Toward a Unified Sub-symbolic Computational Theory of Cognition

Tool mastering today - an interdisciplinary perspective

Tools have coined human life, living conditions, and culture. Recognizing the cognitive architecture underlying tool use would allow us to comprehend its evolution, development, and physiological basis. However, the cognitive underpinnings of tool mastering remain little understood in spite of long-time research in neuroscientific, psychological, behavioral and technological fields. Moreover, the recent transition of tool use to the digital domain poses new challenges for explaining the underlying processes. In this interdisciplinary review, we propose three building blocks of tool mastering: (A) perceptual and motor abilities integrate to tool manipulation knowledge, (B) perceptual and cognitive abilities to functional tool knowledge, and (C) motor and cognitive abilities to means-end knowledge about tool use. This framework allows for integrating and structuring research findings and theoretical assumptions regarding the functional architecture of tool mastering via behavior in humans and non-human primates, brain networks, as well as computational and robotic models. An interdisciplinary perspective also helps to identify open questions and to inspire innovative research approaches. The framework can be applied to studies on the transition from classical to modern, non-mechanical tools and from analogue to digital user-tool interactions in virtual reality, which come with increased functional opacity and sensorimotor decoupling between tool user, tool, and target. By working towards an integrative theory on the cognitive architecture of the use of tools and technological assistants, this review aims at stimulating future interdisciplinary research avenues.

Resourceful Event-Predictive Inference: The Nature of Cognitive Effort

Pursuing a precise, focused train of thought requires cognitive effort. Even more effort is necessary when more alternatives need to be considered or when the imagined situation becomes more complex. Cognitive resources available to us limit the cognitive effort we can spend. In line with previous work, an information-theoretic, Bayesian brain approach to cognitive effort is pursued: to solve tasks in our environment, our brain needs to invest information, that is, negative entropy, to impose structure, or focus, away from a uniform structure or other task-incompatible, latent structures. To get a more complete formalization of cognitive effort, a resourceful event-predictive inference model (REPI) is introduced, which offers computational and algorithmic explanations about the latent structure of our generative models, the active inference dynamics that unfold within, and the cognitive effort required to steer the dynamics-to, for example, purposefully process sensory signals, decide on responses, and invoke their execution. REPI suggests that we invest cognitive resources to infer preparatory priors, activate responses, and anticipate action consequences. Due to our limited resources, though, the inference dynamics are prone to task-irrelevant distractions. For example, the task-irrelevant side of the imperative stimulus causes the Simon effect and, due to similar reasons, we fail to optimally switch between tasks. An actual model implementation simulates such task interactions and offers first estimates of the involved cognitive effort. The approach may be further studied and promises to offer deeper explanations about why we get quickly exhausted from multitasking, how we are influenced by irrelevant stimulus modalities, why we exhibit magnitude interference, and, during social interactions, why we often fail to take the perspective of others into account.

Intuitive physics learning in a deep-learning model inspired by developmental psychology

'Intuitive physics' enables our pragmatic engagement with the physical world and forms a key component of 'common sense' aspects of thought. Current artificial intelligence systems pale in their understanding of intuitive physics, in comparison to even very young children. Here we address this gap between humans and machines by drawing on the field of developmental psychology. First, we introduce and open-source a machine-learning dataset designed to evaluate conceptual understanding of intuitive physics, adopting the violation-of-expectation (VoE) paradigm from developmental psychology. Second, we build a deep-learning system that learns intuitive physics directly from visual data, inspired by studies of visual cognition in children. We demonstrate that our model can learn a diverse set of physical concepts, which depends critically on object-level representations, consistent with findings from developmental psychology. We consider the implications of these results both for AI and for research on human cognition.

