Artificial Intelligence: A Modern Approach, 3e offers the most comprehensive, up-to-date introduction to the theory and practice of artificial intelligence. Number one in its field, this textbook is ideal for one or two-semester, undergraduate or graduate-level courses in Artificial Intelligence. Dr. Peter Norvig, contributing Artificial Intelligence author and Professor Sebastian Thrun, a Pearson author are offering a free online course at Stanford University on artificial intelligence. According to an article in The New York Times, the course on artificial intelligence is (...) “one of three being offered experimentally by the Stanford computer science department to extend technology knowledge and skills beyond this elite campus to the entire world.” One of the other two courses, an introduction to database software, is being taught by Pearson author Dr. Jennifer Widom. Artificial Intelligence: A Modern Approach, 3e is available to purchase as an eText for your Kindle™, NOOK™, and the iPhone®/iPad®. (shrink)

In this essay we provide (1) an argument for why ethics should be naturalized, (2) an analysis of why it is not yet naturalized, (3) a defense of ethical naturalism against two fallacies—Hume’s and Moore’s—that ethical naturalism allegedly commits, and (4) a proposal that normative ethics is best conceived as part of human ecology committed to pluralistic relativism. We explain why naturalizing ethics both entails relativism and also constrains it, and why nihilism about value is not an especially worrisome for (...) ethical naturalists. The substantive view we put forth constitutes the essence of Duke Naturalism. (NOTE: This is a slightly modified reprint of Flangan et al 2007 of the same title.). (shrink)

In this essay we provide (1) an argument for why ethics should be naturalized, (2) an analysis of why it is not yet naturalized, (3) a defense of ethical naturalism against two fallacies—Hume’s and Moore’s—that ethical naturalism allegedly commits, and (4) a proposal that normative ethics is best conceived as part of human ecology committed to pluralistic relativism. We explain why naturalizing ethics both entails relativism and also constrains it, and why nihilism about value is not an especially worrisome for (...) ethical naturalists. The substantive view we put forth constitutes the essence of Duke Naturalism. (shrink)

Many authors have proposed constraining the behaviour of intelligent systems with ‘machine ethics’ to ensure positive social outcomes from the development of such systems. This paper critically analyses the prospects for machine ethics, identifying several inherent limitations. While machine ethics may increase the probability of ethical behaviour in some situations, it cannot guarantee it due to the nature of ethics, the computational limitations of computational agents and the complexity of the world. In addition, machine ethics, even if it were to (...) be ‘solved’ at a technical level, would be insufficient to ensure positive social outcomes from intelligent systems. (shrink)

Trolley cases are widely considered central to the ethics of autonomous vehicles. We caution against this by identifying four problems. Trolley cases, given technical limitations, rest on assumptions that are in tension with one another. Furthermore, trolley cases illuminate only a limited range of ethical issues insofar as they cohere with a certain design framework. Furthermore, trolley cases seem to demand a moral answer when a political answer is called for. Finally, trolley cases might be epistemically problematic in several ways. (...) To put forward a positive proposal, we illustrate how ethical challenges arise from mundane driving situations. We argue that mundane situations are relevant because of the specificity they require and the scale they exhibit. We then illustrate some of the ethical challenges arising from optimizing for safety, balancing safety with other values such as mobility, and adjusting to incentives of legal frameworks. (shrink)

The burgeoning science of ethics has produced a trend toward pessimism. Ordinary moral thought and action, we’re told, are profoundly influenced by arbitrary factors and ultimately driven by unreasoned feelings. This book counters the current orthodoxy on its own terms by carefully engaging with the empirical literature. The resulting view, optimistic rationalism, shows the pervasive role played by reason, and ultimately defuses sweeping debunking arguments in ethics. The science does suggest that moral knowledge and virtue don’t come easily. However, despite (...) the heavy influence of automatic and unconscious processes that have been shaped by evolutionary pressures, we needn’t reject ordinary moral psychology as fundamentally flawed or in need of serious repair. Reason can be corrupted in ethics just as in other domains, but a special pessimism about morality in particular is unwarranted. Moral judgment and motivation are fundamentally rational enterprises not beholden to the passions. (shrink)

