It is a popular view thatpractical deliberation excludes foreknowledge of one's choice. Wolfgang Spohn and Isaac Levi have argued that not even a purely probabilistic self-predictionis available to thedeliberator, if one takes subjective probabilities to be conceptually linked to betting rates. It makes no sense to have a betting rate for an option, for one's willingness to bet on the option depends on the net gain from the bet, in combination with the option's antecedent utility, rather than on the offered (...) odds. And even apart from this consideration, assigning probabilities to the options among which one is choosing is futile since such probabilities could be of no possible use in choice. The paper subjects these arguments to critical examination and suggests that, appearances notwithstanding, practical deliberation need not crowd outself-prediction. (shrink)

“Only a small community has concentratedon general intelligence. No one has tried to make a thinking machine... The bottom line is that we really haven’t progressed too far toward a truly intelligent machine. We have collections of dumb specialists in small domains; the true majesty of general intelligence still awaits our attack.... We have got to get back to the deepest questions of AI and general intelligence... ” –MarvinMinsky as interviewed in Hal’s Legacy, edited (...) by David Stork, 2000. Our goal in creating this edited volume has been to?ll an apparent gap in the scienti?c literature, by providing a coherent presentation of a body of contemporary research that, in spite of its integral importance, has hitherto kept a very low pro?le within the scienti?c and intellectual community. This body of work has not been given a name before; in this book we christen it “Arti?cial General Intelligence”. What distinguishes AGI work from run-of-the-mill “arti?cial intelligence” research is that it is explicitly focused on engineering general intelligence in the short term. We have been active researchers in the AGI?eld for many years, and it has been a pleasure to gather together papers from our colleagues working on related ideas from their own perspectives. In the Introduction we give a conceptual overview of the AGI?eld, and also summarize and interrelate the key ideas of the papers in the subsequent chapters. (shrink)

The science of consciousness has made great strides in recent decades. However, the proliferation of competing theories makes it difficult to reach consensus about artificial consciousness. While for purely scientific purposes we might wish to adopt a ‘wait and see’ attitude, we may soon face practical and ethical questions about whether, for example, agents artificial systems are capable of suffering. Moreover, many of the methods used for assessing consciousness in humans and even non-human animals are not straightforwardly applicable to artificial (...) systems. With these challenges in mind, I propose that we look for ecumenical heuristics for artificial consciousness to enable us to make tentative assessments of the likelihood of consciousness arising in different artificial systems. I argue that such heuristics should have three main features: they should be (i) intuitively plausible, (ii) theoretically-neutral, and (iii) scientifically tractable. I claim that the concept of general intelligence — understood as a capacity for robust, flexible, and integrated cognition and behavior — satisfies these criteria and may thus provide the basis for such a heuristic, allowing us to make initial cautious estimations of which artificial systems are most likely to be conscious. -/- . (shrink)

What will be the next huge leap in humanity's progress? We cannot know for sure, but I am reasonably confident that it will involve the radical extension of technology into the domain of thought. Ray Kurzweil (2000, 2005) has eloquently summarized the arguments in favor of this position.

Current technology has dramatically increased the prevalence of studies to establish the genetic correlates of a wide variety of human characteristics, including not only the physical attributes that determine what we look like and the risk of physiological disease but also the psychological and cognitive characteristics that often define who we are as individuals. Perhaps one of the most deeply personal and often controversial characteristics is the concept of general intelligence, known in the psychological literature as “g.” As (...) with the genetic study of any complex trait, the first step in studying the genetics of g is to carefully define the characteristic of interest. For g, this entails establishing what intelligence means and providing a clear operational definition for how it will be measured. In this paper, we provide a brief historical and theoretical overview of the construct of general intelligence, describe its relationship to the contemporary measurement of intelligence, and discuss these concepts in light of the challenges associated with defining g as a characteristic in the study of genetics. (shrink)

This chapter discusses nine ways to bias open‐source artificial general intelligence (AGI) toward friendliness. There is no way to guarantee that advanced AGI, once created and released into the world, will behave according to human ethical standards. The primary objective of the chapter is to suggest some potential ways to do so. First it discusses an engineer the capability to acquire integrated ethical knowledge, and provides rich ethical interaction and instruction, respecting developmental stages. The chapter creates stable, hierarchy‐dominated (...) goal systems, and ensures that the early stages of recursive self‐improvement occur relatively slowly and with rich human involvement. It tightly links AGI with the Global Brain, and focuses on foster deep, consensus‐building interactions and commensurability between divergent viewpoints. The chapter creates a mutually supportive community of AGIs, and encourages measured co‐advancement of AGI software and AGI ethics theory. Finally it develops advanced AGI sooner. (shrink)

Howe et al.'s rejection of traditional discussion of talent is clearly acceptable, but their alternative has a weakness. They stress practice and hard work while referring vaguely to some basic biological substrate. High scores on a valid test of general intelligence provide a cultural-genetic basis for talented performance in a wide variety of specialties, ranging from engineering to the humanities. These choices may be entirely environmentally determined, and the highest levels of achievement do require practice and hard work.

