We construct a machine that knows its own code, at the price of not knowing its own factivity.

Artificial Knowing challenges the masculine slant in the Artificial Intelligence (AI) view of the world. Alison Adam admirably fills the large gap in science and technology studies by showing us that gender bias is inscribed in AI-based computer systems. Her treatment of feminist epistemology, focusing on the ideas of the knowing subject, the nature of knowledge, rationality and language, are bound to make a significant and powerful contribution to AI studies. Drawing from theories by Donna Haraway and Sherry (...) Turkle, and using tools of feminist epistemology, Adam provides a sustained critique of AI which interestingly re-enforces many of the traditional criticisms of the AI project. Artificial Knowing is an esential read for those interested in gender studies, science and technology studies, and philosophical debates in AI. (shrink)

_Artificial Knowing_ challenges the masculine slant in the Artificial Intelligence view of the world. Alison Adam admirably fills the large gap in science and technology studies by showing us that gender bias is inscribed in AI-based computer systems. Her treatment of feminist epistemology, focusing on the ideas of the knowing subject, the nature of knowledge, rationality and language, are bound to make a significant and powerful contribution to AI studies. Drawing from theories by Donna Haraway and Sherry Turkle, and (...) using tools of feminist epistemology, Adam provides a sustained critique of AI which interestingly re-enforces many of the traditional criticisms of the AI project. _Artificial Knowing_ is an esential read for those interested in gender studies, science and technology studies, and philosophical debates in AI. (shrink)

A book Review of Artificial Knowing Gender and the Thinking Machine, by Alison Adam, Routledge: Taylor and Francis, 1998.

It is nowadays widely understood that undeclared work cannot be efficiently combated without a holistic view on the mechanisms underlying its existence. However, the question remains whether we possess all the pieces of the holistic puzzle. To fill the gap, in this paper, we test if the features so far known to affect the behaviour of taxpayers are sufficient to detect noncompliance with outstanding precision. This is done by training seven supervised machine learning models on the compilation of data from (...) the 2019 Special Eurobarometer on undeclared work and relevant figures from other sources. The conducted analysis not only does attest to the completeness of our knowledge concerning the drivers of undeclared work but also paves the way for wide usage of artificial intelligence in monitoring and confronting this detrimental practice. The study, however, exposes the necessity of having at disposal considerably larger datasets compared to those currently available if successful real-world applications of machine learning are to be achieved in this field. Alongside the apparent theoretical contribution, this paper is thus also expected to be of particular importance for policymakers, whose efforts to tackle tax evasion will have to be expedited in the period after the COVID-19 pandemic. (shrink)

Storrs McCall appeals to a particular true but improvable sentence of formal arithmetic to argue, by appeal to its irrefutability, that human minds transcend Turing machines. Metamathematical oversights in McCall's discussion of the Godel phenomena, however, render invalid his philosophical argument for this transcendentalist conclusion.

Bringing together evolutionary biology, psychology, and philosophy, Henry Plotkin presents a new science of knowledge, one that traces an unbreakable link between instinct and our ability to know.

This article asks the question, ``what is reliable machine learning?'' As I intend to answer it, this is a question about epistemic justification. Reliable machine learning gives justification for believing its output. Current approaches to reliability (e.g., transparency) involve showing the inner workings of an algorithm (functions, variables, etc.) and how they render outputs. We then have justification for believing the output because we know how it was computed. Thus, justification is contingent on what can be shown about the algorithm, (...) its properties, and its behavior. In this paper, I defend computational reliabilism (CR). CR is a computationally-inspired off-shoot of process reliabilism that does not require showing the inner workings of an algorithm. CR credits reliability to machine learning by identifying reliability indicators external to the algorithm (validation methods, knowledge-based integration, etc.). Thus, we have justification for believing the output of machine learning when we have identified the appropriate reliability indicators. CR is advanced as a more suitable epistemology for machine learning. The main goal of this article is to lay the groundwork for CR, how it works, and what we can expect as a justificatory framework for reliable machine learning. (shrink)

Genes are often described by biologists using metaphors derived from computa- tional science: they are thought of as carriers of information, as being the equivalent of ‘‘blueprints’’ for the construction of organisms. Likewise, cells are often characterized as ‘‘factories’’ and organisms themselves become analogous to machines. Accordingly, when the human genome project was initially announced, the promise was that we would soon know how a human being is made, just as we know how to make airplanes and buildings. Impor- (...) tantly, modern proponents of Intelligent Design, the latest version of creationism, have exploited biologists’ use of the language of information and blueprints to make their spurious case, based on pseudoscientific concepts such as ‘‘irreducible complexity’’ and on flawed analogies between living cells and mechanical factories. However, the living organ- ism = machine analogy was criticized already by David Hume in his Dialogues Concerning Natural Religion. In line with Hume’s criticism, over the past several years a more nuanced and accurate understanding of what genes are and how they operate has emerged, ironically in part from the work of computational scientists who take biology, and in particular developmental biology, more seriously than some biologists seem to do. In this article we connect Hume’s original criticism of the living organism = machine analogy with the modern ID movement, and illustrate how the use of misleading and outdated metaphors in science can play into the hands of pseudoscientists. Thus, we argue that dropping the blueprint and similar metaphors will improve both the science of biology and its understanding by the general public. (shrink)