We present a new “reason-based” approach to the formal representation of moral theories, drawing on recent decision-theoretic work. We show that any moral theory within a very large class can be represented in terms of two parameters: a specification of which properties of the objects of moral choice matter in any given context, and a specification of how these properties matter. Reason-based representations provide a very general taxonomy of moral theories, as differences among theories can be attributed to differences in (...) their two key parameters. We can thus formalize several distinctions, such as between consequentialist and non-consequentialist theories, between universalist and relativist theories, between agent-neutral and agent-relative theories, between monistic and pluralistic theories, between atomistic and holistic theories, and between theories with a teleological structure and those without. Reason-based representations also shed light on an important but under-appreciated phenomenon: the “underdetermination of moral theory by deontic content”. (shrink)

Computers are already approving financial transactions, controlling electrical supplies, and driving trains. Soon, service robots will be taking care of the elderly in their homes, and military robots will have their own targeting and firing protocols. Colin Allen and Wendell Wallach argue that as robots take on more and more responsibility, they must be programmed with moral decision-making abilities, for our own safety. Taking a fast paced tour through the latest thinking about philosophical ethics and artificial intelligence, the authors argue (...) that even if full moral agency for machines is a long way off, it is already necessary to start building a kind of functional morality, in which artificial moral agents have some basic ethical sensitivity. But the standard ethical theories don't seem adequate, and more socially engaged and engaging robots will be needed. As the authors show, the quest to build machines that are capable of telling right from wrong has begun. -/- Moral Machines is the first book to examine the challenge of building artificial moral agents, probing deeply into the nature of human decision making and ethics. (shrink)

Self-driving cars hold out the promise of being safer than manually driven cars. Yet they cannot be a 100 % safe. Collisions are sometimes unavoidable. So self-driving cars need to be programmed for how they should respond to scenarios where collisions are highly likely or unavoidable. The accident-scenarios self-driving cars might face have recently been likened to the key examples and dilemmas associated with the trolley problem. In this article, we critically examine this tempting analogy. We identify three important ways (...) in which the ethics of accident-algorithms for self-driving cars and the philosophy of the trolley problem differ from each other. These concern: the basic decision-making situation faced by those who decide how self-driving cars should be programmed to deal with accidents; moral and legal responsibility; and decision-making in the face of risks and uncertainty. In discussing these three areas of disanalogy, we isolate and identify a number of basic issues and complexities that arise within the ethics of the programming of self-driving cars. (shrink)

This book is a provocative contribution to contemporary ethical theory challenging foundational conceptions of character that date back to Aristotle. John Doris draws on behavioral science, especially social psychology, to argue that we misattribute the causes of behavior to personality traits and other fixed aspects of character rather than to the situational context. More often than not it is the situation not the nature of the personality that really counts. The author elaborates the philosophical consequences of this research for a (...) whole array of ethical theories and shows that, once rid of the misleading conception of motivation, moral psychology can support more robust ethical theories and more humane ethical practices. (shrink)

The claim of the multiple realizability of mental states by brain states has been a major feature of the dominant philosophy of mind of the late 20th century. The claim is usually motivated by evidence that mental states are multiply realized, both within humans and between humans and other species. We challenge this contention by focusing on how neuroscientists differentiate brain areas. The fact that they rely centrally on psychological measures in mapping the brain and do so in a comparative (...) fashion undercuts the likelihood that, at least within organic life forms, we are likely to find cases of multiply realized psychological functions. (shrink)

