This paper presents the main topics, arguments, and positions in the philosophy of AI at present (excluding ethics). Apart from the basic concepts of intelligence and computation, the main topics of ar-tificial cognition are perception, action, meaning, rational choice, free will, consciousness, and normativity. Through a better understanding of these topics, the philosophy of AI contributes to our understand-ing of the nature, prospects, and value of AI. Furthermore, these topics can be understood more deeply through the discussion of AI; so (...) we suggest that “AI Philosophy” provides a new method for philosophy. (shrink)

Comparatively easy questions we might ask about creativity are distinguished from the hard question of explaining transformative creativity. Many have focused on the easy questions, offering no reason to think that the imagining relied upon in creative cognition cannot be reduced to more basic folk psychological states. The relevance of associative thought processes to songwriting is then explored as a means for understanding the nature of transformative creativity. Productive artificial neural networks—known as generative antagonistic networks (GANs)—are a recent example of (...) how a system’s ability to generate novel products can both be finely tuned by prior experience and grounded in strategies that cannot be articulated by the system itself. Further, the kinds of processes exploited by GANs need not be seen as incorporating something akin to sui generis imaginative states. The chapter concludes with reflection on the added relevance of personal character to explanations of creativity. [This is Chapter 12 of the book Explaining Imagination.]. (shrink)

Currie’s (2010) argument that “i-desires” must be posited to explain our responses to fiction is critically discussed. It is argued that beliefs and desires featuring ‘in the fiction’ operators—and not sui generis imaginings (or "i-beliefs" or "i-desires")—are the crucial states involved in generating fiction-directed affect. A defense of the “Operator Claim” is mounted, according to which ‘in the fiction’ operators would be also be required within fiction-directed sui generis imaginings (or "i-beliefs" and "i-desires"), were there such. Once we appreciate that (...) even fiction-directed sui generis imaginings would need to incorporate ‘in the fiction’ operators, the main appeal of the idea that sui generis imaginings (or "i-beliefs" or "i-desires") are at work in fiction-appreciation dissipates. [This is Chapter 10 of Explaining Imagination (OUP, 2020)]. (shrink)

The book’s core argument is that an artificial intelligence that could equal or exceed human intelligence—sometimes called artificial general intelligence (AGI)—is for mathematical reasons impossible. It offers two specific reasons for this claim: Human intelligence is a capability of a complex dynamic system—the human brain and central nervous system. Systems of this sort cannot be modelled mathematically in a way that allows them to operate inside a computer. In supporting their claim, the authors, Jobst Landgrebe and Barry Smith, marshal evidence (...) from mathematics, physics, computer science, philosophy, linguistics, and biology, setting up their book around three central questions: What are the essential marks of human intelligence? What is it that researchers try to do when they attempt to achieve "artificial intelligence" (AI)? And why, after more than 50 years, are our most common interactions with AI, for example with our bank’s computers, still so unsatisfactory? Landgrebe and Smith show how a widespread fear about AI’s potential to bring about radical changes in the nature of human beings and in the human social order is founded on an error. There is still, as they demonstrate in a final chapter, a great deal that AI can achieve which will benefit humanity. But these benefits will be achieved without the aid of systems that are more powerful than humans, which are as impossible as AI systems that are intrinsically "evil" or able to "will" a takeover of human society. (shrink)

​Imagination will remain a mystery—we will not be able to explain imagination—until we can break it into parts we already understand. Explaining Imagination is a guidebook for doing just that, where the parts are other ordinary mental states like beliefs, desires, judgments, and decisions. In different combinations and contexts, these states constitute cases of imagining. This reductive approach to imagination is at direct odds with the current orthodoxy, according to which imagination is a sui generis mental state or process—one with (...) its own inscrutable principles of operation. Explaining Imagination upends that view, showing how, on closer inspection, the imaginings at work in hypothetical reasoning, pretense, the enjoyment of fiction, and creativity are reducible to other familiar mental states—judgments, beliefs, desires, and decisions among them. Crisscrossing contemporary philosophy of mind, cognitive science, and aesthetics, Explaining Imagination argues that a clearer understanding of imagination is already well within reach. (shrink)