The notion of computation has changed the world more than any previous expressions of knowledge. However, as know-how in its particular algorithmic embodiment, computation is closed to meaning. Therefore, computer-based data processing can only mimic life’s creative aspects, without being creative itself. AI’s current record of accomplishments shows that it automates tasks associated with intelligence, without being intelligent itself. Mistaking the abstract for the concrete has led to the religion of “everything is an output of computation”—even the humankind that conceived (...) the computer. The hypostatized role of computers explains the increased dependence on them. The convergence machine called deep learning is only the most recent form through which the deterministic theology of the machine claims more than what it actually is: extremely effective data processing. A proper understanding of complexity, as well as the need to distinguish between the reactive nature of the artificial and the anticipatory nature of the living are suggested as practical responses to the challenges posed by machine theology. (shrink)

Reinhardt 's strong mechanistic thesis, a formalization of “I know I am a Turing machine”, is shown to be consistent with Epistemic Arithmetic.

Intelligence is not a property unique to the human brain; rather it represents a spectrum of phenomena. An understanding of the evolution of intelligence makes it clear that the evolution of machine intelligence has no theoretical limits — unlike the evolution of the human brain. Machine intelligence will outpace human intelligence and very likely will do so during the lifetime of our children. The mix of advanced machine intelligence with human individual and communal intelligence will create an evolutionary discontinuity as (...) profound as the origin of life. It will presage the end of the human species as we know it. The question, in the author's view, is not whether this will happen, but when, and what should be our response. (shrink)

There is growing interest in machine ethics in the question of whether and under what circumstances an artificial intelligence would deserve moral consideration. This paper explores a particular type of moral status that the author terms psychological moral patiency, focusing on the epistemological question of what sort of evidence might lead us to reasonably conclude that a given artificial system qualified as having this status. The paper surveys five possible criteria that might be applied: intuitive judgments, assessments of intelligence, the (...) presence of desires and autonomous behavior, evidence of sentience, and behavioral equivalence. The author suggests that despite its limitations, the latter approach offers the best way forward, and defends a variant of that, termed the cognitive equivalence strategy. In short, this holds that an artificial system should be considered a psychological moral patient to the extent that it possesses cognitive mechanisms shared with other beings such as nonhuman animals whom we also consider to be psychological moral patients. -/- . (shrink)

While feminist critiques of AI are increasingly common in the scholarly literature, they are by no means new. Alison Adam’s Artificial Knowing (1998) brought a feminist social and epistemological stance to the analysis of AI, critiquing the symbolic AI systems of her day and proposing constructive alternatives. In this paper, we seek to revisit and renew Adam’s arguments and methodology, exploring their resonances with current feminist concerns and their relevance to contemporary machine learning. Like Adam, we ask how new (...) AI methods could be adapted for feminist purposes and what role new technologies might play in addressing concerns raised by feminist epistemologists and theorists about algorithmic systems. In particular, we highlight distributed and federated learning as providing partial solutions to the power-oriented concerns that have stymied efforts to make machine learning systems more representative and pluralist. (shrink)

This book provides a novel framework for understanding and revising labor markets and education policies in an era of machine learning. It posits that while learning and knowing both require thinking, learning is fundamentally different than knowing because it results in cognitive processes that change over time. Learning, in contrast to knowing, requires time and agency. Therefore, “learning algorithms”—that enable machines to modify their actions based on real-world experiences—are a fundamentally new form of artificial intelligence that (...) have potential to be even more disruptive to labor markets than prior introductions of digital technology. To explore the difference between knowing and learning, Turing’s “Imitation Game,”—that he proposed as a test for machine thinking—is expanded to include time dependence. The arguments presented in the book introduce three novel concepts: (1) Comparative learning advantage: This is a concept analogous to comparative labor advantage but arises from the disparate times required to learn new knowledge bases/skillsets. It is argued that in the future, comparative learning advantages between humans and machines will determine their division of labor. (2) Two dimensions of job performance—expertise and interpersonal: Job tasks can be sorted into two broad categories. Tasks that require expertise have stable endpoints, which makes these tasks inherently repetitive and subject to automation. Tasks that are interpersonal are highly context-dependent and lack stable endpoints, which makes these tasks inherently non-routine. Humans compared to machines have a comparative learning advantage along the interpersonal dimension, which is increasing in value economically. (3) The Learning Game is a time-dependent version of Turing’s “Imitation Game.” It is more than a thought experiment. The “Learning Game” provides a mathematical framework with quantitative criteria for training and assessing comparative learning advantages. The book is highly interdisciplinary—presenting philosophical arguments in economics, artificial intelligence, and education. It also provides data, mathematical analysis, and testable criteria that researchers in these fields will find of practical use. The book calls for a rethinking of how labor markets operate and how the education system should prepare students for future jobs. It concludes with a list of counterintuitive recommendations for future education and labor policies that all stakeholders—employers, employees, educators, students, and political leaders—should heed. (shrink)

