Research into the ethics of artificial intelligence is often categorized into two subareas—robot ethics and machine ethics. Many of the definitions and classifications of the subject matter of these subfields, as found in the literature, are conflated, which I seek to rectify. In this essay, I infer that using the term ‘machine ethics’ is too broad and glosses over issues that the term computational ethics best describes. I show that the subject of inquiry of computational ethics is of great value (...) and indeed is an important frontier in developing ethical artificial intelligence systems. I also show that computational is a distinct, often neglected field in the ethics of AI. In contrast to much of the literature, I argue that the appellation ‘machine ethics’ does not sufficiently capture the entire project of embedding ethics into AI/S, and hence the need for computational ethics. This essay is unique for two reasons; first, it offers a philosophical analysis of the subject of computational ethics that is not found in the literature. Second, it offers a finely grained analysis that shows the thematic distinction among robot ethics, machine ethics and computational ethics. (shrink)

Cognitive science has converged in many ways with cognitive psychology, but while also maintaining a distinctive interdisciplinary nature. Here we further characterize this existing state of the field before proposing how it might be reconceptualized toward a broader and more distinct, and thus more stable, position in the realm of sciences.

Topics in Cognitive Science, Volume 14, Issue 3, Page 451-466, July 2022.

How can we study bounded rationality? We answer this question by proposing rational task analysis —a systematic approach that prevents experimental researchers from drawing premature conclusions regarding the rationality of agents. RTA is a methodology and perspective that is anchored in the notion of bounded rationality and aids in the unbiased interpretation of results and the design of more conclusive experimental paradigms. RTA focuses on concrete tasks as the primary interface between agents and environments and requires explicating essential task elements, (...) specifying rational norms, and bracketing the range of possible performance, before contrasting various benchmarks with actual performance. After describing RTA’s core components we illustrate its use in three case studies that examine human memory updating, multitasking behavior, and melioration. We discuss RTA’s characteristic elements and limitations by comparing it to related approaches. We conclude that RTA provides a useful tool to render the study of bounded rationality more transparent and less prone to theoretical confusion. (shrink)

All forms of cognition, whether natural or artificial, are subject to constraints of their computing architecture. This assumption forms the tenet of virtually all general theories of cognition, including those deriving from bounded optimality and bounded rationality. In this letter, we highlight an unresolved puzzle related to this premise: what are these constraints, and why are cognitive architectures subject to cognitive constraints in the first place? First, we lay out some pieces along the puzzle edge, such as computational tradeoffs inherent (...) to neural architectures that give rise to rational bounds of cognition. We then outline critical next steps for characterizing cognitive bounds, proposing that some of these bounds can be subject to modification by cognition and, as such, are part of what is being optimized when cognitive agents decide how to allocate cognitive resources. We conclude that these emerging views may contribute to a more holistic perspective on the nature of cognitive bounds, as well as their alteration subject to cognition. (shrink)

In this article, a concise argument for computational rationality as a basis for artificial moral agency is advanced. Some ethicists have long argued that rational agents can become artificial moral agents. However, most of their views have come from purely philosophical perspectives, thus making it difficult to transfer their arguments to a scientific and analytical frame of reference. The result has been a disintegrated approach to the conceptualisation and design of artificial moral agents. In this article, I make the argument (...) for computational rationality as an integrative element that effectively combines the philosophical and computational aspects of artificial moral agency. This logically leads to a philosophically coherent and scientifically consistent model for building artificial moral agents. Besides providing a possible answer to the question of how to build artificial moral agents, this model also invites sound debate from multiple disciplines, which should help to advance the field of machine ethics forward. (shrink)

This paper addresses a common challenge with computational cognitive models: identifying parameter values that are both theoretically plausible and generate predictions that match well with empirical data. While computational models can offer deep explanations of cognition, they are computationally complex and often out of reach of traditional parameter fitting methods. Weak methodology may lead to premature rejection of valid models or to acceptance of models that might otherwise be falsified. Mathematically robust fitting methods are, therefore, essential to the progress of (...) computational modeling in cognitive science. In this article, we investigate the capability and role of modern fitting methods—including Bayesian optimization and approximate Bayesian computation—and contrast them to some more commonly used methods: grid search and Nelder–Mead optimization. Our investigation consists of a reanalysis of the fitting of two previous computational models: an Adaptive Control of Thought—Rational model of skill acquisition and a computational rationality model of visual search. The results contrast the efficiency and informativeness of the methods. A key advantage of the Bayesian methods is the ability to estimate the uncertainty of fitted parameter values. We conclude that approximate Bayesian computation is (a) efficient, (b) informative, and (c) offers a path to reproducible results. (shrink)

