Current advances in research, development and application of artificial intelligence systems have yielded a far-reaching discourse on AI ethics. In consequence, a number of ethics guidelines have been released in recent years. These guidelines comprise normative principles and recommendations aimed to harness the “disruptive” potentials of new AI technologies. Designed as a semi-systematic evaluation, this paper analyzes and compares 22 guidelines, highlighting overlaps but also omissions. As a result, I give a detailed overview of the field of AI ethics. Finally, (...) I also examine to what extent the respective ethical principles and values are implemented in the practice of research, development and application of AI systems—and how the effectiveness in the demands of AI ethics can be improved. (shrink)

In this paper, I investigate the phenomenology of awakening in Chinese Zen Buddhism. In this tradition, to awaken is to ‘see your true nature’. In particular, the two aspects of awakening are: (1) seeing that the nature of one’s self or mind is empty or void and (2) an erasing of the usual (though merely apparent) boundary between subject and object. In the early Zen tradition, there are many references to awakening as chopping off your head, not having eyes, nose (...) and tongue, and seeing your ‘Original Face’. These references bear a remarkable resemblance to an approach to awakening developed by Douglas Harding. I will guide the reader through a series of Harding’s first-person experiments which investigate the gap where you cannot see your own head. I will endeavour to show that these methods, although radically different from traditional meditation techniques, result in an experience with striking similarities to Zen accounts of awakening, in particular, as experiencing oneself as empty or void and yet totally united with the given world. The repeatability and apparent reliability of these first-person methods opens up a class of awakening experience to empirical investigation and has the potential to provide new insights into nondual traditions. (shrink)

Phenomenal objectivism explains perceptual phenomenal character by reducing it to an awareness of mind-independent objects, properties, and relations. A challenge for this view is that there is a sense in which a distant tree looks smaller than a closer tree even when they are the same objective size (perceptual size variation). The dual content view is a popular objectivist account in which such experiences are explained by my objective spatial relation to the tree, in particular visual angle (perspectival size). I (...) describe a series of first-person experiments for investigating size experience. I use a ruler as a first-person method for operationalising perspectival size (Experiment 1). I use the corridor illusion (Experiment 2), outlining one’s head in the mirror (Experiment 3), and outlining the size of objects on glass (Experiment 4) to show a phenomenal difference in size for items in different depth contexts, despite being identical in visual angle. These finding demonstrate that visual angle cannot account for these spatial experiences. Psychological evidence provides further support for the thesis that subjects do not experience visual angle when depth information is present. Together this evidence supports the hypothesis that perceptual size variation cannot be accounted for by visual angle, hence undermining a plausible version of the dual content theory. This outcome, combined with problems raised by alternative objectivist accounts of size variation, provides support for a subjectivist account of size experience. (shrink)

Erwin Schrödinger holds a prominent place in the history of science primarily due to his crucial role in the development of quantum physics. What is perhaps lesser known are his insights into subject-object duality, consciousness and mind. He documented himself that these were influenced by the Upanishads, a collection of ancient Hindu spiritual texts. Central to his thoughts in this area is that Mind is only One and there is no separation between subject and object. This chapter aims to bridge (...) Schrödinger’s view on One Mind with the teachings of Dōgen, a twelfth century Zen master. This bridge is formed by addressing the question of how time relates to One Mind, and subject-object duality. Schrödinger describes the experience of One Mind to be like a timeless now, whereas subject-object duality involves a linear continuum of time. We show how these differing positions are unified in the notion of ‘absolute present’, which was put forward in the philosophy of Nishida Kitarō (1871–1945). In addition, we argue that it is in this notion of absolute present that the views of Schrödinger, Dōgen and Nishida meet. (shrink)

Many ethics initiatives have stipulated sets of principles and standards for good technology development in the AI sector. However, several AI ethics researchers have pointed out a lack of practical realization of these principles. Following that, AI ethics underwent a practical turn, but without deviating from the principled approach. This paper proposes a complementary to the principled approach that is based on virtue ethics. It defines four “basic AI virtues”, namely justice, honesty, responsibility and care, all of which represent specific (...) motivational settings that constitute the very precondition for ethical decision making in the AI field. Moreover, it defines two “second-order AI virtues”, prudence and fortitude, that bolster achieving the basic virtues by helping with overcoming bounded ethicality or hidden psychological forces that can impair ethical decision making and that are hitherto disregarded in AI ethics. Lastly, the paper describes measures for successfully cultivating the mentioned virtues in organizations dealing with AI research and development. (shrink)

