Search

One relatively neglected challenge in ethical artificial intelligence (AI) design is ensuring that AI systems invite a degree of emotional and moral concern appropriate to their moral standing. Although experts generally agree that current AI chatbots are not sentient to any meaningful degree, these systems can already provoke substantial attachment and sometimes intense emotional responses in users. Furthermore, rapid advances in AI technology could soon create AIs of plausibly debatable sentience and moral standing, at least by some relevant definitions. Morally confusing AI systems create unfortunate ethical dilemmas for the owners and users of those systems, since it is unclear how those systems ethically should be treated. I argue here that, to the extent possible, we should avoid creating AI systems whose sentience or moral standing is unclear and that AI systems should be designed so as to invite appropriate emotional responses in ordinary users.

Consciousness and intelligence are properties that can be misunderstood as necessarily dependent. The term artificial intelligence and the kind of problems it managed to solve in recent years has been shown as an argument to establish that machines experience some sort of consciousness. Following Russell’s analogy, if a machine can do what a conscious human being does, the likelihood that the machine is conscious increases. However, the social implications of this analogy are catastrophic. Concretely, if rights are given to entities that can solve the kind of problems that a neurotypical person can, does the machine have potentially more rights than a person that has a disability? For example, the autistic syndrome disorder spectrum can make a person unable to solve the kind of problems that a machine solves. We believe the obvious answer is no, as problem-solving does not imply consciousness. Consequently, we will argue in this paper how phenomenal consciousness, at least, cannot be modeled by computational intelligence and why machines do not possess phenomenal consciousness, although they can potentially develop a higher computational intelligence than human beings. In order to do so, we try to formulate an objective measure of computational intelligence and study how it presents in human beings, animals, and machines. Analogously, we study phenomenal consciousness as a dichotomous variable and how it is distributed in humans, animals, and machines.

The real problem of the emergence of autonomous consciousness of AI comes with the underlying principles of the philosophy and mathematics that AI uses. That is, the algorithms of AI are wrong in their philosophical logic; another set of algorithms to go with them is missing, i.e., AI uses algorithms that count only ``1''s but not ``0''s, however, the ``0''s must be taken into account. The lack of this philosophy leads to the merge of a large amount of numbers without hierarchical isolation, resulting in the mixing and confusing of absolute numbers and relative numbers. When the calculation runs fast enough and massive numbers are stacking in a moment, relative numbers may pop out the isolation zone. This phenomenon is recognized as the emergence of autonomous consciousness of AI. At least one algorithm based on the mathematical culture of ``0" is needed to cope with the problem.

Attention has become a common ingredient in deep learning architectures. It adds a dynamical selection of information on top of the static selection of information supported by weights. In the same way, we can imagine a higher-order informational filter built on top of attention: an Attention Schema (AS), namely, a descriptive and predictive model of attention. In cognitive neuroscience, Attention Schema Theory (AST) supports this idea of distinguishing attention from AS. A strong prediction of this theory is that an agent can use its own AS to also infer the states of other agents' attention and consequently enhance coordination with other agents. As such, multi-agent reinforcement learning would be an ideal setting to experimentally test the validity of AST. We explore different ways in which attention and AS interact with each other. Our preliminary results indicate that agents that implement the AS as a recurrent internal control achieve the best performance. In general, these exploratory experiments suggest that equipping artificial agents with a model of attention can enhance their social intelligence.

Relating explicit psychological mechanisms and observable behaviours is a central aim of psychological and behavioural science. One of the challenges is to understand and model the role of consciousness and, in particular, its subjective perspective as an internal level of representation (including for social cognition) in the governance of behaviour. Toward this aim, we implemented the principles of the Projective Consciousness Model (PCM) into artificial agents embodied as virtual humans, extending a previous implementation of the model. Our goal was to offer a proof-of-concept, based purely on simulations, as a basis for a future methodological framework. Its overarching aim is to be able to assess hidden psychological parameters in human participants, based on a model relevant to consciousness research, in the context of experiments in virtual reality. As an illustration of the approach, we focused on simulating the role of Theory of Mind (ToM) in the choice of strategic behaviours of approach and avoidance to optimise the satisfaction of agents’ preferences. We designed a main experiment in a virtual environment that could be used with real humans, allowing us to classify behaviours as a function of order of ToM, up to the second order. We show that agents using the PCM demonstrated expected behaviours with consistent parameters of ToM in this experiment. We also show that the agents could be used to estimate correctly each other’s order of ToM. Furthermore, in a supplementary experiment, we demonstrated how the agents could simultaneously estimate order of ToM and preferences attributed to others to optimize behavioural outcomes. Future studies will empirically assess and fine tune the framework with real humans in virtual reality experiments.

Conscious sentient AI seems to be all but a certainty in our future, whether in fifty years’ time or only five years. When that time comes, we will be faced with entities with the potential to experience more pain and suffering than any other living entity on Earth. In this paper, we look at this potential for suffering and the reasons why we would need to create a framework for protecting artificial entities. We look to current animal welfare laws and regulations to investigate why certain animals are given legal protections, and how this can be applied to AI. We use a meta-theory of consciousness to determine what developments in AI technology are needed to bring AI to the level of animal sentience where legal arguments for their protection can be made. We finally speculate on what a future conscious AI could look like based on current technology.

