The Turing Test (TT), the Chinese Room Argument (CRA), and the Symbol Grounding Problem (SGP) are about the question “can machines think?” We propose to look at these approaches to Artificial Intelligence (AI) by showing that they all address the possibility for Artificial Agents (AAs) to generate meaningful information (meanings) as we humans do. The initial question about thinking machines is then reformulated into “can AAs generate meanings like humans do?” We correspondingly present the TT, the CRA and the SGP (...) as being about generation of human-like meanings. We model and address such possibility by using the Meaning Generator System (MGS) where a system submitted to an internal constraint generates a meaning in order to satisfy the constraint. The system approach of the MGS allows comparing meaning generations in animals, humans and AAs. The comparison shows that in order to have AAs capable of generating human-like meanings, we need the AAs to carry human constraints. And transferring human constraints to AAs raises concerns coming from the unknown natures of life and human mind which are at the root of human constraints. Implications for the TT, the CRA and the SGP are highlighted. It is shown that designing AAs capable of thinking like humans needs an understanding about the natures of life and human mind that we do not have today. Following an evolutionary approach, we propose as a first entry point an investigation about the possibility for extending a “stay alive” constraint into AAs. Ethical concerns are raised from the relations between human constraints and human values. Continuations are proposed. (This paper is an extended version of the proceedings of an AISB/IACAP 2012 presentation). (shrink)

The Chinese room argument has presented a persistent headache in the search for Artificial Intelligence. Since it first appeared in the literature, various interpretations have been made, attempting to understand the problems posed by this thought experiment. Throughout all this time, some researchers in the Artificial Intelligence community have seen Symbol Grounding as proposed by Harnad as a solution to the Chinese room argument. The main thesis in this paper is that although related, these two issues present different problems in (...) the framework presented by Harnad himself. The work presented here attempts to shed some light on the relationship between John Searle’s intentionality notion and Harnad’s Symbol Grounding Problem. (shrink)

Abstract: The core aim of this special issue is to present the philosophy of information as a way of doing philosophy, to focus on the contributions of Luciano Floridi to that area, and most important, to stimulate the debate on the most distinctive and controversial views he has defended in that context. This introduction contains a description of the philosophy of information, a discussion of two common misconceptions about the scope and the ambition of the philosophy of information, and a (...) brief overview of the essays in the issue. (shrink)

The paper presents a paradoxical feature of computational systems that suggests that computationalism cannot explain symbol grounding. If the mind is a digital computer, as computationalism claims, then it can be computing either over meaningful symbols or over meaningless symbols. If it is computing over meaningful symbols its functioning presupposes the existence of meaningful symbols in the system, i.e. it implies semantic nativism. If the mind is computing over meaningless symbols, no intentional cognitive processes are available prior to symbol grounding. (...) In this case, no symbol grounding could take place since any grounding presupposes intentional cognitive processes. So, whether computing in the mind is over meaningless or over meaningful symbols, computationalism implies semantic nativism. (shrink)

This article is the second step in our research into the Symbol Grounding Problem (SGP). In a previous work, we defined the main condition that must be satisfied by any strategy in order to provide a valid solution to the SGP, namely the zero semantic commitment condition (Z condition). We then showed that all the main strategies proposed so far fail to satisfy the Z condition, although they provide several important lessons to be followed by any new proposal. Here, we (...) develop a new solution of the SGP. It is called praxical in order to stress the key role played by the interactions between the agents and their environment. It is based on a new theory of meaning—Action-based Semantics (AbS)—and on a new kind of artificial agents, called two-machine artificial agents (AM²). Thanks to their architecture, AM2s implement AbS, and this allows them to ground their symbols semantically and to develop some fairly advanced semantic abilities, including the development of semantically grounded communication and the elaboration of representations, while still respecting the Z condition. (shrink)

This paper investigates the prospects of Rodney Brooks’ proposal for AI without representation. It turns out that the supposedly characteristic features of “new AI” (embodiment, situatedness, absence of reasoning, and absence of representation) are all present in conventional systems: “New AI” is just like old AI. Brooks proposal boils down to the architectural rejection of central control in intelligent agents—Which, however, turns out to be crucial. Some of more recent cognitive science suggests that we might do well to dispose of (...) the image of intelligent agents as central representation processors. If this paradigm shift is achieved, Brooks’ proposal for cognition without representation appears promising for full-blown intelligent agents—Though not for conscious agents. (shrink)