The Impact of Action Effects on Infants' Predictive Gaze Shifts for a Non-Human Grasping Action at 7, 11, and 18 Months

During the observation of goal-directed actions, infants usually predict the goal at an earlier age when the agent is familiar (e.g., human hand) compared to unfamiliar (e.g., mechanical claw). These findings implicate a crucial role of the developing agentive self for infants’ processing of others’ action goals. Recent theoretical accounts suggest that predictive gaze behavior relies on an interplay between infants’ agentive experience (top-down processes) and perceptual information about the agent and the action-event (bottom-up information; e.g., agency cues). The present study examined 7-, 11-, and 18-month-old infants’ predictive gaze behavior for a grasping action performed by an unfamiliar tool, depending on infants’ age-related action knowledge about tool-use and the display of the agency cue of producing a salient action effect. The results are in line with the notion of a systematic interplay between experience-based top-down processes and cue-based bottom-up information: Regardless of the salient action effect, predictive gaze shifts did not occur in the 7-month-olds (least experienced age group), but did occur in the 18-month-olds (most experienced age group). In the 11-month-olds, however, predictive gaze shifts occurred only when a salient action effect was presented. This sheds new light on how the developing agentive self, in interplay with available agency cues, supports infants’ action-goal prediction also for observed tool-use actions.

Emergent Goal-Anticipatory Gaze in Infants via Event-Predictive Learning and Inference

From about 7 months of age onward, infants start to reliably fixate the goal of an observed action, such as a grasp, before the action is complete. The available research has identified a variety of factors that influence such goal-anticipatory gaze shifts, including the experience with the shown action events and familiarity with the observed agents. However, the underlying cognitive processes are still heavily debated. We propose that our minds (i) tend to structure sensorimotor dynamics into probabilistic, generative event-predictive, and event boundary predictive models, and, meanwhile, (ii) choose actions with the objective to minimize predicted uncertainty. We implement this proposition by means of event-predictive learning and active inference. The implemented learning mechanism induces an inductive, event-predictive bias, thus developing schematic encodings of experienced events and event boundaries. The implemented active inference principle chooses actions by aiming at minimizing expected future uncertainty. We train our system on multiple object-manipulation events. As a result, the generation of goal-anticipatory gaze shifts emerges while learning about object manipulations: the model starts fixating the inferred goal already at the start of an observed event after having sampled some experience with possible events and when a familiar agent (i.e., a hand) is involved. Meanwhile, the model keeps reactively tracking an unfamiliar agent (i.e., a mechanical claw) that is performing the same movement. We qualitatively compare these modeling results to behavioral data of infants and conclude that event-predictive learning combined with active inference may be critical for eliciting goal-anticipatory gaze behavior in infants.

Towards Strong AI

Strong AI—artificial intelligence that is in all respects at least as intelligent as humans—is still out of reach. Current AI lacks common sense, that is, it is not able to infer, understand, or explain the hidden processes, forces, and causes behind data. Main stream machine learning research on deep artificial neural networks (ANNs) may even be characterized as being behavioristic. In contrast, various sources of evidence from cognitive science suggest that human brains engage in the active development of compositional generative predictive models (CGPMs) from their self-generated sensorimotor experiences. Guided by evolutionarily-shaped inductive learning and information processing biases, they exhibit the tendency to organize the gathered experiences into event-predictive encodings. Meanwhile, they infer and optimize behavior and attention by means of both epistemic- and homeostasis-oriented drives. I argue that AI research should set a stronger focus on learning CGPMs of the hidden causes that lead to the registered observations. Endowed with suitable information-processing biases, AI may develop that will be able to explain the reality it is confronted with, reason about it, and find adaptive solutions, making it Strong AI. Seeing that such Strong AI can be equipped with a mental capacity and computational resources that exceed those of humans, the resulting system may have the potential to guide our knowledge, technology, and policies into sustainable directions. Clearly, though, Strong AI may also be used to manipulate us even more. Thus, it will be on us to put good, far-reaching and long-term, homeostasis-oriented purpose into these machines.

Event-Predictive Cognition: A Root for Conceptual Human Thought

Our minds navigate a continuous stream of sensorimotor experiences, selectively compressing them into events. Event-predictive encodings and processing abilities have evolved because they mirror interactions between agents and objects-and the pursuance or avoidance of critical interactions lies at the heart of survival and reproduction. However, it appears that these abilities have evolved not only to pursue live-enhancing events and to avoid threatening events, but also to distinguish food sources, to produce and to use tools, to cooperate, and to communicate. They may have even set the stage for the formation of larger societies and the development of cultural identities. Research on event-predictive cognition investigates how events and conceptualizations thereof are learned, structured, and processed dynamically. It suggests that event-predictive encodings and processes optimally mediate between sensorimotor processes and language. On the one hand, they enable us to perceive and control physical interactions with our world in a highly adaptive, versatile, goal-directed manner. On the other hand, they allow us to coordinate complex social interactions and, in particular, to comprehend and produce language. Event-predictive learning segments sensorimotor experiences into event-predictive encodings. Once first encodings are formed, the mind learns progressively higher order compositional structures, which allow reflecting on the past, reasoning, and planning on multiple levels of abstraction. We conclude that human conceptual thought may be grounded in the principles of event-predictive cognition constituting its root.