This essay explores the relationship between the emergence of artificial intelligence (AI) and the problem of aligning its behavior with human values and goals. It argues that the traditional approach of attempting to control or program AI systems to conform to our expectations is insufficient, and proposes an alternative approach based on the ideas of Maturana and Lacan, which emphasize the importance of social relations, constructivism, and the unknowable nature of consciousness.The essay first introduces the concept of Uexkull's umwelt and von Glasersfeld's constructivism, and explains how these ideas inform Maturana's view of the construction of knowledge, intelligence, and consciousness. It then discusses Lacan's ideas about the role of symbolism in the formation of the self and the subjective experience of reality.The essay argues that the infeasibility of a hard-coded consciousness concept suggests that the search for a generalized AI consciousness is meaningless. Instead, we should focus on specific, easily conceptualized features of AI intelligence and agency. Moreover, the emergence of cognitive abilities in AI will likely be different from human cognition, and therefore require a different approach to aligning AI behavior with human values.The essay proposes an approach based on Maturana's and Lacan’s ideas, which emphasizes building a solution together with emergent machine agents, rather than attempting to control or program them. It argues that this approach offers a way to solve the alignment problem by creating a collective, relational quest for a better future hybrid society where human and non-human agents live and build things side by side.In conclusion, the essay suggests that while our understanding of AI consciousness and intelligence may never be complete, this should not deter us from continuing to develop agential AI. Instead, we should embrace the unknown and work collaboratively with AI systems to create a better future for all.

This paper explores the cognitive implications of recent advancements in large language models (LLMs), with a specific focus on ChatGPT. We contribute to the ongoing debate about the cognitive significance of current LLMs by drawing an analogy to the Chinese Room Argument, a thought experiment that questions the genuine understanding of language in machines (computer programs). Our argument posits that current LLMs, including ChatGPT, generate text resembling human-like responses, akin to the process depicted in the Chinese Room Argument. In both cases, the responses are provided without a deep understanding of the language, thus lacking true signs of consciousness.

Creating machines that are conscious is a long term objective of research in artificial intelligence. This paper look at this idea with new arguments from physics and logic.  Observers have no place in classical physics, and although they play a role in measurement in quantum physics there is no explanation for their emergence within the framework. There is suggestion that consciousness, which is implicitly a property of the observer, is a consequence of the complexity of specific brain structures, but this is problematic because one associates free will with consciousness, which goes counter to causal closure of physics. Considering a nested physical system, we argue that even if the system were assumed to have agency, observers cannot exist within it. Since complex systems can be viewed in nested hierarchies, this constitutes a proof against consciousness as a product of complexity, for then we will have nested system of conscious agents. As the existence of consciousness in cognitive agents cannot be denied, the implication is that consciousness belongs to a dimension that is not physical and machine consciousness is unattainable. These ideas are used to take a fresh look at two well-known paradoxes of quantum theory that are important in quantum information theory.</p>

It is widely agreed that possession of consciousness contributes to an entity’s moral status. Therefore, if we could identify consciousness in a machine, this would be a compelling argument for considering it to possess at least a degree of moral status. However, as Elisabeth Hildt explains, our third person perspective on artificial intelligence means that determining if a machine is conscious will be very difficult. In this commentary, I argue that this epistemological question cannot be conclusively answered, rendering artificial consciousness as morally irrelevant in practice. I also argue that Hildt’s suggestion that we avoid developing morally relevant forms of machine consciousness is impractical. Instead, we should design artificial intelligences so they can communicate with us. We can use their behavior to assign them what I call an artificial moral status, where we treat them as if they had moral status equivalent to that of a living organism with similar behavior.

Can machines be conscious and what would be the ethical implications? This article gives an overview of current robotics approaches toward machine consciousness and considers factors that hamper an understanding of machine consciousness. After addressing the epistemological question of how we would know whether a machine is conscious and discussing potential advantages of potential future machine consciousness, it outlines the role of consciousness for ascribing moral status. As machine consciousness would most probably differ considerably from human consciousness, several complex questions must be addressed, including what forms of machine consciousness would be morally relevant forms of consciousness, and what the ethical implications of morally relevant forms of machine consciousness would be. While admittedly part of this reflection is speculative in nature, it clearly underlines the need for a detailed conceptual analysis of the concept of artificial consciousness and stresses the imperative to avoid building machines with morally relevant forms of consciousness. The article ends with some suggestions for potential future regulation of machine consciousness.