Why should moral philosophers, moral psychologists, and machine ethicists care about computational complexity? Debates on whether artificial intelligence (AI) can or should be used to solve problems in ethical domains have mainly been driven by what AI can or cannot do in terms of human capacities. In this paper, we tackle the problem from the other end by exploring what kind of moral machines are possible based on what computational systems can or cannot do. To do so, we analyze normative (...) ethics through the lens of computational complexity. First, we introduce computational complexity for the uninitiated reader and discuss how the complexity of ethical problems can be framed within Marr’s three levels of analysis. We then study a range of ethical problems based on consequentialism, deontology, and virtue ethics, with the aim of elucidating the complexity associated with the problems themselves (e.g., due to combinatorics, uncertainty, strategic dynamics), the computational methods employed (e.g., probability, logic, learning), and the available resources (e.g., time, knowledge, learning). The results indicate that most problems the normative frameworks pose lead to tractability issues in every category analyzed. Our investigation also provides several insights about the computational nature of normative ethics, including the differences between rule- and outcome-based moral strategies, and the implementation-variance with regard to moral resources. We then discuss the consequences complexity results have for the prospect of moral machines in virtue of the trade-off between optimality and efficiency. Finally, we elucidate how computational complexity can be used to inform both philosophical and cognitive-psychological research on human morality by advancing the moral tractability thesis. (shrink)

This chapter addresses two questions. First, if knowledge is accounted information, how are we supposed (to apply this analysis in order) to understand perceptual knowledge and knowledge by testimony? In the first part of the chapter, I articulate an answer in terms of a re-interpretation of perception and testimony as data providers rather than full-blown cases of knowledge. Second, if perception and testimony are correctly understood as data providers, how are we supposed (to apply this analysis in order) to understand (...) the semantic value of the data provided by such processes? In the second part of the chapter, I argue in favour of a constructionist hypothesis about how data may become meaningful for human cognitive agents through a process of repurposing of natural data/signals. The conclusion of the chapter is that human agents are natural-born data hackers. (shrink)

The goal of the paper is to develop and propose a general model of the state space of AI. Given the breathtaking progress in AI research and technologies in recent years, such conceptual work is of substantial theoretical interest. The present AI hype is mainly driven by the triumph of deep learning neural networks. As the distinguishing feature of such networks is the ability to self-learn, self-learning is identified as one important dimension of the AI state space. Another dimension is (...) recognized as generalization, the possibility to go over from specific to more general types of problems. A third dimension is semantic grounding. Our overall analysis connects to a number of known foundational issues in the philosophy of mind and cognition: the blockhead objection, the Turing test, the symbol grounding problem, the Chinese room argument, and use theories of meaning. It shall finally be argued that the dimension of grounding decomposes into three sub-dimensions. And the dimension of self-learning turns out as only one of a whole range of “self-x-capacities” (based on ideas of organic computing) that span the self-x-subspace of the full AI state space. (shrink)

Fifty years of effort in artificial intelligence (AI) and the formalization of legal reasoning have produced both successes and failures. Considerable success in organizing and displaying evidence and its interrelationships has been accompanied by failure to achieve the original ambition of AI as applied to law: fully automated legal decision-making. The obstacles to formalizing legal reasoning have proved to be the same ones that make the formalization of commonsense reasoning so difficult, and are most evident where legal reasoning has to (...) meld with the vast web of ordinary human knowledge of the world. Underlying many of the problems is the mismatch between the discreteness of symbol manipulation and the continuous nature of imprecise natural language, of degrees of similarity and analogy, and of probabilities. (shrink)

Artificial intelligence research and regulation seek to balance the benefits of innovation against any potential harms and disruption. However, one unintended consequence of the recent surge in AI research is the potential re-orientation of AI technologies to facilitate criminal acts, term in this article AI-Crime. AIC is theoretically feasible thanks to published experiments in automating fraud targeted at social media users, as well as demonstrations of AI-driven manipulation of simulated markets. However, because AIC is still a relatively young and inherently (...) interdisciplinary area—spanning socio-legal studies to formal science—there is little certainty of what an AIC future might look like. This article offers the first systematic, interdisciplinary literature analysis of the foreseeable threats of AIC, providing ethicists, policy-makers, and law enforcement organisations with a synthesis of the current problems, and a possible solution space. (shrink)