While Bayesian models have been applied to an impressive range of cognitive phenomena, methodological challenges have been leveled concerning their role in the program of rational analysis. The focus of the current article is on computational impediments to probabilistic inference and related puzzles about empirical confirmation of these models. The proposal is to rethink the role of Bayesian methods in rational analysis, to adopt an independently motivated notion of rationality appropriate for computationally bounded agents, and to explore broad conditions under (...) which Bayesian agents would be rational. The proposal is illustrated with a characterization of costs inspired by thermodynamics. (shrink)

It is known that, on average, people adapt their choice of memory strategy to the subjective utility of interaction. What is not known is whether an individual's choices are boundedly optimal. Two experiments are reported that test the hypothesis that an individual's decisions about the distribution of remembering between internal and external resources are boundedly optimal where optimality is defined relative to experience, cognitive constraints, and reward. The theory makes predictions that are tested against data, not fitted to it. The (...) experiments use a no-choice/choice utility learning paradigm where the no-choice phase is used to elicit a profile of each participant's performance across the strategy space and the choice phase is used to test predicted choices within this space. They show that the majority of individuals select strategies that are boundedly optimal. Further, individual differences in what people choose to do are successfully predicted by the analysis. Two issues are discussed: the performance of the minority of participants who did not find boundedly optimal adaptations, and the possibility that individuals anticipate what, with practice, will become a bounded optimal strategy, rather than what is boundedly optimal during training. (shrink)

Marr's levels of analysis—computational, algorithmic, and implementation—have served cognitive science well over the last 30 years. But the recent increase in the popularity of the computational level raises a new challenge: How do we begin to relate models at different levels of analysis? We propose that it is possible to define levels of analysis that lie between the computational and the algorithmic, providing a way to build a bridge between computational- and algorithmic-level models. The key idea is to push the (...) notion of rationality, often used in defining computational-level models, deeper toward the algorithmic level. We offer a simple recipe for reverse-engineering the mind's cognitive strategies by deriving optimal algorithms for a series of increasingly more realistic abstract computational architectures, which we call “resource-rational analysis.”. (shrink)

I present a computational-level model of semantic interference effects in online word production within a rate–distortion framework. I consider a bounded-rational agent trying to produce words. The agent's action policy is determined by maximizing accuracy in production subject to computational constraints. These computational constraints are formalized using mutual information. I show that semantic similarity-based interference among words falls out naturally from this setup, and I present a series of simulations showing that the model captures some of the key empirical patterns (...) observed in Stroop and Picture–Word Interference paradigms, including comparisons to human data from previous experiments. (shrink)

We test people's ability to optimize performance across two concurrent tasks. Participants performed a number entry task while controlling a randomly moving cursor with a joystick. Participants received explicit feedback on their performance on these tasks in the form of a single combined score. This payoff function was varied between conditions to change the value of one task relative to the other. We found that participants adapted their strategy for interleaving the two tasks, by varying how long they spent on (...) one task before switching to the other, in order to achieve the near maximum payoff available in each condition. In a second experiment, we show that this behavior is learned quickly and remained stable for as long as the payoff function did not change. The results of this work show that people are adaptive and flexible in how they prioritize and allocate attention in a dual-task setting. However, it also demonstrates some of the limits regarding people's ability to optimize payoff functions. (shrink)

The psychology of reasoning uses norms to categorize responses to reasoning tasks as correct or incorrect in order to interpret the responses and compare them across reasoning tasks. This raises the arbitration problem: any number of norms can be used to evaluate the responses to any reasoning task and there doesn’t seem to be a principled way to arbitrate among them. Elqayam and Evans have argued that this problem is insoluble, so they call for the psychology of reasoning to dispense (...) with norms entirely. Alternatively, Stupple and Ball have argued that norms must be used, but the arbitration problem should be solved by favouring norms that are sensitive to the context, constraints, and goals of human reasoning. In this paper, I argue that the design of reasoning tasks requires the selection of norms that are indifferent to the factors that influence human responses to the tasks—which aren’t knowable during the task design phase, before the task has been given to human subjects. Moreover, I argue that the arbitration problem is easily dissolved: any well-designed task will contain instructions that implicitly or explicitly specify a single determinate norm, which specifies what would count as a solution to the task—independently of the context, constraints, and goals of human reasoning. Finally, I argue that discouraging the use of these a priori task norms may impair the design of novel reasoning tasks. (shrink)