Special Issue “Risks of artificial general intelligence”, Journal of Experimental and Theoretical Artificial Intelligence, 26/3 (2014), ed. Vincent C. Müller. http://www.tandfonline.com/toc/teta20/26/3# - Risks of general artificial intelligence, Vincent C. Müller, pages 297-301 - Autonomous technology and the greater human good - Steve Omohundro - pages 303-315 - - - The errors, insights and lessons of famous AI predictions – and what they mean for the future - Stuart Armstrong, Kaj Sotala & Seán S. Ó hÉigeartaigh (...) - pages 317-342 - - - The path to more general artificial intelligence - Ted Goertzel - pages 343-354 - - - Limitations and risks of machine ethics - Miles Brundage - pages 355-372 - - - Utility function security in artificially intelligent agents - Roman V. Yampolskiy - pages 373-389 - - - GOLEM: towards an AGI meta-architecture enabling both goal preservation and radical self-improvement - Ben Goertzel - pages 391-403 - - - Universal empathy and ethical bias for artificial general intelligence - Alexey Potapov & Sergey Rodionov - pages 405-416 - - - Bounding the impact of AGI - András Kornai - pages 417-438 - - - Ethics of brain emulations - Anders Sandberg - pages 439-457. (shrink)

Burkart et al. present a paradox – general factors of intelligence exist among individual differences (g) in performance in several species, and also at the aggregate level (G); however, there is ambiguous evidence for the existence of g when analyzing data using a mixed approach, that is, when comparing individuals of different species using the same cognitive ability battery. Here, we present an empirical solution to this paradox.

The presence of general intelligence poses a major evolutionary puzzle, which has led to increased interest in its presence in nonhuman animals. The aim of this review is to critically evaluate this question and to explore the implications for current theories about the evolution of cognition. We first review domain-general and domain-specific accounts of human cognition in order to situate attempts to identify general intelligence in nonhuman animals. Recent studies are consistent with the presence of (...) general intelligence in mammals. However, the interpretation of a psychometricgfactor as general intelligence needs to be validated, in particular in primates, and we propose a range of such tests. We then evaluate the implications of general intelligence in nonhuman animals for current theories about its evolution and find support for the cultural intelligence approach, which stresses the critical importance of social inputs during the ontogenetic construction of survival-relevant skills. The presence of general intelligence in nonhumans implies that modular abilities can arise in two ways, primarily through automatic development with fixed content and secondarily through learning and automatization with more variable content. The currently best-supported model, for humans and nonhuman vertebrates alike, thus construes the mind as a mix of skills based on primary and secondary modules. The relative importance of these two components is expected to vary widely among species, and we formulate tests to quantify their strength. (shrink)

We’ve seen a significant increase in the attention AI research is receiving this past decade, in large part due to some of the impressive feats of machine learning, particularly deep learning. This has resulted in something of a hype in the ability of AI’s in tackling various issues. The aim of the current essay is to examine the speculative questions “what would it mean for systems to transition from merely intelligently executing a task to knowledgably executing a task?” and “what (...) is the relation between consciousness and intelligence, such that specific evaluations about conscious AI might be made?” We offer what we hope is a roadmap for how to navigate asking these questions and a route forward in understanding how AI goes from intelligence to knowledge and potentially to some types of consciousness. (shrink)

Machine ethics and robot rights are quickly becoming hot topics in artificial intelligence and robotics communities. We will argue that attempts to attribute moral agency and assign rights to all intelligent machines are misguided, whether applied to infrahuman or superhuman AIs, as are proposals to limit the negative effects of AIs by constraining their behavior. As an alternative, we propose a new science of safety engineering for intelligent artificial agents based on maximizing for what humans value. In particular, we (...) challenge the scientific community to develop intelligent systems that have human-friendly values that they provably retain, even under recursive self-improvement. (shrink)

The Conference on Artificial General Intelligence is the only major conference series devoted wholly and specifically to the creation of AI systems possessing general intelligence at the human level and ultimately beyond. Its second installation, AGI-09, in Arlington, Virginia, March 6-9, 2009, attracted 67 paper submissions, which is a substantial increase from the previous year. Of these submissions, 33 (i.e., 49%) were accepted as full papers for presentation at the conference. Additional 13 papers were included as (...) position papers. The program also included a keynote address by Juergen Schmidhuber on 'The New AI', a post-conference workshop on 'The Future of AI', and a number of pre-conference tutorials on various topics related to AGI. (shrink)

The question of whether fluid intelligence can be differentiated from general intelligence in older adults is addressed. Data indicate that the developmental pattern of performance on fluid tasks differs from the pattern of general intelligence. These results suggest that it is important to identify changes in fluid cognitive functions associated with frontal lobe decline, as they may be early indicators of cognitive decline. (Published Online April 5 2006).