There is a wealth of resources— ideas, insights, discoveries, inventions, traditions, cultures, languages, arts, religions, sciences, narratives, stories, poems, customs and norms, music and songs, games and personal experiences, and advertisements—that we produce, curate, consume, transmit, and inherit as humans. This wealth, which I define as semantic capital, gives meaning to, and makes sense of, our own existence and the world surrounding us. It defines who we are and enables humans to develop an individual and social life. This paper discusses (...) the shift from analog to digital semantic capital, and the extent to which this might affect the semanticisation of our identities, our lives and our realities. -/- . (shrink)

This book reports on the results of the third edition of the premier conference in the field of philosophy of artificial intelligence, PT-AI 2017, held on November 4 - 5, 2017 at the University of Leeds, UK. It covers: advanced knowledge on key AI concepts, including complexity, computation, creativity, embodiment, representation and superintelligence; cutting-edge ethical issues, such as the AI impact on human dignity and society, responsibilities and rights of machines, as well as AI threats to humanity and AI safety; (...) and cutting-edge developments in techniques to achieve AI, including machine learning, neural networks, dynamical systems. The book also discusses important applications of AI, including big data analytics, expert systems, cognitive architectures, and robotics. It offers a timely, yet very comprehensive snapshot of what is going on in the field of AI, especially at the interfaces between philosophy, cognitive science, ethics and computing. (shrink)

Meanings are present everywhere in our environment and within ourselves. But these meanings do not exist by themselves. They are associated to information and have to be created, to be generated by agents. The Meaning Generator System (MGS) has been developed on a system approach to model meaning generation in agents following an evolutionary perspective. The agents can be natural or artificial. The MGS generates meaningful information (a meaning) when it receives information that has a connection with an internal constraint (...) to which the agent is submitted. The generated meaning is to be used by the agent to implement actions aimed at satisfying the constraint. We propose here to highlight some characteristics of the MGS that could be related to items of philosophy of information. (shrink)

The symbol grounding problem (SGP) proposed by Stevan Harnad in 1990, originates from Searle’s “Chinese Room Argument” and refers to the problem of how a pure symbolic system acquires its meaning. While many solutions to this problem have been proposed, all of them have encountered inconsistencies to different extents. A recent approach for resolving the problem is to divide the SGP into hard and easy problems echoing the distinction between hard and easy problems for resolving the enigma of consciousness. This (...) however turns out not to be an ideal strategy: Everything related to consciousness that cannot be well-explained by present theories can be categorized as a hard problem which as a consequence would doom the SGP to irresolvability. We therefore argue that the SGP can be regarded as a general problem of how an AI system can have intentionality, and develop a theoretical direction for its solution. (shrink)

Floridi’s infosphere consisting of informational reality is estimated and delineated by introducing the new notion of Typ-Ken, an undifferentiated amalgam of type and token that can be expressed as either type or token dependent on contingent ontological commitment. First, we elaborate Floridi’s system, level of abstraction (LoA), model, and structure scheme, which is proposed to reconcile ontic with epistemic structural reality, and obtain the duality of type and token inherited in the relationship between LoA and model. While we focus on (...) the ontological commitment that can negotiate and emphasize the discrepancy between type and token, we show that most research focusing on different hierarchical layers (type and token) has converged onto the flattened perspective of type and token to clarify the role of ontological commitment. We propose the idea of Typ-Ken to avoid this flattening. In addition, we elaborate Taddeo and Floridi’s criticism of approaches to the symbol grounding problem and show that their criticism based on the zero semantical commitment condition is specific to the agents affected by the flattening. We then take bird flocks as a metaphor for the living infosphere as if it had one mind. The knowledge recently yielded by image analysis indicates that it is difficult for previous flock models based on agents equipped with the flattening of local (token) and global (type) to show the special feature of a flock of the scale-free proportion (SFP). We show here that agents based on the Typ-Ken can reveal the SFP, which is the essential characteristic of the living infosphere. In other words, Typ-Ken is an expression for informational reality (LoA and model with ontological commitment) that can reveal seamless local and global interactions in the functioning of a flock or infosphere. (shrink)

