The movement of Artificial Intelligence (AI) research from universities to big corporations has had a significant impact on the development of the field. In the past, AI research was primarily conducted in academic institutions, which foster a culture of peer reviewing and collaboration to enhance quality improvements. The growing interest in AI among corporations, especially regarding Machine Learning (ML) technology, has shifted the focus of research from quality to quantity. Corporations have the resources to invest in large-scale ML projects and (...) they are often more interested in fast results than in ensuring that AI algorithms are reliable and safe. This paper proposes that the description of a Darwinian process made by the mathematician and physicist Freeman Dyson can be used to understand the implications of ML research scenario. The context of the Darwinian process—or lack thereof—can be used to draw a parallel between historical and current implications of technology expansion strongly driven by economic interests, making clear the consequences of shifting AI research from academia to industry. Finally, this paper indicates future directions for AI research to create a healthier environment for the evolution of ML technology. (shrink)

Recent years have seen a gradual convergence of seemingly distant research fields over a single goal: understanding and replicating biological intelligence in artifacts. This work presents a general overview on the origin, the state-of-the-art, scientific challenges and the future of Biologically Inspired Cognitive Architecture (BICA) research. Our perspective decomposes the field into the four principal semantic components associated with the BICA challenge that together call for an integration of efforts of researchers across disciplines. Areas and directions of study where new (...) integrated efforts will be primarily needed are summarized. (shrink)

This chapter offers a brief characterization of creativity, followed by a review of some of the reasons people have been skeptical about the possibility of explaining creativity. It surveys some of the recent work on creativity that is naturalistic in the sense that it presumes creativity is natural, as opposed to magical, occult, or supernatural, and is therefore amenable to scientific inquiry. The chapter divides into two categories. The broader category is empirical philosophy, which draws on empirical research while addressing (...) questions that have traditionally been regarded as philosophical. The second category is experimental philosophy, a special branch of empirical philosophy in which experiments are designed by philosophers with the explicit purpose of addressing philosophical questions. This categorization is admittedly somewhat artificial, but it is a familiar one in the prevailing division of intellectual labor, and it will be a useful organizational tool for the chapter. (shrink)

Deep neural networks are said to be opaque, impeding the development of safe and trustworthy artificial intelligence, but where this opacity stems from is less clear. What are the sufficient properties for neural network opacity? Here, I discuss five common properties of deep neural networks and two different kinds of opacity. Which of these properties are sufficient for what type of opacity? I show how each kind of opacity stems from only one of these five properties, and then discuss to (...) what extent the two kinds of opacity can be mitigated by explainability methods. (shrink)

The cerebral cortex is a rich and diverse structure that is the basis of intelligent behavior. One of the deepest mysteries of the function of cortex is that neural processing times are only about one hundred times as fast as the fastest response times for complex behavior. At the very least, this would seem to indicate that the cortex does massive amounts of parallel computation.This paper explores the hypothesis that an important part of the cortex can be modeled as a (...) connectionist computer that is especially suited for parallel problem solving. The connectionist computer uses a special representation, termed value unit encoding, that represents small subsets of parameters in a way that allows parallel access to many different parameter values. This computer can be thought of as computing hierarchies of sensorimotor invariants. The neural substrate can be interpreted as a commitment to data structures and algorithms that compute invariants fast enough to explain the behavioral response times. A detailed consideration of this model has several implications for the underlying anatomy and physiology. (shrink)

The cerebral cortex is a rich and diverse structure that is the basis of intelligent behavior. One of the deepest mysteries of the function of cortex is that neural processing times are only about one hundred times as fast as the fastest response times for complex behavior. At the very least, this would seem to indicate that the cortex does massive amounts of parallel computation.This paper explores the hypothesis that an important part of the cortex can be modeled as a (...) connectionist computer that is especially suited for parallel problem solving. The connectionist computer uses a special representation, termed value unit encoding, that represents small subsets of parameters in a way that allows parallel access to many different parameter values. This computer can be thought of as computing hierarchies of sensorimotor invariants. The neural substrate can be interpreted as a commitment to data structures and algorithms that compute invariants fast enough to explain the behavioral response times. A detailed consideration of this model has several implications for the underlying anatomy and physiology. (shrink)

Technology and Artificial Intelligence, both today and in the near future, are dominated by automated algorithms that combine optimization with models based on the human brain to learn, predict, and even influence the large-scale behavior of human users. Such applications can be understood to be outgrowths of historical trends in industry and academia, yet have far-reaching and even unintended consequences for social and political life around the world. Countries in different parts of the world take different regulatory views for the (...) role and protection of user data, and this will in turn determine the course of development for technology and AI in the near future. (shrink)

This chapter describes the conceptual foundations of cognitive science during its establishment as a science in the 20th century. It is organized around the core ideas of individual agency as its basic explanans and information-processing as its basic explanandum. The latter consists of a package of ideas that provide a mathematico-engineering framework for the philosophical theory of materialism.