Artificial intelligence has historically been conceptualized in anthropomorphic terms. Some algorithms deploy biomimetic designs in a deliberate attempt to effect a sort of digital isomorphism of the human brain. Others leverage more general learning strategies that happen to coincide with popular theories of cognitive science and social epistemology. In this paper, I challenge the anthropomorphic credentials of the neural network algorithm, whose similarities to human cognition I argue are vastly overstated and narrowly construed. I submit that three alternative supervised learning (...) methods—namely lasso penalties, bagging, and boosting—offer subtler, more interesting analogies to human reasoning as both an individual and a social phenomenon. Despite the temptation to fall back on anthropomorphic tropes when discussing AI, however, I conclude that such rhetoric is at best misleading and at worst downright dangerous. The impulse to humanize algorithms is an obstacle to properly conceptualizing the ethical challenges posed by emerging technologies. (shrink)

The ontology of social objects and facts remains a field of continued controversy. This situation complicates the life of social scientists who seek to make predictive models of social phenomena. For the purposes of modelling a social phenomenon, we would like to avoid having to make any controversial ontological commitments. The overwhelming majority of models in the social sciences, including statistical models, are built upon ontological assumptions that can be questioned. Recently, however, artificial neural networks have made their way into (...) the social sciences, raising the question whether they can avoid controversial ontological assumptions. ANNs are largely distinguished from other statistical and machine learning techniques by being a representation-learning technique. That is, researchers can let the neural networks select which features of the data to use for internal representation instead of imposing their preconceptions. On this basis, I argue that neural networks can avoid ontological assumptions to a greater degree than common statistical models in the social sciences. I then go on, however, to establish that ANNs are not ontologically innocent either. The use of ANNs in the social sciences introduces ontological assumptions typically in at least two ways, via the input and via the architecture. (shrink)

There is a vast literature within philosophy of mind that focuses on artificial intelligence, but hardly mentions methodological questions. There is also a growing body of work in philosophy of science about modeling methodology that hardly mentions examples from cognitive science. Here these discussions are connected. Insights developed in the philosophy of science literature about the importance of idealization provide a way of understanding the neural implausibility of connectionist networks. Insights from neurocognitive science illuminate how relevant similarities between models and (...) targets are picked out, how modeling inferences are justified, and the metaphysical status of models. (shrink)

The massive introduction of artificial intelligence has triggered significant societal concerns, ranging from “technological unemployment” and the dominance of algorithms in the work place and in everyday life, among others. While AI is made by humans and is, therefore, dependent on the latter for its purpose, the increasing capabilities of AI to carry out productive activities for humans can lead the latter to unwitting slavish existence. This has become evident, for example, in the area of social media use, where AI (...) programmers tie psychology and persuasion to the human social need for approval and validation in ways that few users can resist. We argue that AI should serve humans with humans as masters and not the other way around. Moreover, we propose that virtue ethics might play a role to solidify the human as master of AI and guard against the alternative of AI as the master. (shrink)

Conceptual engineering means to provide a method to assess and improve our concepts working as cognitive devices. But conceptual engineering still lacks an account of what concepts are (as cognitive devices) and of what engineering is (in the case of cognition). And without such prior understanding of its subject matter, or so it is claimed here, conceptual engineering is bound to remain useless, merely operating as a piecemeal approach, with no overall grip on its target domain. The purpose of this (...) programmatic paper is to overcome this knowledge gap by providing some guidelines for developing the theories of concepts and of cognition that will ground the systematic unified framework needed to effectively implement conceptual engineering as a widely applicable method for the cognitive optimization of our conceptual devices. (shrink)

One goal of this paper is to argue that autobiographical memories are extended and distributed across embodied brains and environmental resources. This is important because such distributed memories play a constitutive role in our narrative identity. So, some of the building blocks of our narrative identity are not brain-bound but extended and distributed. Recognising the distributed nature of memory and narrative identity, invites us to find treatments and strategies focusing on the environment in which dementia patients are situated. A second (...) goal of this paper is to suggest various of such strategies, including lifelogging technologies such as SenseCams, life story books, multimedia biographies, memory boxes, ambient intelligence systems, and virtual reality applications. Such technologies allow dementia patients to remember their personal past in a way that wouldn’t be possible by merely relying on their biological memory, in that way aiding in preserving their narrative identity and positively contributing to their sense of well-being. (shrink)