This paper addresses one of the fundamental problems of the philosophy of information: How does semantic information emerge within the underlying dynamics of the world?—the dynamical semantic information problem. It suggests that the canonical approach to semantic information that defines data before meaning and meaning before use is inadequate for pre-cognitive information media. Instead, we should follow a pragmatic approach to information where one defines the notion of information system as a special kind of purposeful system emerging within the underlying (...) dynamics of the world and define semantic information as the currency of the system. In this way, systems operating with semantic information can be viewed as patterns in the dynamics—semantic information is a dynamical system phenomenon of highly organized systems. In the simplest information systems, the syntax, semantics, and pragmatics of the information medium are co-defined. It proposes a new more general theory of information semantics that focuses on the interface role of the information states in the information system—the interface theory of meaning. Finally, with the new framework, it addresses the debate between weakly semantic and strongly semantic accounts of information, siding with the strongly semantic view because the pragmatic account developed here is a better generalization of it. (shrink)

Moving from a behavioral definition of intelligence, which describes it as the ability to adapt to the surrounding environment and deal effectively with new situations (Anastasi, 1986), this paper explains to what extent the performance obtained by ChatGPT in the linguistic domain can be considered as intelligent behavior and to what extent they cannot. It also explains in what sense the hypothesis of decoupling between cognitive and problem-solving abilities, proposed by Floridi (2017) and Floridi and Chiriatti (2020) should be interpreted. (...) The problem of symbolic grounding (Harnad, 1990) is then addressed to show the problematic relationship between ChatGPT and the natural environment, and thus the impossibility for it to understand the symbols it manipulates. To explain the reasons why ChatGPT does not succeed in this task, an investigation is carried out and a possible solution to the problem in the artificial domain is proposed by making a comparison with the natural ability of living beings to ground their own meanings from some basic cognitive-sensory aspects, which, it is explained, are directly related to the emergence of self-awareness in humans. Thus, the question is raised whether a possible and concrete solution to the Symbol Grounding Problem would involve in the artificial domain the development of cognitive abilities fully comparable to those of humans. Finally, I explain the difficulties that would have to be overcome before such a level could be reached, since human cognitive capacities are intimately linked to intersubjectivity and intercorporeality. (shrink)

Luciano Floridi’s informational structural realism takes a constructionist attitude towards the problems of epistemology and metaphysics, but the question of the nature of the semantical component of his view remains vexing. In this paper, I propose to dispense with the semantical component of ISR completely. I outline a Syntactical version of ISR. The unified entropy-based framework of information has been adopted as the groundwork of SISR. To establish its realist component, SISR should be able to dissolve the latching problem. We (...) have to be able to account for the informational structures–reality relationship in the absence of the standard semantical resources. The paper offers a pragmatic solution to the latching problem. I also take pains to account for the naturalistic plausibility of this solution by grounding it in the recent computational neuroscience of the predictive coding and the free energy principle. (shrink)

The approach the majority of neuroscientists take to the question of how consciousness is generated, it is probably fair to say, is to ignore it. Although there are active research programs looking at correlates of consciousness, and explorations of informational properties of what might be relevant neural ensembles, the tacitly implied mechanism of consciousness in these approaches is that it somehow just happens. This reliance on a “magical emergence” of consciousness does not address the “objectively unreasonable” proposition that elements that (...) have no attributes or properties that can be said to relate to consciousness somehow aggregate to produce it. Neuroscience has furnished evidence that neurons are fundamental to consciousness; at the fine and gross scale, aspects of our conscious experience depend on specific patterns of neural activity – in some way, the connectivity of neurons computes the features of our experience. So how do we get from knowing that some specific configurations of cells produce consciousness to understanding why this would be the case? Behind the voltages and currents electrophysiologists measure is a staggeringly complex system of electromagnetic fields – these are the fundamental physics of neurons and glia in the brain. The brain is entirely made of electromagnetism phenomena from the level of the atoms up. The EM field literally manifests the computations, or signaling, or information processing/activities performed by connected cellular ensembles that generate a 1st-person perspective. An investigation into the EM field at the cellular scale provides the possibility of identifying the outward signs of a mechanism in fundamental terms, as opposed to merely describing the correlates of our mental abstractions of it. (shrink)