The present paper shows how statistical learning theory and machine learning models can be used to enhance understanding of AI-related epistemological issues regarding inductive reasoning and reliability of generalisations. Towards this aim, the paper proceeds as follows. First, it expounds Price’s dual image of representation in terms of the notions of e-representations and i-representations that constitute subject naturalism. For Price, this is not a strictly anti-representationalist position but rather a dualist one (e- and i-representations). Second, the paper links this debate (...) with machine learning in terms of statistical learning theory becoming more viable epistemological tool when it abandons the perspective of object naturalism. The paper then argues that machine learning grounds a form of knowing that can be understood in terms of e- and i-representation learning. Third, this synthesis shows a way of analysing inductive reasoning in terms of reliability of generalisations stemming from a structure of e- and i-representations. In the age of Artificial Intelligence, connecting Price’s dual view of representation with Deep Learning provides an epistemological way forward and even perhaps an approach to how knowing is possible. (shrink)

This chapter provides a classification scheme for artificial intelligence (AI) predictions, and tools for analyzing their reliability and uncertainties. It presents a series of brief case studies of some of the most famous AI predictions: the initial Dartmouth AI conference; Hubert Dreyfus' criticism of AI; Ray Kurzweil's predictions in The Age of Spiritual Machines; and Stephen Omohundro's AI Drives. The chapter takes every falsifiable statement about future AI to be a prediction. Thus the following four categories are all predictions: (...) Timelines and outcome predictions, Scenarios, Plans, and Issues and metastatements. The following schema was suggested by a review of the AI predictions literature : Causal models, Non‐causal models, The outside view, Philosophical arguments, and Expert judgment. The chapter looks at four prominent AI predictions, and analyses their accuracy, attempting to gain insights that will be useful for assessing future prediction. (shrink)

I found the JCS issue on Machine Consciousness, Volume 10, No. 4-5 , frustrating and alienating. There seems to be a consensus building that consciousness is accessible to scientific scrutiny, so much so that it is already understood well enough to be modeled and even synthesized. I'm not so sure. It could be instead that the vocabulary of consciousness is being subtly redefined to be amenable to scientific investigation and explicit modeling. Such semantic revisionism is confusing and often misleading. Whatever (...) else consciousness is, it is at least a certain quality of life apparent from personal reflection. Introspection is, after all, the only way we know that consciousness even exists. Scientific and technical redefinitions that fail to account for its phenomenal quality are at best incomplete. In my view, all but one of the ten articles in the JCS volume on Machine Consciousness commit various degrees of Protean distortion. (shrink)

Seeing, Doing, and Knowing is an original and comprehensive philosophical treatment of sense perception as it is currently investigated by cognitive neuroscientists. Its central theme is the task-oriented specialization of sensory systems across the biological domain. Sensory systems are automatic sorting machines; they engage in a process of classification. Human vision sorts and orders external objects in terms of a specialized, proprietary scheme of categories - colours, shapes, speeds and directions of movement, etc. This 'Sensory Classification Thesis' implies (...) that sensation is not a naturally caused image from which an organism must infer the state of the world beyond; it is more like an internal communication, a signal concerning the state of the world issued by a sensory system, in accordance with internal conventions, for the use of an organism's other systems. This is why sensory states are both easily understood and persuasive.Sensory classification schemes are purpose-built to serve the knowledge-gathering and pragmatic needs of particular types of organisms. They are specialized: a bee or a bird does not see exactly what a human does. The Sensory Classification Thesis helps clarify this specialization in perceptual content and supports a new form of realism about the deliverances of sensation: 'Pluralistic Realism' is based on the idea that sensory systems coevolve with an organism's other systems; they are not simply moulded to the external world. The last part of the book deals with reference in vision. Cognitive scientists now believe that vision guides the limbs by means of a subsystem that links up with the objects of physical manipulation in ways that bypass sensory categories. In a novel extension of this theory, Matthen argues that 'motion-guiding vision' is integrated with sensory classification in conscious vision. This accounts for the quasi-demonstrative form of visual states: 'This particular object is red', and so on. He uses this idea to cast new light on the nature of perceptual objects, pictorial representation, and the visual representation of space. (shrink)

Philosophy has long been fascinated by miracles, and with good reason. Where, however, the problem of the miracle once offered unparalleled insight into the inner workings of natural laws and of human knowledge, today, the attention commanded by it is essentially political. The sovereign’s miraculous suspension is the most well studied of these political dimensions, but this formulation is, in fact, ill-suited to the complexities inherent in the concept of the miracle. Political theology understands the miracle poorly, for it captures (...) only the inaugural movement of exception; it knows nothing of the social and political conditions it inspires. A key claim defended in this article is that the miracle possesses continuing interest for philosophy precisely because it lies in the heart of contemporary political formations, instituting a bond that tethers a collective to its present. It is thus more interesting as a figure of inception than of exception. The miraculous is uniformly distinguished as the unengendered, the neutral point of equilibrium against which any distribution of beings is measured. This is the meaning given the miraculous in the thought of Gilles Deleuze and Félix Guattari, and this article frames the political aspect of that theme by developing their concept of miraculation, which foregrounds the movement of inception. The article works out a conceptual genealogy, tracing the descent of the miracle back from Deleuze and Guattari to Baruch Spinoza and David Hume. Each contributes an element essential for the constitution of miraculation. It is argued that the miracle has always bound ‘desire and the social’, though in diverse ways. (shrink)