The ongoing explosion of interest in artificial intelligence is fueled in part by recently developed techniques in machine learning. Those techniques allow automated systems to process huge amounts of data, utilizing mathematical methods that depart from traditional statistical approaches, and resulting in impressive advancements in our ability to make predictions and uncover correlations across a host of interesting domains. But as is now widely discussed, the way that those systems arrive at their outputs is often opaque, even to the experts (...) who design and deploy them. Is it morally problematic to make use of opaque automated methods when making high-stakes decisions, like whether to issue a loan to an applicant, or whether to approve a parole request? Many scholars answer in the affirmative. However, there is no widely accepted explanation for why transparent systems are morally preferable to opaque systems. We argue that the use of automated decision-making systems sometimes violates duties of consideration that are owed by decision-makers to decision-subjects, duties that are both epistemic and practical in character. Violations of that kind generate a weighty consideration against the use of opaque decision systems. In the course of defending our approach, we show that it is able to address three major challenges sometimes leveled against attempts to defend the moral import of transparency in automated decision-making. (shrink)

There is a distinction between merely having the right belief, and further basing that belief on the right reasons. Any adequate epistemology needs to be able to accommodate the basing relation that marks this distinction. However, trouble arises for Bayesianism. I argue that when we combine Bayesianism with the standard approaches to the basing relation, we get the result that no agent forms their credences in the right way; indeed, no agent even gets close. This is a serious problem, for (...) it prevents us from making epistemic distinctions between agents that are doing a reasonably good job at forming their credences and those that are forming them in clearly bad ways. I argue that if this result holds, then we have a problem for Bayesianism. However, I show how the Bayesian can avoid this problem by rejecting the standard approaches to the basing relation. By drawing on recent work on the basing relation, we can develop an account of the relation that allows us to avoid the result that no agent comes close to forming their credences in the right way. The Bayesian can successfully accommodate the basing relation. (shrink)

The article provides an analysis of Paul and Patricia Churchland’s eliminative materialism. I will distinguish two lines of argument in their eliminativism: one seeking to eliminate folk psychology and the second criticising Jerry Fodor’s language of thought hypothesis. Then I will closely examine the second line of argument, and show that it represents the main motive of Churchland’s work since the end of 1980s and demonstrate why the success of the argument against the language of thought hypothesis does not constitute (...) a reason for the elimination of folk psychology. Finally, I will examine the consequences of this approach for the role of folk psychology in the study of mind and show that the weakened eliminativist position still fulfils the original aim of Churchland’s program. (shrink)

Human behavior and thought often exhibit a familiar pattern of within group similarity and between group difference. Many of these patterns are attributed to cultural differences. For much of the history of its investigation into behavior and thought, however, cognitive science has been disproportionately focused on uncovering and explaining the more universal features of human minds—or the universal features of minds in general. -/- This entry charts out the ways in which this has changed over recent decades. It sketches the (...) motivation behind the cultural turn in cognitive science, and situates some of its central findings with respect to the questions that animate it and the debates that it has inspired. Woven throughout the entry are examples of how the cognitive science of culture, and especially its elevated concern with different forms of diversity and variation, continues to influence and be influenced by philosophers. -/- One cluster of philosophical work falls within the traditional subject matter of philosophy of science, in this case of the cognitive and social sciences. Philosophers have analyzed and assessed the methods and evidence central to the scientific study of cognition and culture, and have offered conceptual scrutiny, clarification, and synthesis. Research in a second vein sees philosophers themselves contributing more directly to cognitive scientific projects, (co)constructing theories, helping build computational models, even gathering empirical data. A third kind of work is naturalistic philosophy or philosophy of nature, wherein philosophers seek to use results from the cognitive science of culture to inform or transform debates over long-standing philosophical questions, including questions about the nature of philosophy and philosophical methodology itself. (shrink)

Paul Cilliers has developed a novel post-structural approach to complexity that has influenced several writers contributing to the current complexity literature. Concomitantly however, Cilliers advocates for modelling complex systems using connectionist neural networks (rather than analytic, rule-based models). In this paper, I argue that it is dilemmic to simultaneously hold these two positions. Cilliers’ post-structural interpretation of complexity states that models of complex systems are always contextual and provisional; there is no exclusive model of complex systems. This sentiment however appears (...) at odds with Cilliers’ promotion of connectionist neural networks as the best way to model complex systems. The lesson is that those who currently follow Cilliers’ post-structural approach to complexity cannot also develop a preferred model of complex systems, and those who currently advocate for some preferred model of complex systems cannot adopt the post-structural approach to complexity without giving up the purported objectivity and/or superiority of their preferred model. (shrink)