This article deals with the links between the enaction paradigm and artificial intelligence. Enaction is considered a metaphor for artificial intelligence, as a number of the notions which it deals with are deemed incompatible with the phenomenal field of the virtual. After explaining this stance, we shall review previous works regarding this issue in terms of artificial life and robotics. We shall focus on the lack of recognition of co-evolution at the heart of these approaches. We propose to explicitly integrate (...) the evolution of the environment into our approach in order to refine the ontogenesis of the artificial system, and to compare it with the enaction paradigm. The growing complexity of the ontogenetic mechanisms to be activated can therefore be compensated by an interactive guidance system emanating from the environment. This proposition does not however, resolve that of the relevance of the meaning created by the machine (sense-making). Such reflections lead us to integrate human interaction into this environment in order to construct relevant meaning in terms of participative artificial intelligence. This raises a number of questions with regards to setting up an enactive interaction. The article concludes by exploring a number of issues, thereby enabling us to associate current approaches with the principles of morphogenesis, guidance, the phenomenology of interactions and the use of minimal enactive interfaces in setting up experiments which will deal with the problem of artificial intelligence in a variety of enaction-based ways. (shrink)

Book Description (Blurb): Cognitive Design for Artificial Minds explains the crucial role that human cognition research plays in the design and realization of artificial intelligence systems, illustrating the steps necessary for the design of artificial models of cognition. It bridges the gap between the theoretical, experimental and technological issues addressed in the context of AI of cognitive inspiration and computational cognitive science. -/- Beginning with an overview of the historical, methodological and technical issues in the field of Cognitively-Inspired Artificial Intelligence, (...) Lieto illustrates how the cognitive design approach has an important role to play in the development of intelligent AI technologies and plausible computational models of cognition. Introducing a unique perspective that draws upon Cybernetics and early AI principles, Lieto emphasizes the need for an equivalence between cognitive processes and implemented AI procedures, in order to realise biologically and cognitively inspired artificial minds. He also introduces the Minimal Cognitive Grid, a pragmatic method to rank the different degrees of biologically and cognitive accuracy of artificial systems in order project and predict their explanatory power with respect to the natural systems taken as source of inspiration. -/- Providing a comprehensive overview of cognitive design principles in constructing artificial minds, this text will be essential reading for students and researchers of artificial intelligence and cognitive science. (shrink)

Deep neural networks (DNNs) are powerful computational models, which generate complex, high-level representations that were missing in previous models of human cognition. By studying these high-level representations, psychologists can now gain new insights into the nature and origin of human high-level vision, which was not possible with traditional handcrafted models. Abandoning DNNs would be a huge oversight for psychological sciences.

Studies comparing movement-related cortical potentials, post-excitatory inhibition after transcranial magnetic brain stimulation, and PET findings in normal controls and patients with cerebellar degeneration demonstrate plasticity of cerebro-cerebellar interactions and hereby support Thach's theory that the cerebellum has the ability to play a role in building behavioral context-response linkages and to build up appropriate responses from simpler constitutive elements, [THACH].

Several target articles in this BBS special issue address the topic of cerebellar and olivary functions, especially as they pertain to motor earning. Another important topic is the neural interaction between the limbic system and the cerebellum during associative learning. In this commentary we present some of our data on olivo-cerebellar and limbic-cerebellar interactions during eyeblink conditioning. [HOUK et al.; SIMPSON et al.; THACH].