Consciousness is widely perceived as one of the most fundamental, interesting and difficult problems of our time. However, we still know next to nothing about the relationship between consciousness and the brain and we can only speculate about the consciousness of animals and machines. Human and Machine Consciousness presents a new foundation for the scientific study of consciousness. It sets out a bold interpretation of consciousness that neutralizes the philosophical problems and explains how we can make scientific predictions about (...) the consciousness of animals, brain-damaged patients and machines. Gamez interprets the scientific study of consciousness as a search for mathematical theories that map between measurements of consciousness and measurements of the physical world. We can use artificial intelligence to discover these theories and they could make accurate predictions about the consciousness of humans, animals and artificial systems. Human and Machine Consciousness also provides original insights into unusual conscious experiences, such as hallucinations, religious experiences and out-of-body states, and demonstrates how 'designer' states of consciousness could be created in the future. Gamez explains difficult concepts in a clear way that closely engages with scientific research. His punchy, concise prose is packed with vivid examples, making it suitable for the educated general reader as well as philosophers and scientists. Problems are brought to life in colourful illustrations and a helpful summary is given at the end of each chapter. The endnotes provide detailed discussions of individual points and full references to the scientific and philosophical literature."--Publisher's website. (shrink)

The advent of the intelligent robot has occupied a significant position in society over the past decades and has given rise to new issues in society. As we know, the primary aim of artificial intelligence or robotic research is not only to develop advanced programs to solve our problems but also to reproduce mental qualities in machines. The critical claim of artificial intelligence advocates is that there is no distinction between mind and machines and thus they argue that (...) there are possibilities for machine ethics, just as human ethics. Unlike computer ethics, which has traditionally focused on ethical issues surrounding human use of machines, AI or machine ethics is concerned with the behaviour of machines towards human users and perhaps other machines as well, and the ethicality of these interactions. The ultimate goal of machine ethics, according to the AI scientists, is to create a machine that itself follows an ideal ethical principle or a set of principles; that is to say, it is guided by this principle or these principles in decisions it makes about possible courses of action it could takea. Thus, machine ethics task of ensuring ethical behaviour of an artificial agent. Although, there are many philosophical issues related to artificial intelligence, but our attempt in this paper is to discuss, first, whether ethics is the sort of thing that can be computed. Second, if we are ascribing mind to machines, it gives rise to ethical issues regarding machines. And if we are not drawing the difference between mind and machines, we are not only redefining specifically human mind but also the society as a whole. Having a mind is, among other things, having the capacity to make voluntary decisions and actions. The notion of mind is central to our ethical thinking, and this is because the human mind is self-conscious, and this is a property that machines lack, as yet. (shrink)

Data science is not simply a method but an organising idea. Commitment to the new paradigm overrides concerns caused by collateral damage, and only a counterculture can constitute an effective critique. Understanding data science requires an appreciation of what algorithms actually do; in particular, how machine learning learns. The resulting ‘insight through opacity’ drives the observable problems of algorithmic discrimination and the evasion of due process. But attempts to stem the tide have not grasped the nature of data science as (...) both metaphysical and machinic. Data science strongly echoes the neoplatonism that informed the early science of Copernicus and Galileo. It appears to reveal a hidden mathematical order in the world that is superior to our direct experience. The new symmetry of these orderings is more compelling than the actual results. Data science does not only make possible a new way of knowing but acts directly on it; by converting predictions to pre-emptions, it becomes a machinic metaphysics. The people enrolled in this apparatus risk an abstraction of accountability and the production of ‘thoughtlessness’. Susceptibility to data science can be contested through critiques of science, especially standpoint theory, which opposes the ‘view from nowhere’ without abandoning the empirical methods. But a counterculture of data science must be material as well as discursive. Karen Barad’s idea of agential realism can reconfigure data science to produce both non-dualistic philosophy and participatory agency. An example of relevant praxis points to the real possibility of ‘machine learning for the people’. (shrink)

A "machine" is any causal physical system, hence we are machines, hence machines can be conscious. The question is: which kinds of machines can be conscious? Chances are that robots that can pass the Turing Test -- completely indistinguishable from us in their behavioral capacities -- can be conscious (i.e. feel), but we can never be sure (because of the "other-minds" problem). And we can never know HOW they have minds, because of the "mind/body" problem. We can (...) only know how they pass the Turing Test, but not how, why or whether that makes them feel. (shrink)