Conceptual combination performs a fundamental role in creating the broad range of compound phrases utilised in everyday language. This article provides a novel probabilistic framework for assessing whether the semantics of conceptual combinations are compositional, and so can be considered as a function of the semantics of the constituent concepts, or not. While the systematicity and productivity of language provide a strong argument in favor of assuming compositionality, this very assumption is still regularly questioned in both cognitive science and philosophy. (...) Additionally, the principle of semantic compositionality is underspecifi ed, which means that notions of both "strong" and "weak" compositionality appear in the literature. Rather than adjudicating between different grades of compositionality, the framework presented here contributes formal methods for determining a clear dividing line between compositional and non-compositional semantics. In addition, we suggest that the distinction between these is contextually sensitive. Compositionality is equated with a joint probability distribution modeling how the constituent concepts in the combination are interpreted. Marginal selectivity is introduced as a pivotal probabilistic constraint for the application of the Bell/CH and CHSH systems of inequalities. Non-compositionality is equated with a failure of marginal selectivity, or violation of either system of inequalities in the presence of marginal selectivity. This means that the conceptual combination cannot be modeled in a joint probability distribution, the variables of which correspond to how the constituent concepts are being interpreted. The formal analysis methods are demonstrated by applying them to an empirical illustration of twenty-four non-lexicalised conceptual combinations. (shrink)

I am aware of the red and orange autumn leaves. Am I aware of my awareness of the leaves? Not so according to many philosophers. By contrast, many meditative traditions report an experience of awareness itself. I argue that such a pure awareness experience must have a non-sensory phenomenal character. I use Douglas Harding’s first-person experiments for assisting in recognising pure awareness. In particular, I investigate the gap where one cannot see one’s head. This is not a mere gap because (...) I seem to be looking from here. Critically, I claim, the experience of looking from here has a non-sensory phenomenal character. I argue that this sense of being aware cannot be reduced to egocentric visual-spatial relations nor the viewpoint because it continues when I close my eyes. Neither is a multisensory origin sufficient to explain why I seem to be at this central point rather than elsewhere. Traditionally, claims of a pure awareness experience have been restricted to highly trained individuals in very restricted circumstances. The innovation of Harding’s approach is that it reliably isolates a candidate for pure awareness using methods which can be replicated at any time. (shrink)

Certain research strands can yield “forbidden knowledge”. This term refers to knowledge that is considered too sensitive, dangerous or taboo to be produced or shared. Discourses about such publication restrictions are already entrenched in scientific fields like IT security, synthetic biology or nuclear physics research. This paper makes the case for transferring this discourse to machine learning research. Some machine learning applications can very easily be misused and unfold harmful consequences, for instance, with regard to generative video or text synthesis, (...) personality analysis, behavior manipulation, software vulnerability detection and the like. Up till now, the machine learning research community embraces the idea of open access. However, this is opposed to precautionary efforts to prevent the malicious use of machine learning applications. Information about or from such applications may, if improperly disclosed, cause harm to people, organizations or whole societies. Hence, the goal of this work is to outline deliberations on how to deal with questions concerning the dissemination of such information. It proposes a tentative ethical framework for the machine learning community on how to deal with forbidden knowledge and dual-use applications. (shrink)

The current “AI Summer” is marked by scientific breakthroughs and economic successes in the fields of research, development, and application of systems with artificial intelligence. But, aside from the great hopes and promises associated with artificial intelligence, there are a number of challenges, shortcomings and even limitations of the technology. For one, these challenges arise from methodological and epistemological misconceptions about the capabilities of artificial intelligence. Secondly, they result from restrictions of the social context in which the development of applications (...) of machine learning is embedded. And third, they are a consequence of current technical limitations in the development and use of artificial intelligence. The paper intends to provide an overview of current challenges which the research and development of applications in the field of artificial intelligence and machine learning have to face, whereas all three mentioned areas are to be further explored in this paper. (shrink)

