Search

Following arguments put forward in my book (Why red doesn’t sound like a bell: understanding the feel of consciousness. Oxford University Press, New York, USA, 2011), this article takes a pragmatic, scientist’s point of view about the concepts of consciousness and “feel”, pinning down what people generally mean when they talk about these concepts, and then investigating to what extent these capacities could be implemented in non-biological machines. Although the question of “feel”, or “phenomenal consciousness” as it is called by some philosophers, is generally considered to be the “hard” problem of consciousness, the article shows that by taking a “sensorimotor” approach, the difficulties can be overcome. What remains to account for are the notions of so-called “access consciousness” and the self. I claim that though they are undoubtedly very difficult, these are not logically impossible to implement in robots.

The potential for the near-future development of two technologies — artificial forms of intelligence, as well as the ability to "upload" human minds into artificial forms — raises several ethical questions regarding the proper treatment and understanding of these artificial minds. The crux of the dilemma is whether or not such creations should be accorded the same rights we currently grant humans, and this question seems to hinge upon whether they will exhibit their own "subjectivity", or internal viewpoints. Recognizing this as the essential factor yields some ethical guidance, but these issues need further exploration before such technologies become available.

The main motivation for this work is to investigate the advantages provided by machine consciousness, while in the control of software agents. In order to pursue this goal, we developed a cognitive architecture, with different levels of machine consciousness, targeting the control of artificial creatures. As a standard guideline, we applied cognitive neuroscience concepts to incrementally develop the cognitive architecture, following the evolutionary steps taken by the animal brain. The triune brain theory proposed by MacLean, together with Arrabale's "ConsScale", serve as roadmaps to achieve each developmental stage, while iCub — a humanoid robot and its simulator — serve as a platform for the experiments. A completely codelet-based system "Core" has been implemented, serving the whole architecture.

The Terasem Mind Uploading Experiment is a multi-decade test of the comparability of single person actual human consciousness as assessed by expert psychological review of their digitized interactions with same person purported human consciousness as assessed by expert psychological interviews of personality software that draws upon a database comprised of the original actual person's digitized interactions. The experiment is based upon a hypothesis that the paper analyzes for its conformance with scientific testability in accordance with the criteria set forth by Karl Popper. Strengths and weaknesses of both the hypothesis and the experiment are assessed in terms of other tests of digital consciousness, scientific rigor and good clinical practices. Recommendations for improvement include stronger parametrization of endpoint assessment and better attention to compliance with informed consent in the event there is emergence of software-based consciousness.

We answer the question raised by the title by developing a neural architecture for the attention control system in animals in a hierarchical manner, following what we conjecture is an evolutionary path. The resulting evolutionary model (based on CODAM at the highest level) and answer to the question allow us to consider both different forms of consciousness as well as how machine consciousness could itself possess a variety of forms.

In this paper, the notion of super-intelligence (or "AI++", as Chalmers has termed it) is considered in the context of machine consciousness (MC) research. Suppose AI++ were to come about, would real MC have then also arrived, "for free"? (I call this the "drop-out question".) Does the idea tempt you, as an MC investigator? What are the various positions that might be adopted on the issue of whether an AI++ would necessarily (or with strong likelihood) be a conscious AI++? Would a conscious super-intelligence also be a super-consciousness? (Indeed, what meaning might be attached to the notion of "super-consciousness"?) What ethical and social consequences might be drawn from the idea of conscious super-AIs or from that of artificial super-consciousness? And what implications does this issue have for technical progress on MC in a pre-AI++ world? These and other questions are considered.

In this article, we propose the Conscious-Emotional Learning Tutoring System technology, a biologically plausible cognitive agent based on human brain functions. This agent is capable of learning and remembering events and any related information such as corresponding procedures, stimuli and their emotional valences. In our model, emotions play a role in the encoding and remembering of events. This allows the agent to improve its behaviour or by remembering previously selected behaviours which were influenced by its emotional mechanism. Moreover, the architecture incorporates a realistic memory consolidation process based on a data mining algorithm.

Most philosophers of mind follow Thomas Nagel and hold that subjective experiences are characterised by the fact that there is “something it is like” to have them. Philosophers of mind have sometimes speculated that ordinary people endorse, perhaps implicitly, this conception of subjective experiences. Some recent findings cast doubt on this view.

We present a two-level model of concurrent communicating systems (CCS) to serve as a basis for machine consciousness. A language implementing threads within logic programming is first introduced. This high-level framework allows for the definition of abstract processes that can be executed on a virtual machine. We then look for a possible grounding of these processes into the brain. Towards this end, we map abstract definitions (including logical expressions representing compiled knowledge) into a variant of the π-calculus. We illustrate this approach through a series of examples extending from a purely reactive behavior to patterns of consciousness.

Scientific behavior is used as a benchmark to examine the truth status of computationalism (COMP) as a law of nature. A COMP-based artificial scientist is examined from three simple perspectives to see if they shed light on the truth or falsehood of COMP through its ability or otherwise, to deliver authentic original science on the a priori unknown like humans do. The first perspective (A) looks at the handling of ignorance and supports a claim that COMP is "trivially true" or "pragmatically false" in the sense that you can simulate a scientist if you already know everything, which is a state that renders the simulation possible but pointless. The second scenario (B) is more conclusive and unusual in that it reveals that the COMP scientist can never propose/debate that COMP is a law of nature. This marked difference between the human and the artificial scientist in this single, very specific circumstance, means that COMP cannot be true as a general claim. The third scenario (C) examines the artificial scientist's ability to do science on itself/humans to uncover the "law of nature" which results in itself. This scenario reveals that a successful test for scientific behavior by a COMP-based artificial scientist supports a claim that COMP is true. Such a test is quite practical and can be applied to an artificial scientist based on any design principle, not merely COMP. Scenario (C) also reveals a practical example of the COMP scientist's inability to handle informal systems (in the form of liars), which further undermines COMP. Overall, the result is that COMP is false, with certainty in one very specific, critical place. This lends support to the claims (i) that artificial general intelligence will not succeed based on COMP principles, and (ii) computationally enacted abstract models of human cognition will never create a mind.