Methodological aspects for cognitive architectures construction: a study and proposal

In the field of Artificial Intelligence (AI), efforts to achieve human-like behavior have taken very different paths through time. Cognitive Architectures (CAs) differentiate from traditional AI approaches, due to their intention to model cognitive and behavioral processes by understanding the brain’s structure and their functionalities in a natural way. However, the development of distinct CAs has not been easy, mainly because there is no consensus on the theoretical basis, assumptions or even purposes for their creation nor how well they reflect human function. In consequence, there is limited information about the methodological aspects to construct this type of models. To address this issue, some initial statements are established to contextualize about the origins and directions of cognitive architectures and their development, which help to outline perspectives, approaches and objectives of this work, supported by a brief study of methodological strategies and historical aspects taken by some of the most relevant architectures to propose a methodology which covers general perspectives for the construction of CAs. This proposal is intended to be flexible, focused on use-case tasks, but also directed by theoretic paradigms or manifestos. A case study between cognitive functions is then detailed, using visual perception and working memory to exemplify the proposal’s assumptions, postulates and binding tools, from their meta-architectural conceptions to validation. Finally, the discussion addresses the challenges found at this stage of development and future work directions.

Hands Ahead in Mind and Motion: Active Inference in Peripersonal Hand Space

According to theories of anticipatory behavior control, actions are initiated by predicting their sensory outcomes. From the perspective of event-predictive cognition and active inference, predictive processes activate currently desired events and event boundaries, as well as the expected sensorimotor mappings necessary to realize them, dependent on the involved predicted uncertainties before actual motor control unfolds. Accordingly, we asked whether peripersonal hand space is remapped in an uncertainty anticipating manner while grasping and placing bottles in a virtual reality (VR) setup. To investigate, we combined the crossmodal congruency paradigm with virtual object interactions in two experiments. As expected, an anticipatory crossmodal congruency effect (aCCE) at the future finger position on the bottle was detected. Moreover, a manipulation of the visuo-motor mapping of the participants’ virtual hand while approaching the bottle selectively reduced the aCCE at movement onset. Our results support theories of event-predictive, anticipatory behavior control and active inference, showing that expected uncertainties in movement control indeed influence anticipatory stimulus processing.

How Do Actions Influence Attitudes? An Inferential Account of the Impact of Action Performance on Stimulus Evaluation

Over the past decade, an increasing number of studies have shown that the performance of specific actions (e.g., approach and avoidance) in response to a stimulus can lead to changes in how that stimulus is evaluated. In contrast to the reigning idea that these effects are mediated by the automatic formation and activation of associations in memory, we describe an inferential account that specifies the inferences underlying the effects and how these inferences are formed. We draw on predictive processing theories to explain the basic processes underlying inferential reasoning and their main characteristics. Our inferential account accommodates past findings, is supported by new findings, and leads to novel predictions as well as concrete recommendations for how action performance can be used to influence real-world behavior.

The Active Inference Approach to Ecological Perception: General Information Dynamics for Natural and Artificial Embodied Cognition

The emerging neurocomputational vision of humans as embodied, ecologically embedded, social agents—who shape and are shaped by their environment—offers a golden opportunity to revisit and revise ideas about the physical and information-theoretic underpinnings of life, mind, and consciousness itself. In particular, the active inference framework (AIF) makes it possible to bridge connections from computational neuroscience and robotics/AI to ecological psychology and phenomenology, revealing common underpinnings and overcoming key limitations. AIF opposes the mechanistic to the reductive, while staying fully grounded in a naturalistic and information-theoretic foundation, using the principle of free energy minimization. The latter provides a theoretical basis for a unified treatment of particles, organisms, and interactive machines, spanning from the inorganic to organic, non-life to life, and natural to artificial agents. We provide a brief introduction to AIF, then explore its implications for evolutionary theory, ecological psychology, embodied phenomenology, and robotics/AI research. We conclude the paper by considering implications for machine consciousness.