Industry involvement in the machine learning (ML) community seems to be increasing. However, the quantitative scale and ethical implications of this influence are rather unknown. For this purpose, we have not only carried out an informed ethical analysis of the field, but have inspected all papers of the main ML conferences NeurIPS, CVPR, and ICML of the last 5 years—almost 11,000 papers in total. Our statistical approach focuses on conflicts of interest, innovation, and gender equality. We have obtained four main (...) findings. (1) Academic–corporate collaborations are growing in numbers. At the same time, we found that conflicts of interest are rarely disclosed. (2) Industry papers amply mention terms that relate to particular trending machine learning topics earlier than academia does. (3) Industry papers are not lagging behind academic papers with regard to how often they mention keywords that are proxies for social impact considerations. (4) Finally, we demonstrate that industry papers fall short of their academic counterparts with respect to the ratio of gender diversity. We believe that this work is a starting point for an informed debate within and outside of the ML community. (shrink)

Consider the concept combination ‘pet human’. In word association experiments, human subjects produce the associate ‘slave’ in relation to this combination. The striking aspect of this associate is that it is not produced as an associate of ‘pet’, or ‘human’ in isolation. In other words, the associate ‘slave’ seems to be emergent. Such emergent associations sometimes have a creative character and cognitive science is largely silent about how we produce them. Departing from a dimensional model of human conceptual space, this (...) article will explore concept combinations, and will argue that emergent associations are a result of abductive reasoning within conceptual space, that is, below the symbolic level of cognition. A tensor-based approach is used to model concept combinations allowing such combinations to be formalized as interacting quantum systems. Free association norm data is used to motivate the underlying basis of the conceptual space. It is shown by analogy how some concept combinations may behave like quantum-entangled particles. Two methods of analysis were presented for empirically validating the presence of non-separable concept combinations in human cognition. One method is based on quantum theory and another based on comparing a joint probability distribution with another distribution based on a separability assumption using a chi-square goodness-of-fit test. Although these methods were inconclusive in relation to an empirical study of bi-ambiguous concept combinations, avenues for further refinement of these methods are identified. (shrink)

In the original publication of this article, the Table 1 has been published in a low resolution. Now a larger version of Table 1 is published in this correction. The publisher apologizes for the error made during production.

In the original publication of this article, the Table 1 has been published in a low resolution. Now a larger version of Table 1 is published in this correction. The publisher apologizes for the error made during production.

Machine behavior that is based on learning algorithms can be significantly influenced by the exposure to data of different qualities. Up to now, those qualities are solely measured in technical terms, but not in ethical ones, despite the significant role of training and annotation data in supervised machine learning. This is the first study to fill this gap by describing new dimensions of data quality for supervised machine learning applications. Based on the rationale that different social and psychological backgrounds of (...) individuals correlate in practice with different modes of human–computer-interaction, the paper describes from an ethical perspective how varying qualities of behavioral data that individuals leave behind while using digital technologies have socially relevant ramification for the development of machine learning applications. The specific objective of this study is to describe how training data can be selected according to ethical assessments of the behavior it originates from, establishing an innovative filter regime to transition from the big data rationale n = all to a more selective way of processing data for training sets in machine learning. The overarching aim of this research is to promote methods for achieving beneficial machine learning applications that could be widely useful for industry as well as academia. (shrink)

The term complementarity plays a central role in quantum physics, not least in various approaches to defining entanglement and the conditions for its occurrence. It has, however, been used in a variety of ways by different authors, denoting different concepts and relationships. Here we describe and clarify some of them and analyze the role they play with respect to the phenomenon of entanglement. Based on these considerations we discuss the recently proposed system-theoretical generalization of the concepts entanglement and complementarity (Atmanspacher (...) et al. in Found Phys 32(3):379–406, 2002; von Lucadou et al. in J Conscious Stud 14(4):50–74, 2007; Filk and Römer in Axiomathes 21(2):211–220, 2011; Walach and Von Stillfried in Axiomathes 21(2): 185–209, 2011). We hope that a clarification regarding the specific meaning of these terms can be useful to the growing engagement with this interesting hypothesis and its critical investigation. (shrink)