The objective of this thesis is to present a naturalised metaphysics of information, or to naturalise information, by way of deploying a scientific metaphysics according to which contingency is privileged and a-priori conceptual analysis is excluded (or at least greatly diminished) in favour of contingent and defeasible metaphysics. The ontology of information is established according to the premises and mandate of the scientific metaphysics by inference to the best explanation, and in accordance with the idea that the primacy of physics (...) constraint accommodates defeasibility of theorising in physics. This metametaphysical approach is used to establish a field ontology as a basis for an informational structural realism. This is in turn, in combination with information theory and specifically mathematical and algorithmic theories of information, becomes the foundation of what will be called a source ontology, according to which the world is the totality of information sources. Information sources are to be understood as causally induced configurations of structure that are, or else reduce to and/or supervene upon, bounded (including distributed and non-contiguous) regions of the heterogeneous quantum field (all quantum fields combined) and fluctuating vacuum, all in accordance with the above-mentioned quantum field-ontic informational structural realism (FOSIR.) Arguments are presented for realism, physicalism, and reductionism about information on the basis of the stated contingent scientific metaphysics. In terms of philosophical argumentation, realism about information is argued for primarily by way of an indispensability argument that defers to the practice of scientists and regards concepts of information as just as indispensable in their theories as contingent representations of structure. Physicalism and reductionism about information are adduced by way of the identity thesis that identifies the substance of the structure of ontic structural realism as identical to selections of structure existing in re to combined heterogeneous quantum fields, and to the total heterogeneous quantum field comprised of all such fields. Adjunctly, an informational statement of physicalism is arrived at, and a theory of semantic information is proposed, according to which information is intrinsically semantic and alethically neutral. (shrink)

In this paper I am advocating a structuralist theory of mental representation. For a structuralist theory of mental representation to be defended satisfactorily, the naturalistic and causal constraints have to be satisfied first. The more intractable of the two, i.e., the naturalistic constraint, indicates that to account for the mental representation, we should not invoke “a full-blown interpreting mind”. So, the aim of the paper is to show how the naturalistic and causal constraints could be satisfied. It aims to offer (...) a strategy for grounding the structure of the mental representations in nature. The strategy that I offer is inspired by Marcello Barbieri’s code model of biosemiotics. (shrink)

Symbol Grounding Problem is commonly considered one of the central challenges in the philosophy of artificial intelligence as its resolution is deemed necessary for bridging the gap between simple data processing and understanding of meaning and language. SGP has been addressed on numerous occasions with varying results, all resolution attempts having been severely, but for the most part justifiably, restricted by the Zero Semantic Commitment Condition. A further condition that demands explanatory power in terms of machine-to-human communication is the Non-Stupidity (...) Condition that demands an SG approach to be able to account for plausibility of higher-level language use and understanding, such as pragmatics. In this article, we undertake the endeavour of attempting to explain how merging certain early requirements for SG, such as embodiment, environmental interaction, and compliance with the Z-Condition with symbol emergence rather than direct attempts at symbol grounding can help emulate human language acquisition. Along with the presumption that mind and language are both symbolic and computational, we argue that some rather abstract aspects of language can be logically formalised and finally, that this melange of approaches can yield the explanatory power necessary to satisfy the Non-Stupidity Condition without breaking any previous conditions. (shrink)

One of the central debates in cognitive science is the dispute over the role of representation in cognition: on computational/representational accounts, representations are theoretically central; on dynamic systems approaches in which cognition is investigated as a particular sort of physical process, representations play either no role, or, at best, a derivative one. But these two perspectives lead to a deeply unsatisfying theoretical divide: accounts situated in the representational camp are plagued by the inscrutable problem of intentionality, while those hedging towards (...) anti-representationalism seem incapable of saying anything theoretically interesting about high-level cognition. This unhelpful polarization is due in part, at least, to a muddy debate; while some take representationalism to be a commitment to the necessity of conceptual representations for cognition, representations the having of which require certain conceptual capacities, others do not. Recently, there has been a surge of work on non-conceptual representation. This article aims to add to this movement by suggesting a particular cognitive mechanism for non-conceptual representations, one that plays a pivotal role in making conceptual representations possible. One of the central consequences of this new view of representation is the possibility of a non-question-begging naturalistic account of intentionality. (shrink)

