on the concept of data entityrather than life. It follows that the novelty of CE issues suchas AE can be appreciated properly because IE provides a newperspective (though not vice versa). In light of the discussionprovided in this paper, it is concluded that Computer Ethics isworthy of independent study because it requires its ownapplication-specific knowledge and is capable of supporting amethodological foundation, Information Ethics.     

Against the moral Turing test: accountable design and the moral reasoning of autonomous systems

This paper argues against the moral Turing test (MTT) as a framework for evaluating the moral performance of autonomous systems. Though the term has been carefully introduced, considered, and cautioned about in previous discussions (Allen et al. in J Exp Theor Artif Intell 12(3):251–261, 2000; Allen and Wallach 2009), it has lingered on as a touchstone for developing computational approaches to moral reasoning (Gerdes and Øhrstrøm in J Inf Commun Ethics Soc 13(2):98–109, 2015). While these efforts have not led to the detailed development of an MTT, they nonetheless retain the idea to discuss what kinds of action and reasoning should be demanded of autonomous systems. We explore the flawed basis of an MTT in imitation, even one based on scenarios of morally accountable actions. MTT-based evaluations are vulnerable to deception, inadequate reasoning, and inferior moral performance vis a vis a system’s capabilities. We propose verification—which demands the design of transparent, accountable processes of reasoning that reliably prefigure the performance of autonomous systems—serves as a superior framework for both designer and system alike. As autonomous social robots in particular take on an increasing range of critical roles within society, we conclude that verification offers an essential, albeit challenging, moral measure of their design and performance.     

Ethics and Artificial life: From Modeling to Moral Agents

Artificial Life (ALife) has two goals. One attempts to describe fundamental qualities of living systems through agent based computer models. And the second studies whether or not we can artificially create living things in computational mediums that can be realized either, virtually in software, or through biotechnology. The study of ALife has recently branched into two further subdivisions, one is “dry” ALife, which is the study of living systems “in silico” through the use of computer simulations, and the other is “wet” ALife that uses biological material to realize what has only been simulated on computers, effectively wet ALife uses biological material as a kind of computer. This is challenging to the field of computer ethics as it points towards a future in which computer and bioethics might have shared concerns. The emerging studies into wet ALife are likely to provide strong empirical evidence for ALife’s most challenging hypothesis: that life is a certain set of computable functions that can be duplicated in any medium. I believe this will propel ALife into the midst of the mother of all cultural battles that has been gathering around the emergence of biotechnology. Philosophers need to pay close attention to this debate and can serve a vital role in clarifying and resolving the dispute. But even if ALife is merely a computer modeling technique that sheds light on living systems, it still has a number of significant ethical implications such as its use in the modeling of moral and ethical systems, as well as in the creation of artificial moral agents.     

试论孝道传承与高校德育教育相融合的必要性--以常州机电职业技术学院的实践活动为例

“孝道”是中华民族文化和中国传统道德的一个基本的、重要的内容,是人们道德行为的生长点。文章论述了将孝道传承与高校德育教育相融合的必要性及重大意义,分析了二者融合的路径和策略,以期为提高大学生思想素质、提高高校德育教育的实效性提供借鉴。     

Memory objects in project environments: Storing, retrieving and adapting learning in project-based firms

This paper investigates the role of objects holding representations of knowledge in the transfer of learning across projects. On the basis of an in-depth case study, this paper shows that the way in which relatively simple artifacts, such as Excel workbooks, represent knowledge enables them to act as boundary objects across occupations and as memory devices across projects. It is the temporal capacity of these boundary objects that makes them points of juncture in a widely distributed memory system, enabling project-based firms to balance preservation and adaptation of knowledge. The mechanisms for the preservation of learning are not missing from project environments, rather they are less visible and less direct than in other settings, and therefore less docile in the face of managerial action.     

德育教育需细致入微——如何做一名贴近学生心灵的德育教师

在现代教育中,德育教育一直被置于首位.现今教育的对象是那些在生活水平提高、科技高度发展的大环境下生活的,自我意识颇为强烈的学生,这也就对德育教师提出了更高的要求.要加强德育工作,细致入微地引导学生自觉接受德育教育,使得教师在获得德育教育成效的同时能够贴近学生的心灵.     

Theft of virtual items in online multiplayer computer games: an ontological and moral analysis

In 2009 Dutch judges convicted several minors for theft of virtual items in the virtual worlds of online multiplayer computer games. From a legal point of view these convictions gave rise to the question whether virtual items should count as “objects” that can be “stolen” under criminal law. This legal question has both an ontological and a moral component. The question whether or not virtual items count as “objects” that can be “stolen” is an ontological question. The question whether or not they should count as such under criminal law is of a moral nature. The purpose of this paper is to answer both the ontological question and the moral question underlying the legal question.     

论高校德育视域下科尔伯格德育理论的现实意义

认知发展道德教育理论是20世纪50年代以来西方最有影响力的一种教育理论派别。本文立足于我国高校德育现状,在概述该理论的主要内容的基础上,分析该理论对我国高校德育的启示。     

LINC: biology's revolutionary little computer

The 1963 LINC (Laboratory INstrument Computer) stands at the center of two stories: the computerization of the biologist's laboratory and the advent of small-scale computing. The brainchild of Wesley Clark, 'the most brilliant computer designer of his generation', LINC was developed specifically to address the failure of biologists to adopt computer technology. To meet their unique needs, Clark built a machine the radical design of which defied and subverted the then dominant conventions of computer architecture.