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 first issue of Artificial Intelligence and Law journal was published in 1992. This paper offers some commentaries on papers drawn from the Journal’s third decade. They indicate a major shift within Artificial Intelligence, both generally and in AI and Law: away from symbolic techniques to those based on Machine Learning approaches, especially those based on Natural Language texts rather than feature sets. Eight papers are discussed: two concern the management and use of documents available on the World Wide Web, (...) and six apply machine learning techniques to a variety of legal applications. (shrink)

PET studies using classical conditioning paradigms are reported. It is emphasized that PET studies show and not in learning paradigms. The importance of dissociating motor performance and learning deficits in human lesions studies is demonstrated in two exemplary studies. The different role of the cerebellum in adaptation of postural reflexes and learning of complex voluntary arm movements is discussed, [THACH].

For reasons I have never understood, some students of the cerebellum have been unwilling to accept the now overwhelming evidence that the cerebellum exhibits lasting synaptic plasticity and plays an essential role in some forms of learning and memory. With a few exceptions (e.g., target article by SIMPSON et al.) this is no longer the case, as is clear in the excellent target articles on cerebellar LTD and the excellent target review by HOUK et al. [CRÉPEL et al.; HOUR et (...) al.; KANO; LINDEN; SIMPSON et al.; SMITH; VINCENT]. (shrink)

In the behavioral literature on human movement, a distinction is made between the learning of parameters and the learning of new movement forms or topologies. Whereas the target articles by Thach, Smith, and Houk et al. provide evidence for cerebellar involvement in parametrization learning and adaptation, the evidence in favor of its involvement in the generation of new movement patterns is less straightforward. A case is made for focusing more attention on the latter issue in the future. This would directly (...) help to bridge the gap between current neurophysiological approaches to the role of the cerebellum and the behavioral expressions of human motor learning, [HOUK et al.; SMITH; THACH]. (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)

In response to several requests from commentators, an unambiguous definition of time-varying joint stiffness is provided. However, since a variety of different operations can be used to measure stiffness, a problem for quantification admittedly still exists. Several commentaries pointed out the advantage of controlling joint stiffness in optimizing the speed-accuracy trade-off known as Fittss law. The deficit in rapid reciprocal movements and the impact on joint stiffness inhibition caused by cerebellar lesions is clarified here, as the target article was apparently (...) misinterpreted by some readers. In response to the challenge that there is little consensus among cerebellar physiologists, several areas of tacit agreement with other theories of cerebellar function are enumerated. An alternative interpretation of studies showing a transient activation of the cerebellum in motor learning is suggested. Finally, the relationship between the command signals generated by supraspinal centers such as the cerebellum and spinal interneuron networks controlling muscle synergies is discussed. (shrink)

Several themes can be identified in the commentaries. The first is that the climbing fibers may have more than one function; the second is that the climbing fibers provide sensory rather than motor signals. We accept the possibility that climbing fibers may have more than one function consequence(s)’ in the title. Until we know more about the function of the inhibitory input to the inferior olive from the cerebellar nuclei, which are motor structures, we have to keep open the possibility (...) that the climbing fiber signals can be a combination of sensory and motor signals. (shrink)

Researchers studying complex cognition have grown increasingly interested in mapping symbolic cognitive architectures onto subsymbolic brain models. Such a mapping seems essential for understanding cognition under all but the most extreme viewpoints (namely, that cognition consists exclusively of digitally implemented rules; or instead, involves no rules whatsoever). Making this mapping reduces to specifying an interface between symbolic and subsymbolic descriptions of brain activity. To that end, we propose parameterization techniques for building cognitive models as programmable, structured, recurrent neural networks. Feedback (...) strength in these models determines whether their components implement classically subsymbolic neural network functions (e.g., pattern recognition), or instead, logical rules and digital memory. These techniques support the implementation of limited production systems. Though inherently sequential and symbolic, these neural production systems can exploit principles of parallel, analog processing from decision-making models in psychology and neuroscience to explain the effects of brain damage on problem solving behavior. (shrink)

The purpose of this article is to show why consciousness and thought are not manifested in digital computers. Analyzing the rationale for claiming that the formal manipulation of physical symbols in Turing machines would emulate human thought, the article attempts to show why this proved false. This is because the reinterpretation of designation and meaning to accommodate physical symbol manipulation eliminated their crucial functions in human discourse. Words have denotations and intensional meanings because the brain transforms the physical stimuli received (...) from the microworld into a qualitative, macroscopic representation for consciousness. Lacking this capacity as programmed machines, computers have no representations for their symbols to designate and mean. Unlike human beings in which consciousness and thought, with their inherent content, have emerged because of their organic natures, serial processing computers or parallel distributed processing systems, as programmed electrical machines, lack these causal capacities. (shrink)