Behavioral scientists routinely publish broad claims about human psychology and behavior in the world's top journals based on samples drawn entirely from Western, Educated, Industrialized, Rich, and Democratic (WEIRD) societies. Researchers – often implicitly – assume that either there is little variation across human populations, or that these “standard subjects” are as representative of the species as any other population. Are these assumptions justified? Here, our review of the comparative database from across the behavioral sciences suggests both that there is (...) substantial variability in experimental results across populations and that WEIRD subjects are particularly unusual compared with the rest of the species – frequent outliers. The domains reviewed include visual perception, fairness, cooperation, spatial reasoning, categorization and inferential induction, moral reasoning, reasoning styles, self-concepts and related motivations, and the heritability of IQ. The findings suggest that members of WEIRD societies, including young children, are among the least representative populations one could find for generalizing about humans. Many of these findings involve domains that are associated with fundamental aspects of psychology, motivation, and behavior – hence, there are no obviousa priorigrounds for claiming that a particular behavioral phenomenon is universal based on sampling from a single subpopulation. Overall, these empirical patterns suggests that we need to be less cavalier in addressing questions ofhumannature on the basis of data drawn from this particularly thin, and rather unusual, slice of humanity. We close by proposing ways to structurally re-organize the behavioral sciences to best tackle these challenges. (shrink)

Recently, there has been a resurgence of interest in general, comprehensive models of human cognition. Such models aim to explain higher-order cognitive faculties, such as deliberation and planning. Given a computational representation, the validity of these models can be tested in computer simulations such as software agents or embodied robots. The push to implement computational models of this kind has created the field of artificial general intelligence (AGI). Moral decision making is arguably one of the most challenging tasks for computational (...) approaches to higher-order cognition. The need for increasingly autonomous artificial agents to factor moral considerations into their choices and actions has given rise to another new field of inquiry variously known as Machine Morality, Machine Ethics, Roboethics, or Friendly AI. In this study, we discuss how LIDA, an AGI model of human cognition, can be adapted to model both affective and rational features of moral decision making. Using the LIDA model, we will demonstrate how moral decisions can be made in many domains using the same mechanisms that enable general decision making. Comprehensive models of human cognition typically aim for compatibility with recent research in the cognitive and neural sciences. Global workspace theory, proposed by the neuropsychologist Bernard Baars (1988), is a highly regarded model of human cognition that is currently being computationally instantiated in several software implementations. LIDA (Franklin, Baars, Ramamurthy, & Ventura, 2005) is one such computational implementation. LIDA is both a set of computational tools and an underlying model of human cognition, which provides mechanisms that are capable of explaining how an agent’s selection of its next action arises from bottom-up collection of sensory data and top-down processes for making sense of its current situation. We will describe how the LIDA model helps integrate emotions into the human decision-making process, and we will elucidate a process whereby an agent can work through an ethical problem to reach a solution that takes account of ethically relevant factors. (shrink)

In Intricate Ethics, Kamm questions the moral importance of some non-consequentialist distinctions and then introduces and argues for the moral importance of ...

Woodward's long awaited book is an attempt to construct a comprehensive account of causation explanation that applies to a wide variety of causal and explanatory claims in different areas of science and everyday life. The book engages some of the relevant literature from other disciplines, as Woodward weaves together examples, counterexamples, criticisms, defenses, objections, and replies into a convincing defense of the core of his theory, which is that we can analyze causation by appeal to the notion of manipulation.

Simulation techniques, especially those implemented on a computer, are frequently employed in natural as well as in social sciences with considerable success. There is mounting evidence that the "model-building era" (J. Niehans) that dominated the theoretical activities of the sciences for a long time is about to be succeeded or at least lastingly supplemented by the "simulation era". But what exactly are models? What is a simulation and what is the difference and the relation between a model and a simulation? (...) These are some of the questions addressed in this article. I maintain that the most significant feature of a simulation is that it allows scientists to imitate one process by another process. "Process" here refers solely to a temporal sequence of states of a system. Given the observation that processes are dealt with by all sorts of scientists, it is apparent that simulations prove to be a powerful interdisciplinarily acknowledged tool. Accordingly, simulations are best suited to investigate the various research strategies in different sciences more carefully. To this end, I focus on the function of simulations in the research process. Finally, a somewhat detailed case-study from nuclear physics is presented which, in my view, illustrates elements of a typical simulation in physics. (shrink)