Are there aspects of human intelligence that artificial intelligence cannot emulate? Harry Collins uses a distinction between tacit aspects of knowing, which cannot be digitized, and explicit aspects, which can be, to formulate an answer to this question. He postulates three purported areas of the tacit and argues that only “collective tacit knowing” cannot be adequately digitized. I argue, first, that Collins’s approach rests upon problematic Cartesian assumptions—particularly his claim that animal knowing is strictly deterministic and, thus, (...) radically different from human knowing. I offer evidence that human linguistic intelligence depends upon embodied forms of animal intelligence. Second, I suggest the development of deep-learning algorithms means that Collins’s mimesis assumption is no longer appropriate; equivalent accomplishment of goals is what counts. However, persons must realize and integrate many goals and also deal with failures; a general-purpose AI capable of integrating all the needs and goals of human existence resists development. Third, I explain how Michael Polanyi’s understanding of tacit knowing, quite different than Collins’s concept of the tacit, exemplifies features that are missing in contemporary AI. I make use of evolutionary theory, studies of animal intelligence, biological insights, individual construction of meaning, and notions of human responsibility to argue for the existence of three categories that distinguish human from artificial intelligence. (shrink)

Experimental engineering models have been used both to model general phenomena, such as the onset of turbulence in fluid flow, and to predict the performance of machines of particular size and configuration in particular contexts. Various sorts of knowledge are involved in the method - logical consistency, general scientific principles, laws of specific sciences, and experience. I critically examine three different accounts of the foundations of the method of experimental engineering models (scale models), and examine how theory, practice, and (...) experience are involved in employing the method to obtain practical results. Models of machines and mechanisms can be (and generally are) involved in establishing criteria for similar phenomena, which provide guidance in using events to model other events. Conversely, models of phenomena such as events that model other events can be (and generally are) involved in experimentation on models of machines. I conclude that often it is not more detailed models or the more precise equations they engender that leads to better understanding, but rather an insightful use of knowledge at hand to determine which similarity principles are appropriate in allowing us to infer what we do not know from what we are able to observe. (shrink)

We define a notion of the intelligence level of an idealized mechanical knowing agent. This is motivated by efforts within artificial intelligence research to define real-number intelligence levels of compli- cated intelligent systems. Our agents are more idealized, which allows us to define a much simpler measure of intelligence level for them. In short, we define the intelligence level of a mechanical knowing agent to be the supremum of the computable ordinals that have codes the agent knows to (...) be codes of computable ordinals. We prove that if one agent knows certain things about another agent, then the former necessarily has a higher intelligence level than the latter. This allows our intelligence no- tion to serve as a stepping stone to obtain results which, by themselves, are not stated in terms of our intelligence notion (results of potential in- terest even to readers totally skeptical that our notion correctly captures intelligence). As an application, we argue that these results comprise evidence against the possibility of intelligence explosion (that is, the no- tion that sufficiently intelligent machines will eventually be capable of designing even more intelligent machines, which can then design even more intelligent machines, and so on). (shrink)

In our recent article ‘The Ethics of the Algorithmic Prediction of Goal of Care Preferences: From Theory to Practice’1, we aimed to ignite a critical discussion on why and how to design artificial intelligence (AI) systems assisting clinicians and next-of-kin by predicting goal of care preferences for incapacitated patients. Here, we would like to thank the commentators for their valuable responses to our work. We identified three core themes in their commentaries: (1) the risks of AI paternalism, (2) worries about (...) attacks to our humanity stemming from the use of AI and (3) the possibility of designing AI systems for more relevant use cases than the one we consider in our work. We shall focus on these themes, leaving aside some other interesting suggestions, given the limited space available for our response. Diaz Milian and Bhattacharyya discuss the risks of AI paternalism highlighting how the use of an AI to predict goal of care preferences may incentivise clinicians and surrogates to shift the burden of decision making to the machine.2 In particular, ‘[i]f the AI-generated response is given priority above the other agents’2 and the AI makes decisions autonomously, then ‘AI paternalism’ would emerge.2 Against this possibility, the authors suggest four safeguards to be implemented in the AI life-cycle.2 These are procedures that may improve the trustworthiness of the AI and human control over it.3 We agree that it is important …. (shrink)

Thinking about time travel is an entertaining way to explore how to understand time and its location in the broad conceptual landscape that includes causation, fate, action, possibility, experience, and reality. It is uncontroversial that time travel towards the future exists, and time travel to the past is generally recognized as permitted by Einstein’s general theory of relativity, though no one knows yet whether nature truly allows it. Coherent time travel stories have added flair to traditional debates over the metaphysical (...) status of the past, the reality of temporal passage, and the existence of free will. Moreover, plausible models of time travel and time machines can be used to investigate the subtle relation between space-time structure and causality. -/- It surveys some philosophical issues concerning time travel and should serves as a quick introduction. It includes a new and improved way to define a time machine. (shrink)

Machines have always been a tool or technical instrument for human beings to facilitate and to accelerate processes through mechanical power. The same applies to robots nowadays – the next step in the evolution of machines. Over the course of the last few years, robot usage in society has expanded enormously, and they now carry out a remarkable number of tasks for us. It seems we are on the eve of a historic revolution that will change everything we (...) know right now. But not only robots have an impact on our life. It is digitization in its entirety, including smart applications and games, that confronts us with new spaces. This special volume of Ethics in Progress tries to broaden our understanding of a philosophical field – robots and digitization – that is still in its infancy in terms of it research and literature. (shrink)