This paper introduces a special issue of Cognitive Science initiated on the 25th anniversary of the publication of Parallel Distributed Processing (PDP), a two-volume work that introduced the use of neural network models as vehicles for understanding cognition. The collection surveys the core commitments of the PDP framework, the key issues the framework has addressed, and the debates the framework has spawned, and presents viewpoints on the current status of these issues. The articles focus on both historical roots and contemporary (...) developments in learning, optimality theory, perception, memory, language, conceptual knowledge, cognitive control, and consciousness. Here we consider the approach more generally, reviewing the original motivations, the resulting framework, and the central tenets of the underlying theory. We then evaluate the impact of PDP both on the field at large and within specific subdomains of cognitive science and consider the current role of PDP models within the broader landscape of contemporary theoretical frameworks in cognitive science. Looking to the future, we consider the implications for cognitive science of the recent success of machine learning systems called “deep networks”—systems that build on key ideas presented in the PDP volumes. (shrink)

This commentary gives a personal perspective on modeling and modeling developments in cognitive science, starting in the 1950s, but focusing on the author’s personal views of modeling since training in the late 1960s, and particularly focusing on advances since the official founding of the Cognitive Science Society. The range and variety of modeling approaches in use today are remarkable, and for many, bewildering. Yet to come to anything approaching adequate insights into the infinitely complex fields of mind, brain, and intelligent (...) systems, an extremely wide array of modeling approaches is vital and necessary. (shrink)

As an emerging discipline, neuroeconomics faces considerable methodological and practical challenges. In this paper, I suggest that these challenges can be understood by exploring the similarities and dissimilarities between the emergence of neuroeconomics and the emergence of cognitive and computational neuroscience two decades ago. From these parallels, I suggest the major challenge facing theory formation in the neural and behavioural sciences is that of being under-constrained by data, making a detailed understanding of physical implementation necessary for theory construction in neuroeconomics. (...) Rather than following a top-down strategy, neuroeconomists should be pragmatic in the use of available data from animal models, information regarding neural pathways and projections, computational models of neural function, functional imaging and behavioural data. By providing convergent evidence across multiple levels of organization, neuroeconomics will have its most promising prospects of success. (shrink)

This paper analyzes the rapid and unexpected rise of deep learning within Artificial Intelligence and its applications. It tackles the possible reasons for this remarkable success, providing candidate paths towards a satisfactory explanation of why it works so well, at least in some domains. A historical account is given for the ups and downs, which have characterized neural networks research and its evolution from “shallow” to “deep” learning architectures. A precise account of “success” is given, in order to sieve out (...) aspects pertaining to marketing or sociology of research, and the remaining aspects seem to certify a genuine value of deep learning, calling for explanation. The alleged two main propelling factors for deep learning, namely computing hardware performance and neuroscience findings, are scrutinized, and evaluated as relevant but insufficient for a comprehensive explanation. We review various attempts that have been made to provide mathematical foundations able to justify the efficiency of deep learning, and we deem this is the most promising road to follow, even if the current achievements are too scattered and relevant for very limited classes of deep neural models. The authors’ take is that most of what can explain the very nature of why deep learning works at all and even very well across so many domains of application is still to be understood and further research, which addresses the theoretical foundation of artificial learning, is still very much needed. (shrink)

In recent years, the family of algorithms collected under the term ``deep learning'' has revolutionized artificial intelligence, enabling machines to reach human-like performances in many complex cognitive tasks. Although deep learning models are grounded in the connectionist paradigm, their recent advances were basically developed with engineering goals in mind. Despite of their applied focus, deep learning models eventually seem fruitful for cognitive purposes. This can be thought as a kind of biological exaptation, where a physiological structure becomes applicable for a (...) function different from that for which it was selected. In this paper, it will be argued that it is time for cognitive science to seriously come to terms with deep learning, and we try to spell out the reasons why this is the case. First, the path of the evolution of deep learning from the connectionist project is traced, demonstrating the remarkable continuity, and the differences as well. Then, it will be considered how deep learning models can be useful for many cognitive topics, especially those where it has achieved performance comparable to humans, from perception to language. It will be maintained that deep learning poses questions that cognitive sciences should try to answer. One of such questions is the reasons why deep convolutional models that are disembodied, inactive, unaware of context, and static, are by far the closest to the patterns of activation in the brain visual system. (shrink)