We consider approaches to explanation within the cognitive sciences that begin with Marr's computational level or Marr's implementational level and argue that each is subject to fundamental limitations which impair their ability to provide adequate explanations of cognitive phenomena. For this reason, it is argued, explanation cannot proceed at either level without tight coupling to the algorithmic and representation level. Even at this level, however, we argue that additional constraints relating to the decomposition of the cognitive system into a set (...) of interacting subfunctions are required. Integrated cognitive architectures that permit abstract specification of the functions of components and that make contact with the neural level provide a powerful bridge for linking the algorithmic and representational level to both the computational level and the implementational level. (shrink)

A widespread need to explain the behavior and outcomes of AI-based systems has emerged, due to their ubiquitous presence. Thus, providing renewed momentum to the relatively new research area of eXplainable AI (XAI). Nowadays, the importance of XAI lies in the fact that the increasing control transference to this kind of system for decision making -or, at least, its use for assisting executive stakeholders- already affects many sensitive realms (as in Politics, Social Sciences, or Law). The decision-making power handover to (...) opaque AI systems makes mandatory explaining those, primarily in application scenarios where the stakeholders are unaware of both the high technology applied and the basic principles governing the technological solutions. The issue should not be reduced to a merely technical problem; the explainer would be compelled to transmit richer knowledge about the system (including its role within the informational ecosystem where he/she works). To achieve such an aim, the explainer could exploit, if necessary, practices from other scientific and humanistic areas. The first aim of the paper is to emphasize and justify the need for a multidisciplinary approach that is beneficiated from part of the scientific and philosophical corpus on Explaining, underscoring the particular nuances of the issue within the field of Data Science. The second objective is to develop some arguments justifying the authors’ bet by a more relevant role of ideas inspired by, on the one hand, formal techniques from Knowledge Representation and Reasoning, and on the other hand, the modeling of human reasoning when facing the explanation. This way, explaining modeling practices would seek a sound balance between the pure technical justification and the explainer-explainee agreement. (shrink)

Theories of rational decision making often abstract away from computational and other resource limitations faced by real agents. An alternative approach known as resource rationality puts such matters front and center, grounding choice and decision in the rational use of finite resources. Anticipated by earlier work in economics and in computer science, this approach has recently seen rapid development and application in the cognitive sciences. Here, the theory of rationality plays a dual role, both as a framework for normative assessment (...) and as a source of scientific hypotheses about how mental processes in fact work. The latter project, often called rational analysis, depends for its success on a fine-grained characterization of the computational problem facing a decision maker, which may in turn depend on realistic assumptions about what the relevant agent is like. As a consequence, resource rationality involves a delicate, but often fruitful interplay between the normative and the descriptive. (shrink)

Resource rationality may explain suboptimal patterns of reasoning; but what of “anti-Bayesian” effects where the mind updates in a direction opposite the one it should? We present two phenomena — belief polarization and the size-weight illusion — that are not obviously explained by performance- or resource-based constraints, nor by the authors’ brief discussion of reference repulsion. Can resource rationality accommodate them?

This paper explores how the Leviathan that projects power through nuclear arms exercises a unique nuclearized sovereignty. In the case of nuclear superpowers, this sovereignty extends to wielding the power to destroy human civilization as we know it across the globe. Nuclearized sovereignty depends on a hybrid form of power encompassing human decision-makers in a hierarchical chain of command, and all of the technical and computerized functions necessary to maintain command and control at every moment of the sovereign's existence: this (...) sovereign power cannot sleep. This article analyzes how the form of rationality that informs this hybrid exercise of power historically developed to be computable. By definition, computable rationality must be able to function without any intelligible grasp of the context or the comprehensive significance of decision-making outcomes. Thus, maintaining nuclearized sovereignty necessarily must be able to execute momentous life and death decisions without the type of sentience we usually associate with ethical individual and collective decisions. (shrink)