As the accompanying articles in the Special Issue on Semiotic Scaffolding will attest, my colleagues in biosemiotics have done an exemplary job in showing us how to think about the critically generative role that semiotic scaffolding plays “vertically” – i.e., in evolutionary and developmental terms – by “allowing access to the upper floors” of biological complexity, cognition and evolution.In addition to such diachronic considerations of semiotic scaffolding, I wish to offer here a consideration of semiotic scaffolding’s synchronic power, as well (...) – and in particular the ability that it can afford its users to access new and other sign relations “horizontally” as a function of the way that multiple semiotically scaffolded relations intertwine to result in a “definite semantic topology that determines the ways that symbols modify each other’s referential functions in different combinations”.Taking up, in turn, Terrence Deacon’s later challenge that what the sciences of cognition – and biology more generally – needs to “come to grips with [is] the process of semiosis; the dynamic of interpretive activity by which semiotic relationships emerge from other semiotic relationships [as] intrinsically dynamic phases in a generative process”, I attempt here to show how Deacon’s own Peirce-inspired matrix of referential sign relations as presented in The Symbolic Species, when viewed as a semiotic scaffold of interactional constraints and possibility biases, provides the key to understanding the essentially thirdness-manifesting nature of symbol reference, formation and growth. (shrink)

Abstract: Luciano Floridi has impressively applied the concept of information to problems in semantics and epistemology, among other areas. In this essay, I briefly review two areas where I think one may usefully raise questions about some of Floridi's conclusions. One area is in the project to naturalize semantics and Floridi's use of the derived versus nonderived notion of semantic content. The other area is in the logic of information and knowledge and whether knowledge based on information necessarily supports closure, (...) in every instance. I suggest that it does not and, thereby, raise a challenge to Floridi's logic of being informed. (shrink)

We consider the symbol grounding problem, and apply to it philosophical arguments against Cartesianism developed by Sellars and McDowell: the problematic issue is the dichotomy between inside and outside which the definition of a physical symbol system presupposes. Surprisingly, one can question this dichotomy and still do symbolic computation: a detailed examination of the hardware and software of serial ports shows this.

Our chronically impoverished explanatory capacity in respect of P-consciousness is highly suggestive of a problem with science itself, rather than its lack of acquisition of some particular knowledge. The hidden assumption built into science is that science itself is a completed human behaviour. Removal of this assumption is achieved through a simple revision to our science model which is constructed, outlined and named ‘dual aspect science’ (DAS). It is constructed with reference to existing science being ‘single aspect science’. DAS is (...) consistent with and predictive of the very explanatory poverty that generated it and is simultaneously a seamless upgrade; no existing law of nature is altered or lost. The framework is completely empirically self-consistent and is validated empirically. DAS eliminates the behavioural inconsistencies currently inhabiting a world in which single aspect science has been inherited rather than chosen and in which its presuppositions are implemented through habit rather than by scientific examination of options by the scientists actually carrying out science. The proposed DAS framework provides a working vantage point from which an explanation of P-consciousness becomes expected and meaningful. The framework requires that we rediscover what we scientists do and then discover something new about ourselves: that how we have been doing science is not the entire story. Dual aspect science shows us what we have not been doing. (shrink)

The paper develops a new holistic theory of the word by integrating semiotic, linguistic, and psychological perspectives and introduces the Cogitative-Sensory Word Model, the CogSens Model, that unites the human mind and body. Saussure’s two-sided sign is replaced by a Peirce-inspired three-sided conception in which the expression unit mediates two content units, namely, an idea content connected to the human mind and an image content linked to the human body. It is argued that it is the word that makes human (...) language a unique tool of communication. Moreover, it is demonstrated how words interact with grammar to create an utterance. Finally, it is suggested that speech perception, i.e., fission of idea and image content, is the mirror image of speech production, i.e., fusion of idea and image content. (shrink)