Truth is standardly considered a requirement on epistemic acceptability. But science and philosophy deploy models, idealizations and thought experiments that prescind from truth to achieve other cognitive ends. I argue that such felicitous falsehoods function as cognitively useful fictions. They are cognitively useful because they exemplify and afford epistemic access to features they share with the relevant facts. They are falsehoods in that they diverge from the facts. Nonetheless, they are true enough to serve their epistemic purposes. Theories that contain (...) them have testable consequences, hence are factually defeasible. (shrink)

One of the objectives in the field of artificial intelligence for some decades has been the development of artificial agents capable of coexisting in harmony with people and other systems. The computing research community has made efforts to design artificial agents capable of doing tasks the way people do, tasks requiring cognitive mechanisms such as planning, decision-making, and learning. The application domains of such software agents are evident nowadays. Humans are experiencing the inclusion of artificial agents in their environment as (...) unmanned vehicles, intelligent houses, and humanoid robots capable of caring for people. In this context, research in the field of machine ethics has become more than a hot topic. Machine ethics focuses on developing ethical mechanisms for artificial agents to be capable of engaging in moral behavior. However, there are still crucial challenges in the development of truly Artificial Moral Agents. This paper aims to show the current status of Artificial Moral Agents by analyzing models proposed over the past two decades. As a result of this review, a taxonomy to classify Artificial Moral Agents according to the strategies and criteria used to deal with ethical problems is proposed. The presented review aims to illustrate (1) the complexity of designing and developing ethical mechanisms for this type of agent, and (2) that there is a long way to go (from a technological perspective) before this type of artificial agent can replace human judgment in difficult, surprising or ambiguous moral situations. (shrink)

Intimate relationships between robots and human beings may begin to form in the near future. Market forces, customer demand, and other factors may drive the creation of various forms of robots to which humans may form strong emotional attachments. Yet prior to the technology becoming fully actualized, numerous ethical, legal, and social issues must be addressed. This could be accomplished in part by establishing a rigorous scientific research agenda in the realm of intimate robotics, the aim of which would be (...) to explore what effects the technology may have on users and on society more generally. Our goal is not to resolve whether the development of intimate robots is ethically appropriate. Rather, we contend that if such robots are going to be designed, then an obligation emerges to prevent harm that the technology could cause. (shrink)

The recent progress in the development of autonomous cars has seen ethical questions come to the forefront. In particular, life and death decisions regarding the behavior of self-driving cars in trolley dilemma situations are attracting widespread interest in the recent debate. In this essay we want to ask whether we should implement a mandatory ethics setting for the whole of society or, whether every driver should have the choice to select his own personal ethics setting. While the consensus view seems (...) to be that people would not be willing to use an automated car that might sacrifice themselves in a dilemma situation, we will defend the somewhat contra-intuitive claim that this would be nevertheless in their best interest. The reason is, simply put, that a PES regime would most likely result in a prisoner’s dilemma. (shrink)

Robot ethics encompasses ethical questions about how humans should design, deploy, and treat robots; machine morality encompasses questions about what moral capacities a robot should have and how these capacities could be computationally implemented. Publications on both of these topics have doubled twice in the past 10 years but have often remained separate from one another. In an attempt to better integrate the two, I offer a framework for what a morally competent robot would look like and discuss a number (...) of ethical questions about the design, use, and treatment of such moral robots in society. Instead of searching for a fixed set of criteria of a robot’s moral competence I identify the multiple elements that make up human moral competence and probe the possibility of designing robots that have one or more of these human elements, which include: moral vocabulary; a system of norms; moral cognition and affect; moral decision making and action; moral communication. Juxtaposing empirical research, philosophical debates, and computational challenges, this article adopts an optimistic perspective: if robotic design truly commits to building morally competent robots, then those robots could be trustworthy and productive partners, caretakers, educators, and members of the human community. Moral competence does not resolve all ethical concerns over robots in society, but it may be a prerequisite to resolve at least some of them. (shrink)