The concept of qualia poses a central problem in the framework of consciousness studies. Despite it being a controversial issue even in the study of human consciousness, we argue that qualia can be complementarily studied using artificial cognitive architectures. In this work we address the problem of defining qualia in the domain of artificial systems, providing a model of “artificial qualia”. Furthermore, we partially apply the proposed model to the generation of visual qualia using the cognitive architecture CERA-CRANIUM, which is modeled after the global workspace theory of consciousness. It is our aim to define, characterize and identify artificial qualia as direct products of a simulated conscious perception process. Simple forms of the apparent motion effect are used as the basis for a preliminary experimental setting focused on the simulation and analysis of synthetic visual experience. In contrast with the study of biological brains, the inspection of the dynamics and transient inner states of the artificial cognitive architecture can be performed effectively, thus enabling the detailed analysis of covert and overt percepts generated by the system when it is confronted with specific visual stimuli. The observed states in the artificial cognitive architecture during the simulation of apparent motion effects are used to discuss the existence of possible analogous mechanisms in human cognition processes.

A novel theory of reflective consciousness, will and self is presented, based on modeling each of these entities using self-referential mathematical structures called hypersets. Pattern theory is used to argue that these exotic mathematical structures may meaningfully be considered as parts of the minds of physical systems, even finite computational systems. The hyperset models presented are hypothesized to occur as patterns within the ”moving bubble of attention” of the human brain and any brainlike AI system. They appear to be compatible with both panpsychist and materialist views of consciousness, and probably other views as well.

Self-aware individuals are more likely to consider whether their actions are appropriate in terms of public self-consciousness, and to use that information to execute behaviors that match external standards and/or expectations. The learning concepts through which individuals monitor themselves have generally been overlooked by artificial intelligence researchers. Here we report on our attempt to integrate a self-awareness mechanism into an agent's learning architecture. Specifically, we describe (a) our proposal for a self-aware agent model that includes an external learning mechanism and internal cognitive capacity with super-ego and ego characteristics; and (b) our application of a version of the iterated prisoner's dilemma representing conflicts between the public good and private interests to analyze the effects of self-awareness on an agent's individual performance and cooperative behavior. Our results indicate that self-aware agents that consider public self-consciousness utilize rational analysis in a manner that promotes cooperative behavior and supports faster societal movement toward stability. We found that a small number of self-aware agents are sufficient for improving social benefits and resolving problems associated with collective irrational behaviors.

This paper addresses the problem of human–computer interactions when the computer can interpret and express a kind of human-like behavior, offering natural communication. A conceptual framework for incorporating emotions with rationality is proposed. A model of affective social interactions is described. The model utilizes the SAIBA framework, which distinguishes among several stages of processing of information. The SAIBA framework is extended, and a model is realized in human behavior detection, human behavior interpretation, intention planning, attention tracking behavior planning, and behavior realization components. Two models of incorporating emotions with rationality into a virtual artifact are presented. The first one uses an implicit implementation of emotions. The second one has an explicit realization of a three-layered model of emotions, which is highly interconnected with other components of the system. Details of the model with implicit implementation of emotional behavior are shown as well as evaluation methodology and results. Discussions about the extended model of an agent are given in the final part of the paper.

Shanahan's work admirably and convincingly supports Baars' global workspace by means of plausible and updated neural models. Yet little of his work is related with the issue of consciousness as phenomenal experience. He focuses his effort mostly on the behavioral correlates of consciousness like autonomy, flexibility, and information integration. Moreover, although the importance of embodiment and situated cognition is emphasized, most of the conceptual tools suggested (dynamic systems, complex networks, global workspace) require the external world only during their development. Leaving aside the issue of phenomenal experience, the book fleshes out a convincing and thought-provoking model for many aspects of conscious behaviour.

A brain model based on glial-neuronal interactions is proposed. Glial-neuronal synaptic units are interpreted as elementary reflection mechanisms, called proemial synapses. In glial networks (syncytia), cyclic intentional programs are generated, interpreted as auto-reflective intentional programming. Both types of reflection mechanisms are formally described and may be implementable in a robot brain. Based on the logic of acceptance and rejection, the robot is capable of rejecting irrelevant environmental information, showing at least a "touch" of subjective behavior. Since reflective intentional programming generates both relevant and irrelevant structures already within the brain, ontological gaps arise which must be integrated. In the human brain, the act of self-reference may exert a holistic function enabling self-consciousness. However, since the act of self-reference is a mysterious function not experimentally testable in brain research, it cannot be implemented in a robot brain. Therefore, the creation of self-conscious robots may never be possible. Finally, some philosophical implications are discussed.

Despite many efforts, there are no computational models of consciousness that can be used to design conscious intelligent machines. This is mainly attributed to available definitions of consciousness being human centered, vague, and incomplete. Through a biological analysis of consciousness and concept of machine intelligence, we propose a physical definition of consciousness with the hope to model it in intelligent machines. We propose a computational model of consciousness driven by competing motivations, goals, and attention-switching. We propose a concept of mental saccades that is useful for explaining the attention-switching and focusing mechanism from computational perspective. Finally, we compare our model with other computational models of consciousness.