The claims that Dr. F. H. George makes on behalf of his machines are obscurely stated. Does he claim that a machine has been made and has actually produced a kind of response which is incalculable, given the specification to which it has been built and also the prescribed conditions, what is put in for the particular performance in question? “Incalculable” does not mean that nobody has bothered to calculate, but that somebody has bothered, that the calculations show that (...) the response in question is not among the calculated ones, that it is not chance or mechanical breakdown, but a specific response to something specific among the given conditions. If he could substantiate this claim, then I would reluctantly agree that some machines are cleverer and more thoughtful than I thought possible, and some men sillier and less thoughtful, for they deliberately design machines to do they know not what. (shrink)

This work has been selected by scholars as being culturally important and is part of the knowledge base of civilization as we know it. This work is in the public domain in the United States of America, and possibly other nations. Within the United States, you may freely copy and distribute this work, as no entity has a copyright on the body of the work. Scholars believe, and we concur, that this work is important enough to be preserved, reproduced, and (...) made generally available to the public. To ensure a quality reading experience, this work has been proofread and republished using a format that seamlessly blends the original graphical elements with text in an easy-to-read typeface. We appreciate your support of the preservation process, and thank you for being an important part of keeping this knowledge alive and relevant. (shrink)

One of the main worries with machine learning model opacity is that we cannot know enough about how the model works to fully understand the decisions they make. But how much is model opacity really a problem? This chapter argues that the problem of machine learning model opacity is entangled with non-epistemic values. The chapter considers three different stages of the machine learning modeling process that corresponds to understanding phenomena: (i) model acceptance and linking the model to the phenomenon, (ii) (...) explanation, and (iii) attributions of understanding. At each of these stages, non-epistemic values can, in part, determine how much machine learning model opacity poses a problem. (shrink)

In this paper, we address the question of what current computers are from the point of view of human-computer interaction. In the early days of computing, the Turing machine (TM) has been the cornerstone of the understanding of computers. The TM defines what can be computed and how computation can be carried out. However, in the last decades, computers have evolved and increasingly become interactive systems, reacting in real-time to external events in an ongoing loop. We argue that the TM (...) does not provide a mechanistic explanation for interactive computing. The reason is that the fundamental phenomena relevant to interactive computing are out of the scope of classical computability theory. Part of the explanatory power of the TM relies on what we propose to call an execution model. An execution model belongs to a level of abstraction where it is possible to describe both the functional architecture and the execution in mechanistic terms. An updated execution model is warranted to provide the minimal mechanistic description for interactive computation as a counterpart of what the TM could explain regarding Church-Turing computation. It would support an explanation of the ubiquitous computing devices we know - those interacting with humans, e.g., through digital interfaces. We show that such a model is not available within interactive models of computation and that relevant abstractions and concerns are available in computer engineering but need to be identified and gathered. To fill this void, we propose to reflect on the level of abstraction required to support the mechanistic description of an interactive execution and propose some preliminary requirements. (shrink)

There is a great deal in Man Machine and Other Writings that will delight the reader. Thomson has managed to capture much of La Mettrie’s wit and poetic use of language, which is no easy task; as La Mettrie himself comments on his “figurative style,” it “is often necessary in order to express better what is felt and to add grace to truth itself”. The central thesis of Man Machine needs little introduction. Inspired by the suggestion in Part 5 of (...) Descartes’s Discourse on Method that animals are machines, La Mettrie extends the metaphor to man. Man is like a watch whose springs and wheels, properly organized, move to perform its function. But before the reader concludes that La Mettrie is embracing Descartes’s reduction of biological phenomena to “lifeless” mechanism, La Mettrie further claims that matter itself cannot tell time. Only a body appropriately organized can have the property of self-motion, and, hence, perform such functions as telling time, or feeling, or thinking. La Mettrie’s primary task is to vitalize Cartesian mechanism to show that self-directed motion could be a property of organized matter: “I believe thought to be so little incompatible with organized matter that it seems to be one of its properties, like electricity, motive power, impenetrability, extension, etc.”. As for how inert, simple matter becomes active and composed of organs, and, in some cases, endowed with feeling and thought, La Mettrie appeals to our ignorance of causes which prevents us from knowing ultimate beginnings or ends. Such ignorance, however, does not prevent us from admitting such “incontrovertible observations” as that organized matter possesses a motive principle, whereas unorganized matter does not. Other themes in the work include his account of the faculty of imagination as the origin of thought, observations in comparative psychology, a sketch of thinking as a symbolic process, and a discussion of whether apes could speak. Man Machine, as Thomson explains in the introduction, is not a formal treatise, but, rather, a loosely structured polemic, often given to rhetorical flourishes. However, where Man Machine falls short on argumentation it will surely provoke and engage the reader. (shrink)