Science relies on models and idealizations that are known not to be true. Even so, science is epistemically reputable. To accommodate science, epistemology should focus on understanding rather than knowledge and should recognize that the understanding of a topic need not be factive. This requires reconfiguring the norms of epistemic acceptability. If epistemology has the resources to accommodate science, it will also have the resources to show that art too advances understanding.

We are moral apes, a difference between humans and our relatives that has received significant recent attention in the evolutionary literature. Evolutionary accounts of morality have often been recruited in support of error theory: moral language is truth-apt, but substantive moral claims are never true. In this article, we: locate evolutionary error theory within the broader framework of the relationship between folk conceptions of a domain and our best scientific conception of that same domain; within that broader framework, argue that (...) error theory and vindication are two ends of a continuum, and that in the light of our best science, many folk conceptual structures are neither hopelessly wrong nor fully vindicated; and argue that while there is no full vindication of morality, no seamless reduction of normative facts to natural facts, nevertheless one important strand in the evolutionary history of moral thinking does support reductive naturalism—moral facts are facts about cooperation, and the conditions and practices that support or undermine it. In making our case for, we first respond to the important error theoretic argument that the appeal to moral facts is explanatorily redundant, and second, we make a positive case that true moral beliefs are a ‘fuel for success’, a map by which we steer, flexibly, in a variety of social interactions. The vindication, we stress, is at most partial: moral cognition is a complex mosaic, with a complex genealogy, and selection for truth-tracking is only one thread in that genealogy. (shrink)

In Intricate Ethics, Kamm questions the moral importance of some non-consequentialist distinctions and then introduces and argues for the moral importance of other distinctions. The first section discusses nonconsequentialist ethical theory and the trolley problem; the second deals with the notions of moral status and rights; the third takes up the issues of responsibility and complicity and the possible moral significance of distance; and the fourth section analyzes the views of others in the non-consequentialist and consequentialist camps.

Humans are unique in the animal world in the extent to which their day-to-day behavior is governed by a complex set of rules and principles commonly called norms. Norms delimit the bounds of proper behavior in a host of domains, providing an invisible web of normative structure embracing virtually all aspects of social life. People also find many norms to be deeply meaningful. Norms give rise to powerful subjective feelings that, in the view of many, are an important part of (...) what it is to be a human agent. Despite the vital role of norms in human lives and human behavior, and the central role they play in explanations in the social sciences, there has been very little systematic attention devoted to norms in cognitive science. Much existing research is partial and piecemeal, making it difficult to know how individual findings cohere into a comprehensive picture. Our goal in this essay is to offer an account of the psychological mechanisms and processes underlying norms that integrates what is known and can serve as a framework for future research. (shrink)

In this chapter, we begin with a discussion of motivation itself, and use that discussion to sketch four possible theories of distinctively moral motivation: caricature versions of familiar instrumentalist, cognitivist, sentimentalist, and personalist theories about morally worthy motivation. To test these theories, we turn to a wealth of scientific, particularly neuroscientific, evidence. Our conclusions are that (1) although the scientific evidence does not at present mandate a unique philosophical conclusion, it does present formidable obstacles to a number of popular philosophical (...) approaches, and (2) theories of morally worthy motivation that best fit the current scientific picture are ones that owe much more to Hume or Aristotle than to Kant. (shrink)