Computational complexity is a discipline of computer science and mathematics which classifies computational problems depending on their inherent difficulty, i.e. categorizes algorithms according to their performance, and relates these classes to each other. P problems are a class of computational problems that can be solved in polynomial time using a deterministic Turing machine while solutions to NP problems can be verified in polynomial time, but we still do not know whether they can be solved in polynomial time as well. A (...) solution for the so-called NP-complete problems will also be a solution for any other such problems. Its artificial-intelligence analogue is the class of AI-complete problems, for which a complete mathematical formalization still does not exist. In this chapter we will focus on analysing computational classes to better understand possible formalizations of AI-complete problems, and to see whether a universal algorithm, such as a Turing test, could exist for all AI-complete problems. In order to better observe how modern computer science tries to deal with computational complexity issues, we present several different deep-learning strategies involving optimization methods to see that the inability to exactly solve a problem from a higher order computational class does not mean there is not a satisfactory solution using state-of-the-art machine-learning techniques. Such methods are compared to philosophical issues and psychological research regarding human abilities of solving analogous NP-complete problems, to fortify the claim that we do not need to have an exact and correct way of solving AI-complete problems to nevertheless possibly achieve the notion of strong AI. (shrink)

Hierarchies of evidence are an important and influential tool for appraising evidence in medicine. In recent years, hierarchies have been formally adopted by organizations including the Cochrane Collaboration [1], NICE [2,3], the WHO [4], the US Preventive Services Task Force [5], and the Australian NHMRC [6,7]. The development of such hierarchies has been regarded as a central part of Evidence-Based Medicine, a movement within healthcare which prioritises the use of epidemiological evidence such as that provided by Randomised Controlled Trials. Philosophical (...) work on the methodology of medicine has so far mostly focused on claims about the superiority of RCTs, and hence has largely neglected the questions of what hierarchies are, what assumptions they require, and how they affect clinical practice. This thesis shows that there is great variation in the hierarchies defended and in the interpretations they are, and can be, given. The interpretative assumptions made in using hierarchies are crucial to the content and defensibility of the underlying philosophical commitments concerning evidence and medical practice. Once this variation is been identified, it becomes clear that the little philosophical work that has been done so far affects only some hierarchies, under some interpretations. Modest interpretations offered by La Caze [11], conditional hierarchies like GRADE [12-14], and heuristic approaches such as that defended by Howick et al. [15,16] all survive previous philosophical criticism. This thesis extends previous criticisms by arguing that modest interpretations are so weak as to be unhelpful for clinical practice; that GRADE and similar conditional models omit clinically relevant information, such as information about variation in treatments’ effects and the causes of different responses to therapy; and that heuristic approaches lack the necessary empirical support. The conclusion is that hierarchies in general embed untenable philosophical assumptions: principally that information about average treatment effects backed by high-quality evidence can justify strong recommendations, and that the impact of evidence from individual studies can and should be appraised in isolation. Hierarchies are a poor basis for the application of evidence in clinical practice. The Evidence-Based Medicine movement should move beyond them and explore alternative tools for appraising the overall evidence for therapeutic claims. (shrink)

Models of memory in cognitive science and philosophy have traditionally explained human remembering in terms of storage and retrieval. This tendency has been entrenched by reliance on computationalist explanations over the course of the twentieth century; even research programs that eschew computationalism in name, or attempt the revision of traditional models, demonstrate tacit commitment to computationalist assumptions. It is assumed that memory must be stored by means of an isomorphic trace, that memory processes must divide into conceptually distinct systems and (...) phases, and that human remembering consists in inner, cognitive processes that are implemented by distinct neural processes. This dissertation draws on recent empirical work, and on philosophical arguments from Ludwig Wittgenstein and others, to demonstrate that this latent computationalism in the study of memory is problematic, and that it can and should be eliminated. Cognitive psychologists studying memory have encountered numerous data in recent decades that belie archival models. In cognitive neuroscience, establishing the neural basis of storage and retrieval processes has proven elusive. A number of revised models on offer in memory science, that have taken these issues into account, fail to sufficiently extricate the archival framework. Several impasses in memory science are products of these underlying computationalist assumptions. Wittgenstein and other philosophers offer a number of arguments against the need for, and the efficacy of, the storage and retrieval of traces in human remembering. A study of these arguments clarifies the ways that these computationalist assumptions are presently impeding the science of memory, and provides ways forward in removing them. We can and should characterize and model human memory without invoking the storage and retrieval of traces. A range of work in connectionism, dynamical systems theory, and recent philosophical accounts of memory demonstrate how the science of memory can proceed without these assumptions, toward non-archival models of remembering. (shrink)