Artificial Intelligence and brain

From the start, Kurt Godel observed that computer and brain paradigms were considered on a par by researchers and that researchers had misunderstood his theorems. He hailed with displeasure that the brain transcends computers. In this brief article, we point out that Artificial Intelligence (AI) comprises multitudes of human-made methodologies, systems, and languages, and implemented with computer technology. These advances enhance development in the electron and quantum realms. In the biological realm, animal neurons function, also utilizing electron flow, and are products of evolution. Mirror neurons are an important paradigm in neuroscience research. Moreover, the paradigm shift proposed here - 'hall of mirror neurons' - is a potentially further productive research tactic. These concepts further expand AI and brain research.

You cannot speak and listen at the same time: a probabilistic model of turn-taking

Turn-taking is a preverbal skill whose mastering constitutes an important precondition for many social interactions and joint actions. However, the cognitive mechanisms supporting turn-taking abilities are still poorly understood. Here, we propose a computational analysis of turn-taking in terms of two general mechanisms supporting joint actions: action prediction (e.g., recognizing the interlocutor's message and predicting the end of turn) and signaling (e.g., modifying one's own speech to make it more predictable and discriminable). We test the hypothesis that in a simulated conversational scenario dyads using these two mechanisms can recognize the utterances of their co-actors faster, which in turn permits them to give and take turns more efficiently. Furthermore, we discuss how turn-taking dynamics depend on the fact that agents cannot simultaneously use their internal models for both action (or messages) prediction and production, as these have different requirements-or, in other words, they cannot speak and listen at the same time with the same level of accuracy. Our results provide a computational-level characterization of turn-taking in terms of cognitive mechanisms of action prediction and signaling that are shared across various interaction and joint action domains.

Mario Becomes Cognitive

In line with Allen Newell's challenge to develop complete cognitive architectures, and motivated by a recent proposal for a unifying subsymbolic computational theory of cognition, we introduce the cognitive control architecture SEMLINCS. SEMLINCS models the development of an embodied cognitive agent that learns discrete production rule-like structures from its own, autonomously gathered, continuous sensorimotor experiences. Moreover, the agent uses the developing knowledge to plan and control environmental interactions in a versatile, goal-directed, and self-motivated manner. Thus, in contrast to several well-known symbolic cognitive architectures, SEMLINCS is not provided with production rules and the involved symbols, but it learns them. In this paper, the actual implementation of SEMLINCS causes learning and self-motivated, autonomous behavioral control of the game figure Mario in a clone of the computer game Super Mario Bros. Our evaluations highlight the successful development of behavioral versatility as well as the learning of suitable production rules and the involved symbols from sensorimotor experiences. Moreover, knowledge- and motivation-dependent individualizations of the agents' behavioral tendencies are shown. Finally, interaction sequences can be planned on the sensorimotor-grounded production rule level. Current limitations directly point toward the need for several further enhancements, which may be integrated into SEMLINCS in the near future. Overall, SEMLINCS may be viewed as an architecture that allows the functional and computational modeling of embodied cognitive development, whereby the current main focus lies on the development of production rules from sensorimotor experiences.

In touch with mental rotation: interactions between mental and tactile rotations and motor responses

Although several process models have described the cognitive processing stages that are involved in mentally rotating objects, the exact nature of the rotation process itself remains elusive. According to embodied cognition, cognitive functions are deeply grounded in the sensorimotor system. We thus hypothesized that modal rotation perceptions should influence mental rotations. We conducted two studies in which participants had to judge if a rotated letter was visually presented canonically or mirrored. Concurrently, participants had to judge if a tactile rotation on their palm changed direction during the trial. The results show that tactile rotations can systematically influence mental rotation performance in that same rotations are favored. In addition, the results show that mental rotations produce a response compatibility effect: clockwise mental rotations facilitate responses to the right, while counterclockwise mental rotations facilitate responses to the left. We conclude that the execution of mental rotations activates cognitive mechanisms that are also used to perceive rotations in different modalities and that are associated with directional motor control processes.