The traditional philosophical doctrine of double effect claims that agents’ intentions affect whether acts are morally wrong. Our behavioral study reveals that agents’ intentions do affect whether acts are judged morally wrong, whereas the temporal order of good and bad effects affects whether acts are classified as killings. This finding suggests that the moral judgments are not based on the classifications. Our results also undermine recent claims that prior moral judgments determine whether agents are seen as causing effects intentionally rather (...) than as side effects. (shrink)

This paper explores the ambiguous impact of new information and communications technologies on the cultivation of moral skills in human beings. Just as twentieth century advances in machine automation resulted in the economic devaluation of practical knowledge and skillsets historically cultivated by machinists, artisans, and other highly trained workers , while also driving the cultivation of new skills in a variety of engineering and white collar occupations, ICTs are also recognized as potential causes of a complex pattern of economic deskilling, (...) reskilling, and upskilling. In this paper, I adapt the conceptual apparatus of sociological debates over economic deskilling to illuminate a different potential for technological deskilling/upskilling, namely the ability of ICTs to contribute to the moral deskilling of human users, a potential that exists alongside rich but currently underrealized possibilities for moral reskilling and/or upskilling. I flesh out this general hypothesis by means of examples involving automated weapons technology, new media practices, and social robotics. I conclude that since moral skills are essential prerequisites for the effective development of practical wisdom and virtuous character, and since market and cultural forces are not presently aligned to bring about the more salutary of the ambiguous potentials presented here, the future shape of these developments warrants our close attention—and perhaps active intervention. (shrink)

How should autonomous vehicles be programmed to behave in the event of an unavoidable accident in which the only choice open is one between causing different damages or losses to different objects or persons? This paper addresses this ethical question starting from the normative principles elaborated in the law to regulate difficult choices in other emergency scenarios. In particular, the paper offers a rational reconstruction of some major principles and norms embedded in the Anglo-American jurisprudence and case law on the (...) “doctrine of necessity”; and assesses which, if any, of these principles and norms can be utilized to find reasonable guidelines for solving the ethical issue of the regulation of the programming of autonomous vehicles in emergency situations. The paper covers the following topics: the distinction between “justification” and “excuse”, the legal prohibition of intentional killing outside self-defence, the incommensurability of goods, and the legal constrains to the use of lethal force set by normative positions: obligations, responsibility, rights, and authority. For each of these principles and constrains the possible application to the programming of autonomous vehicles is discussed. Based on the analysis, some practical suggestions are offered. (shrink)

A principal goal of the discipline of artificial morality is to design artificial agents to act as if they are moral agents. Intermediate goals of artificial morality are directed at building into AI systems sensitivity to the values, ethics, and legality of activities. The development of an effective foundation for the field of artificial morality involves exploring the technological and philosophical issues involved in making computers into explicit moral reasoners. The goal of this paper is to discuss strategies for implementing (...) artificial morality and the differing criteria for success that are appropriate to different strategies. (shrink)

In response, I elaborate on my conception of moral reasoning, as well as clarify the structure of debunking arguments and how my cautious optimism is only of the “glass half full” sort. I also explain how rationalism can capture insights purportedly only explained by sentimentalist and Humean views. The reply concludes by clarifying and admitting some limits of the book's scope.

As the capabilities of artificial intelligence systems improve, it becomes important to constrain their actions to ensure their behaviour remains beneficial to humanity. A variety of ethical, legal and safety-based frameworks have been proposed as a basis for designing these constraints. Despite their variations, these frameworks share the common characteristic that decision-making must consider multiple potentially conflicting factors. We demonstrate that these alignment frameworks can be represented as utility functions, but that the widely used Maximum Expected Utility paradigm provides insufficient (...) support for such multiobjective decision-making. We show that a Multiobjective Maximum Expected Utility paradigm based on the combination of vector utilities and non-linear action–selection can overcome many of the issues which limit MEU’s effectiveness in implementing aligned AI. We examine existing approaches to multiobjective AI, and identify how these can contribute to the development of human-aligned intelligent agents. (shrink)