The term complementarity plays a central role in quantum physics, not least in various approaches to defining entanglement and the conditions for its occurrence. It has, however, been used in a variety of ways by different authors, denoting different concepts and relationships. Here we describe and clarify some of them and analyze the role they play with respect to the phenomenon of entanglement. Based on these considerations we discuss the recently proposed system-theoretical generalization of the concepts entanglement and complementarity (Atmanspacher (...) et al. in Found Phys 32(3):379–406, 2002; von Lucadou et al. in J Conscious Stud 14(4):50–74, 2007; Filk and Römer in Axiomathes 21(2):211–220, 2011; Walach and Von Stillfried in Axiomathes 21(2): 185–209, 2011). We hope that a clarification regarding the specific meaning of these terms can be useful to the growing engagement with this interesting hypothesis and its critical investigation. (shrink)

P. Clodius’ lex frumentaria of 58 BC tends to be viewed as a populist law that greatly exacerbated, if not largely created, the problems associated with the provision of grain in the 50s. Yet the law was intended to alleviate the high costs due to increased demand created by Cato’s lex frumentaria of 62. Moreover, it was not conceived as a one-shot, populist solution, but was part of a larger strategy aimed at mitigating the problem of supply. In ways hitherto (...) unrealised, Clodius’ annexation of Cyprus and the consular exemption of Delos from taxation formed part of a wider debate on tackling the root cause of high costs and unreliable supply: the resurgence of piracy in the eastern Mediterranean. (shrink)

Due to the technology of machine learning, new breakthroughs are currently being achieved with constant regularity. By using machine learning techniques, computer applications can be developed and used to solve tasks that have hitherto been assumed not to be solvable by computers. If these achievements consider applications that collect and process personal data, this is typically perceived as a threat to information privacy. This paper aims to discuss applications from both fields of personality and image analysis. These applications are often (...) criticized by reference to the protection of privacy. This paper critically questions this approach. Instead of solely using the concept of privacy to address the risks of machine learning, it is increasingly necessary to consider and implement ethical anti-discrimination concepts, too. In many ways, informational privacy requires individual information control. However, not least because of machine learning technologies, information control has become obsolete. Hence, societies need stronger anti-discrimination tenets to counteract the risks of machine learning. (shrink)

While first-person methods are essential for a science of consciousness, it is controversial what form these methods should take and whether any such methods are reliable. I propose that first-person experiments are a reliable method for investigating conscious experience. I outline the history of these methods and describe their characteristics. In particular, a first-person experiment is an intervention on a subject's experience in which independent variables are manipulated, extraneous variables are held fixed, and in which the subject makes a phenomenal (...) judgement about the target experience of the investigation. I examine historical and contemporary examples of first-person experiments: Mariotte’s demonstration of the visual blind spot, Kanizsa’s subjective contours, the Tse Illusion, and investigations of the non-uniform resolution of the visual field. I discuss the role that phenomenal contrast plays in these methods, and how they overcome typical introspective errors. I argue that their intersubjective repeatability is an important factor in their scientific status, however, it is not the only factor. That they control for extraneous factors and confounds is another factor which sets them apart from pseudoscience (e.g., the perception of auras), and hence another reason for classifying them as genuine experiments. Furthermore, by systematically mapping out the structure of visual experience, these methods make scientific progress. Praises of such first-person experimental approaches may not always be sung by philosophers and psychologists, but they continue to flourish as respectable scientific methods nevertheless. (shrink)

We introduce the Bounded Axiom A Forcing Axiom . It turns out that it is equiconsistent with the existence of a regular ∑2-correct cardinal and hence also equiconsistent with BPFA. Furthermore we show that, if consistent, it does not imply the Bounded Proper Forcing Axiom.