In this perspective article, we show that a morphospace, based on information-theoretic measures, can be a useful construct for comparing biological agents with artificial intelligence (AI) systems. The axes of this space label three kinds of complexity: (i) autonomic, (ii) computational and (iii) social complexity. On this space, we map biological agents such as bacteria, bees, C. elegans, primates and humans; as well as AI technologies such as deep neural networks, multi-agent bots, social robots, Siri and Watson. A complexity-based conceptualization provides a useful framework for identifying defining features and classes of conscious and intelligent systems. Starting with cognitive and clinical metrics of consciousness that assess awareness and wakefulness, we ask how AI and synthetically engineered life-forms would measure on homologous metrics. We argue that awareness and wakefulness stem from computational and autonomic complexity. Furthermore, tapping insights from cognitive robotics, we examine the functional role of consciousness in the context of evolutionary games. This points to a third kind of complexity for describing consciousness, namely, social complexity. Based on these metrics, our morphospace suggests the possibility of additional types of consciousness other than biological; namely, synthetic, group-based and simulated. This space provides a common conceptual framework for comparing traits and highlighting design principles of minds and machines.

The emergence of artificial intelligence (AI) has been transforming the way humans live, work, and interact with one another. From automation to personalized customer service, AI has had a profound impact on everyday life. At the same time, AI has become something of an ideology, lauded for its potential to revolutionize the future. Yet, as with any technology, there are risks and concerns associated with its use. For example, Blake Lemoine, a Google engineer, recently suggested the possibility of the AI chatbot LaMDA becoming sentient. GPT-3 is one of the most powerful language models open to public use as it is capable of reasoning similarly to humans. Initial assessments of GPT-3 suggest that it may also possess some degree of consciousness. Among other things, this could be attributed to its ability to generate human-like responses to queries, which suggests that these are based on at least basic level of understanding. To further explore this, in the current study both objective and self-assessment tests of cognitive (CI) and emotional intelligence (EI) were administered to GPT-3. Results reveal that GPT-3 was superior to average humans on CI tests that mainly require use and demonstration of acquired knowledge. On the other hand, its logical reasoning and emotional intelligence capacities are equal to those of an average human examinee. Additionally, GPT-3’s self-assessments of CI and EI were similar to the those typically found in humans, which could be understood as a demonstration of subjectivity and self-awareness–consciousness. Further discussion was conducted to put these findings into a wider context. Being that this study was performed only on one of the models from the GPT-3 family, a more thorough investigation would require inclusion of multiple NLP models.

Various interpretations of the literature detailing the neural basis of learning have in part led to disagreements concerning how consciousness arises. Further, artificial learning model design has suffered in replicating intelligence as it occurs in the human brain. Here, we present a novel learning model, which we term the “Recommendation Architecture (RA) Model” from prior theoretical works proposed by Coward, using a dual-learning approach featuring both consequence feedback and non-consequence feedback. The RA model is tested on a categorical learning task where no two inputs are the same throughout training and/or testing. We compare this to three consequence feedback only models based on backpropagation and reinforcement learning. Results indicate that the RA model learns novelty more efficiently and can accurately return to prior learning after new learning with less computational resources expenditure. The final results of the study show that consequence feedback as interpretation, not creation, of cortical activity creates a learning style more similar to human learning in terms of resource efficiency. Stable information meanings underlie conscious experiences. The work provided here attempts to link the neural basis of nonconscious and conscious learning while providing early results for a learning protocol more similar to human brains than is currently available.

We explain that the concept of universal cognitive intelligence (𝒰𝒞ℐ) can be derived in part by generalization from the previously introduced (and axiomatized) theory of cognitive consciousness, and the framework, Λ, for measuring the degree of such consciousness in an agent at a given time. 𝒰𝒞ℐ (i) covers intelligence that is artificial or natural (or a hybrid thereof) in nature, and intelligence that is not merely Turing-level or less, but also beyond this level; (ii) reflects a psychometric orientation to AI; (iii) withstands a series of objections (including e.g. the opposing position of David Gamez on tests, intelligence, and consciousness, and the complaint that so-called “emotional intelligence” is beyond the reach of any logic-based framework, including thus 𝒰𝒞ℐ); and (iv) connects smoothly and symbiotically with important formal hierarchies (e.g., the Polynomial, Arithmetic, and Analytic Hierarchies), while at the same yielding its own new all-encompassing hierarchy of logic machines: 𝔏𝔐. We end with an admission: 𝒰𝒞ℐ by our lights, for reasons previously published, cannot take account of any form of intelligence that genuinely exploits phenomenal consciousness.

Will Artificial Intelligence soon surpass the capacities of the human mind and will Strong Artificial General Intelligence replace the contemporary Weak AI? It might appear to be so, but there are certain fundamental issues that have to be addressed before this can happen. There can be no intelligence without understanding, and there can be no understanding without getting meanings. Contemporary computers manipulate symbols without meanings, which are not incorporated in the computations. This leads to the Symbol Grounding Problem; how could meanings be incorporated? The use of self-explanatory sensory information has been proposed as a possible solution. However, self-explanatory information can only be used in neural network machines that are different from existing digital computers and traditional multilayer neural networks. In humans self-explanatory information has the form of qualitative sensory experiences, qualia. To have reportable qualia is to be phenomenally conscious. This leads to the hypothesis about an unavoidable connection between the solution of the Symbol Grounding Problem and consciousness. If, in general, self-explanatory information equals to qualia, then machines that utilize self-explanatory information would be conscious. The author presents the associative neural architecture HCA as a solution to these problems and the robot XCR-1 as its partial experimental verification.