Globally, robots can be described as some sets of moving parts that are dedicated to a task while using their own energy. Yet, humans commonly qualify those machines as being intelligent, autonomous or being able to learn, know, feel, make decisions, etc. Is it merely a way of talking or does it mean that robots could eventually be more than a complex set of moving parts? On the one hand, the language of robotics allows multiple interpretations (leading sometimes to (...) misreading or confusion in various contexts). On the other hand, the status of robots is challenged more and more by technical achievements and humans’ own empirical beliefs. In this paper, we follow a linguistic approach in order to explore the relevance of these words when talking about robots. Since we note that the words impose themselves (even if opposed), we discuss the efficiency of a rhetorical strategy in order to work with such a lexicon in robotics. More precisely, we explore the argumentative technique of the dissociation of notions through the study of a practical case: the case of robot lawn mowers versus hedgehogs. (shrink)

Michael Polanyi’s conceptions of tacit knowing and emergent being are used to correct a reductionism that developed from, or reacted against, the excesses of several Cartesian assumptions: (a) the method of universal doubt; (b) the emphasis on reductive analysis to unshakeable foundations, via connections between clear and distinct ideas; (c) the notion that what is real are the basic atomic substances out of which all else is composed; (d) a sharp body-mind substance dualism; and (e) the notion that the (...) seat of consciousness can be traced to a point in the human body. The reductivist project in biology began with the emphasis Descartes put on the body as a machine. Polanyi reappropriates the machine metaphor to demonstrate how mechanistic explanations are not fully reductive. He shows how an eliminative materialism that would reduce mind to brain is unwarranted if either an interlevel mechanistic reduction or an intralevel successional reduction is posited. (shrink)

The recent emergence of machine-manipulated media raises an important societal question: How can we know whether a video that we watch is real or fake? In two online studies with 15,016 participants, we present authentic videos and deepfakes and ask participants to identify which is which. We compare the performance of ordinary human observers with the leading computer vision deepfake detection model and find them similarly accurate, while making different kinds of mistakes. Together, participants with access to the model’s prediction (...) are more accurate than either alone, but inaccurate model predictions often decrease participants’ accuracy. To probe the relative strengths and weaknesses of humans and machines as detectors of deepfakes, we examine human and machine performance across video-level features, and we evaluate the impact of preregistered randomized interventions on deepfake detection. We find that manipulations designed to disrupt visual processing of faces hinder human participants’ performance while mostly not affecting the model’s performance, suggesting a role for specialized cognitive capacities in explaining human deepfake detection performance. -/- . (shrink)

An argument that what is usually dismissed as the “mystical shell” of Hegel's thought—the concept of absolute knowledge—is actually its most “rational kernel.” This book sets out from a counterintuitive premise: the “mystical shell” of Hegel's system proves to be its most “rational kernel.” Hegel's radicalism is located precisely at the point where his thought seems to regress most. Most current readings try to update Hegel's thought by pruning back his grandiose claims to “absolute knowing.” Comay and Ruda invert (...) this deflationary gesture by inflating what seems to be most trivial: the absolute is grasped only in the minutiae of its most mundane appearances. Reading Hegel without presupposition, without eliminating anything in advance or making any decision about what is essential and what is inessential, what is living and what is dead, they explore his presentation of the absolute to the letter. The Dash is organized around a pair of seemingly innocuous details. Hegel punctuates strangely. He ends the Phenomenology of Spirit with a dash, and he begins the Science of Logic with a dash. This distinctive punctuation reveals an ambiguity at the heart of absolute knowing. The dash combines hesitation and acceleration. Its orientation is simultaneously retrospective and prospective. It both holds back and propels. It severs and connects. It demurs and insists. It interrupts and prolongs. It generates nonsequiturs and produces explanations. It leads in all directions: continuation, deviation, meaningless termination. This challenges every cliché about the Hegelian dialectic as a machine of uninterrupted teleological progress. The dialectical movement is, rather, structured by intermittency, interruption, hesitation, blockage, abruption, and random, unpredictable change—a rhythm that displays all the vicissitudes of the Freudian drive. (shrink)

Does the human mind resemble the machines that can behave like it? Biologically inspired machine-learning systems approach “human-level” accuracy in an astounding variety of domains, and even predict human brain activity—raising the exciting possibility that such systems represent the world like we do. However, even seemingly intelligent machines fail in strange and “unhumanlike” ways, threatening their status as models of our minds. How can we know when human–machine behavioral differences reflect deep disparities in their underlying capacities, vs. when (...) such failures are only superficial or peripheral? This article draws on a foundational insight from cognitive science—the distinction between performance and competence—to encourage “species-fair” comparisons between humans and machines. The performance/competence distinction urges us to consider whether the failure of a system to behave as ideally hypothesized, or the failure of one creature to behave like another, arises not because the system lacks the relevant knowledge or internal capacities (“competence”), but instead because of superficial constraints on demonstrating that knowledge (“performance”). I argue that this distinction has been neglected by research comparing human and machine behavior, and that it should be essential to any such comparison. Focusing on the domain of image classification, I identify three factors contributing to the species-fairness of human–machine comparisons, extracted from recent work that equates such constraints. Species-fair comparisons level the playing field between natural and artificial intelligence, so that we can separate more superficial differences from those that may be deep and enduring. (shrink)