Various geoengineering technologies that would deliberately alter the climate system have been proposed as a way to alleviate risks of global warming. Technologies that would shield incoming sunlight to cool the planet, so called solar radiation management, are particularly controversial. Considering insights from social studies of simulation modeling and research on expectations in science and technology, I argue that climate modeling has a central role in producing visions of SRM. I draw upon an empirical analysis of scientific research on SRM (...) to examine how a creative play with technological ideas becomes possible through climate modeling. This enables scientists to project and study environmental impacts of speculative SRM methods in virtual experiments and to develop and refine ideas for adjusting sunlight. Hence, while climate models are used to improve scientific understandings of climate system behavior and to anticipate possible environmental impacts of SRM, they also become inventive tools, allowing scientists to envision novel ways of climate control and optimization. Given the importance of simulation studies to knowledge production on SRM, I critically reflect on the challenges that arise when visions about an engineered climate future are first and foremost produced in climate simulations. (shrink)

In this paper, I argue for a version of panpsychist idealism on first-person experiential grounds. As things always appear in my field of consciousness, there is prima facie empirical support for idealism. Furthermore, by assuming that all things correspond to a conscious perspective or perspectives (i.e., panpsychism), realism about the world is arguably safeguarded without the need to appeal to God (as per Berkeley’s idealism). Panpsychist idealism also has a phenomenological advantage over traditional panpsychist views as it does not commit (...) perceptual experience to massive error by denying that perceived colours are properties of things. Finally, I argue that the subject combination problem for panpsychism has been motivated by the problematic assumption that consciousness is in things. Thinking about subject combination from the first-person perspective is fruitful for reframing the subject combination problem and for seeing how subjects could potentially combine for the idealist. (shrink)

Fairness in machine learning (ML) is an ever-growing field of research due to the manifold potential for harm from algorithmic discrimination. To prevent such harm, a large body of literature develops new approaches to quantify fairness. Here, we investigate how one can divert the quantification of fairness by describing a practice we call “fairness hacking” for the purpose of shrouding unfairness in algorithms. This impacts end-users who rely on learning algorithms, as well as the broader community interested in fair AI (...) practices. We introduce two different categories of fairness hacking in reference to the established concept of p-hacking. The first category, intra-metric fairness hacking, describes the misuse of a particular metric by adding or removing sensitive attributes from the analysis. In this context, countermeasures that have been developed to prevent or reduce p-hacking can be applied to similarly prevent or reduce fairness hacking. The second category of fairness hacking is inter-metric fairness hacking. Inter-metric fairness hacking is the search for a specific fair metric with given attributes. We argue that countermeasures to prevent or reduce inter-metric fairness hacking are still in their infancy. Finally, we demonstrate both types of fairness hacking using real datasets. Our paper intends to serve as a guidance for discussions within the fair ML community to prevent or reduce the misuse of fairness metrics, and thus reduce overall harm from ML applications. (shrink)

I am aware of the tree and its leaves, but am I aware of my awareness of these things? When I try to introspect my awareness, I just find myself attending to objects and their properties. This observation is known as the ‘transparency of experience’. On the other hand, I seem to directly know that I am aware. Given the first observation, it is not clear how I know that I am aware. Fred Dretske thought that the problem was so (...) acute that he issued the challenge of answering ‘How do you know that you are not a zombie?’ I propose that a view found in the Advaita Vedanta, that awareness is self luminous, reconciles these two observations. I understand self-luminosity as the thesis that: (1) I am implicitly aware of my awareness and (2) I am phenomenally aware of a distinct phenomenal character of my awareness. In support of the first claim, that I apparently only attend to objects in the world when introspecting perceptual experience, suggests that I do not know my awareness explicitly, but rather that I must know it implicitly. In support of the second claim, I argue that the mere fact that I am perceptually conscious is not sufficient to allow me to know that I am perceptually conscious. In particular, the qualities I am perceptually aware of do not tell me that I aware of them, rather they just seem to be properties of objects. I also assess whether strategies for responding to Cartesian sceptical scenarios can be employed against Dretske’s consciousness scepticism. I argue that these strategies either fail to distinguish me from a zombie or they do not adequately describe my epistemic situation. By contrast to other accounts, if awareness has its own distinct phenomenal character, then it cannot be considered to be a prima facie property of the world, hence the self-luminosity of awareness provides a plausible account of how I know that I’m not a zombie. (shrink)

Kants Religionsphilosophie by Ch. A. Thilo. This book is a reproduction of the original book published in 1905 and may have some imperfections such as marks or hand-written notes.