In a 1955 lecture the physicist Richard Feynman reflected on the place of doubt within scientific practice. ‘Permit us to question, to doubt, to not be sure’, proposed Feynman, ‘it is possible to live and not to know’. In our contemporary world, the science of machine learning algorithms appears to transform the relations between science, knowledge and doubt, to make even the most doubtful event amenable to action. What might it mean to ‘leave room for doubt’ or ‘to live and (...) not to know’ in our contemporary culture, where the algorithm plays a major role in the calculability of doubts? I propose a posthuman mode of doubt that decentres the liberal humanist subject. In the science of machine learning algorithms the doubts of human and technological beings nonetheless dwell together, opening onto a future that is never fully reduced to the single output signal, to the optimised target. (shrink)

With Artificial Intelligence entering our lives in novel ways—both known and unknown to us—there is both the enhancement of existing ethical issues associated with AI as well as the rise of new ethical issues. There is much focus on opening up the ‘black box’ of modern machine-learning algorithms to understand the reasoning behind their decisions—especially morally salient decisions. However, some applications of AI which are no doubt beneficial to society rely upon these black boxes. Rather than requiring algorithms to be (...) transparent we should focus on constraining AI and those machines powered by AI within microenvironments—both physical and virtual—which allow these machines to realize their function whilst preventing harm to humans. In the field of robotics this is called ‘envelopment’. However, to put an ‘envelope’ around AI-powered machines we need to know some basic things about them which we are often in the dark about. The properties we need to know are the: training data, inputs, functions, outputs, and boundaries. This knowledge is a necessary first step towards the envelopment of AI-powered machines. It is only with this knowledge that we can responsibly regulate, use, and live in a world populated by these machines. (shrink)

A fundamental problem in artificial intelligence is that nobody really knows what intelligence is. The problem is especially acute when we need to consider artificial systems which are significantly different to humans. In this paper we approach this problem in the following way: we take a number of well known informal definitions of human intelligence that have been given by experts, and extract their essential features. These are then mathematically formalised to produce a general measure of intelligence for arbitrary (...) class='Hi'>machines. We believe that this equation formally captures the concept of machine intelligence in the broadest reasonable sense. We then show how this formal definition is related to the theory of universal optimal learning agents. Finally, we survey the many other tests and definitions of intelligence that have been proposed for machines. (shrink)

The question of the imagination is rather like the question Augustine raised with regard to the nature of time. We all seem to know what it involves, yet find it difficult to define. For Descartes, the imagination was simply our faculty for producing a mental image. He distinguished it from the understanding by noting that while the notion of a thousand sided figure was comprehensible—that is, was sufficiently clear and distinct to be differentiated from a thousand and one sided figure—the (...) figure could not be clearly pictured in our mind. The representation of its sides exceeded our powers of imagination.[i] This view of the imagination as our ability to produce a mental image fails, however, to distinguish it from remembering. Let us say that I see an object and then I close my eyes, maintaining the image of the object. Is this imagining or short term memory? What about the case when I recall this image an hour later? Am I imagining or remembering it? Such examples make it clear that imagination, as distinct from memory, implies something more than the ability to produce a mental image. It involves, as Sartre pointed out, a certain attitude towards this image. Engaging in it, we deny its reality. In Sartre’s words, imagination “carries within it a double negation; first, it is the nihilation of the world (since the world is not offering the imagined object as an actual object of perception), secondly, the nihilation of the object of the image (it is posited as not actual) ...” (BN, p. 62). Imagination, then, represents the imagined as nonactual. (shrink)

In this paper the author tries to disentangle some of the problems tied up in John Searle's famous Chinese-room-argument. In a first step to answer the question what it would be for a system to have not only syntax, but also semantics the author gives a brief account of the functioning of the language understanding systems (LUS) so far developed in the framework of AI research thereby making clear that systems like Winograd's SHRDLU are indeed doing little more than mere (...) number crunching. But things would be entirely different, the author argues, if the database of a LUS were built up by the system itself via some perceptual component-at least, if this perceptual component had the capacity to distinguish objects having a certain property F from objects which do not. For in this case the system could store an internal representation of the fact that the object has the property F in its database if and only if the object in fact has that property. And this would be a good basis for calling such a system a genuine LUS. But Searle has objected to a very similar account of J. Fodor that nothing could be further from true language understanding. The reason for this complaint seems to be that Searle holds the view that a true LUS must e.g., know that the word hamburgers refers to hamburgers and that he moreover claims that this knowledge must be explicit or that the system must be aware of the reference of hamburgers to hamburgers. The author argues that this is asking too much. For it seems plausible to say that a system is able to understand e.g., the word hamburger even if it has only implicit knowledge of the fact that hamburger refers to hamburgers in the sense that it has the capacity to tell hamburgers from non hamburgers and the capacity to bring the word hamburger together just with objects of the former kind. (shrink)