Hume famously denied that he could experience the self. Most subsequent philosophers have concurred with this finding. I argue that if the subject is to function as a bearer of experience it must (1) lack sensory qualities in itself to be compatible with bearing sensory qualities and (2) be single so that it can unify experience. I use Douglas Harding’s first-person experiments to investigate the visual gap where one cannot see one’s own head. I argue that this open space conforms (...) to the above criteria and hence is consistent with being the subject. I respond to the objection that this location is merely a lack of visual experience. I argue that this space also encompasses sound and touch properties and hence functions as a bearer for other sensory modalities. These first-person findings provide prima facie support for the view that the subject is a thin bearer of experience. (shrink)

Measures and theories of information abound, but there are few formalised methods for treating the contextuality that can manifest in different information systems. Quantum theory provides one possible formalism for treating information in context. This paper introduces a quantum inspired model of the human mental lexicon. This model is currently being experimentally investigated and we present a preliminary set of pilot data suggesting that concept combinations can indeed behave non-separably.

This article explores the potential ways in which artificial intelligence (AI) can support veganism, a lifestyle that aims to promote the protection of animals and also avoids the consumption of animal products for environmental and health reasons. The first part of the article discusses the technical requirements for utilizing AI technologies in the mentioned field. The second part provides an overview of potential use cases, including facilitating consumer change with the help of AI, technologically augmenting undercover investigations in factory farms, (...) raising the efficiency of nongovernment organizations promoting plant-based lifestyles, and so forth. The article acknowledges that the deployment of AI should not happen in a “solutionist” manner, meaning to always consider nontechnical means for achieving desired outcomes. However, it is important for organizations promoting veganism to realize the potential of modern data-driven tools and to merge and share their data to reach common goals. (shrink)

Objectives Practitioners in contemplative traditions commonly report experiencing an awareness that is distinct from sensory objects, thoughts, and emotions (“awareness itself”). They also report experiences of a void or underlying silence that is closely associated with this awareness. Subjects who carry out the Headless Way exercises frequently report an experience of emptiness or void at the same time as other contents (void-like experiences). The goals of this study were to (1) assess the reliability of these methods in eliciting the recognition (...) of awareness and void-like experiences in participants who had no prior exposure to these techniques, (2) investigate the prevalence of these experiences in these tasks, and (3) to differentiate these experiences from closely related and potential precursor experiences. -/- Method Twenty adults participated in in-depth individual interviews in which they were guided through the Headless Way exercises. A thematic analysis was conducted on the interview transcripts. -/- Results Twelve of the participants reported a void-like experience, and five participants reported an experience of awareness itself. These experiences were respectively categorized as subsets of the more general categories of perceptual absences and the sense of not being person-like. Another novel finding was the real-time reports of awareness and void-like experiences during the exercises. -/- Conclusions Our findings provide preliminary evidence that the Headless Way exercises can effectively induce experiences of emptiness and awareness in participants without prior experience. The findings suggest that such experiences can be elicited outside of a traditional meditation context, including in non-meditators. Furthermore, the experience of not being person-like and of perceptual absences may be precursors and more general forms of recognizing awareness itself and the void-like nature of the mind. (shrink)

Eric Schwitzgebel (2011) argues that phenomenal judgments are in general less reliable than perceptual judgments. This paper distinguishes two versions of this unreliability thesis. The process unreliability thesis says that unreliability in phenomenal judgments is due to faulty domain-specific mechanisms involved in producing these judgments, whereas the statistical unreliability thesis says that it is simply a matter of higher numbers of errors. Against the process unreliability thesis, I argue that the main errors and limitations in making phenomenal judgments can be (...) accounted for by domain-general factors: attention, working memory limits and conceptualization. As these factors are shared with the production of perceptual judgments, errors in phenomenal judgments are not due to faulty domain-specific processes. Furthermore, this account defends phenomenal judgments against general scepticism by providing criteria for distinguishing between reliable and unreliable phenomenal judgments. (shrink)