To what extent general intelligence mechanisms are associated with causal thinking is unclear. There has been little work done experimentally to determine which developing cognitive capacities help to integrate causal knowledge into explicit systems. To investigate this neglected aspect of development, 138 children aged 5–11 studying at mainstream primary schools completed a battery of three intelligence tests: one investigating verbal ability, another looking at verbal analogical, and a third assessing non-verbal/fluid reasoning. Children were also interviewed over the (...) course of three causal tasks, with the results showing that the developmental paths exhibited uneven profiles across the three causal phenomena. Children consistently found that explaining solution, where substances disappeared toward the end of the process, was more challenging. The confirmatory factor analyses suggested that the impact of cognitive ability factor in explicitly identifying causal relations was large. The proportion of the direct effect of general intelligence was 66% and it subsumed the variances of both verbal measures. Of this, 37% was the indirect effect of age. Fluid reasoning explained a further 28% of the variance, playing a unique role in causal explanation. The results suggested that, overall, cognitive abilities are substantially related to causal reasoning, but not entirely due to developmental differences in “g” during the age periods studied. (shrink)

Pei Wang, Ben Goertzel. [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [ 18] [19] [20] [21] [22] [23] [24] [25] [26] [27] Bach, J. (2009). Principles ofSynthetic Intelligence PSI: An Architecture ofMotivated Cognition (Oxford University Press,  ...

This article offers comprehensive criticism of the Turing test and develops quality criteria for new artificial general intelligence (AGI) assessment tests. It is shown that the prerequisites A. Turing drew upon when reducing personality and human consciousness to “suitable branches of thought” re-flected the engineering level of his time. In fact, the Turing “imitation game” employed only symbolic communication and ignored the physical world. This paper suggests that by restricting thinking ability to symbolic systems alone Turing unknowingly constructed (...) “the wall” that excludes any possi-bility of transition from a complex observable phenomenon to an abstract image or concept. It is, therefore, sensible to factor in new requirements for AI (artificial intelligence) maturity assessment when approaching the Tu-ring test. Such AI must support all forms of communication with a human being, and it should be able to comprehend abstract images and specify con-cepts as well as participate in social practices. (shrink)

An arms race for an artificial general intelligence would be detrimental for and even pose an existential threat to humanity if it results in an unfriendly AGI. In this paper, an all-pay contest model is developed to derive implications for public policy to avoid such an outcome. It is established that, in a winner-takes-all race, where players must invest in R&D, only the most competitive teams will participate. Thus, given the difficulty of AGI, the number of competing teams (...) is unlikely ever to be very large. It is also established that the intention of teams competing in an AGI race, as well as the possibility of an intermediate outcome, is important. The possibility of an intermediate prize will raise the probability of finding the dominant AGI application and, hence, will make public control more urgent. It is recommended that the danger of an unfriendly AGI can be reduced by taxing AI and using public procurement. This would reduce the pay-off of contestants, raise the amount of R&D needed to compete, and coordinate and incentivize co-operation. This will help to alleviate the control and political problems in AI. Future research is needed to elaborate the design of systems of public procurement of AI innovation and for appropriately adjusting the legal frameworks underpinning high-tech innovation, in particular dealing with patenting by AI. (shrink)

The article explores the transformative impact on human and social sciences in response to anticipated societal shifts driven by the forthcoming proliferation of artificial systems, whose intelligence will match human capabilities. Initially, it was posited that artificial intelligence (AI) would excel beyond human abilities in computational tasks and algorithmic operations, leaving creativity and humanities as uniquely human domains. However, recent advancements in large language models have significantly challenged these conventional beliefs about AI’s limitations and strengths. It is projected (...) that, in the near future, generative AI models will adeptly replicate individual qualities, desires, beliefs, opinions, and the essence of human identity and consciousness to a degree that is nearly indistinguishable from that of humans. This lends support to the connectionist approach to understanding consciousness, suggesting an inherent similarity between biological and artificial neural networks. The discussion posits two innovative areas of scientific inquiry: “mathematical anthropology” and “multi-dimensional calculus of value.” The former suggests that viewing human nature through a mathematical and calculative lens not only preserves but enriches our understanding of the complexity of anthropological experience and its perceived contradictions. The latter hypothesis explores how mathematical models could facilitate various social interactions. The advanced simulation capabilities of neural networks suggest that traditional social and political frameworks face a growing vulnerability to AI-driven manipulations. This trend underscores the urgency of developing social interaction models that incorporate explicitly defined calculative rules. In conclusion, the paper advocates for a paradigm shift in how mathematics is perceived – not merely as a tool for computation but as a foundational science for crafting sophisticated models. In conclusion, the article advocates for transitioning from perceiving mathematics solely as a science of computation to viewing it as a discipline dedicated to constructing various formalized models, thereby deepening our insight into the complexities of human and social phenomena. (shrink)

This target article considers the relation of fluid cognitive functioning to general intelligence. A neurobiological model differentiating working memory/executive function cognitive processes of the prefrontal cortex from aspects of psychometrically defined general intelligence is presented. Work examining the rise in mean intelligence-test performance between normative cohorts, the neuropsychology and neuroscience of cognitive function in typically and atypically developing human populations, and stress, brain development, and corticolimbic connectivity in human and nonhuman animal models is reviewed and (...) found to provide evidence of mechanisms through which early experience affects the development of an aspect of cognition closely related to, but distinct from, general intelligence. Particular emphasis is placed on the role of emotion in fluid cognition and on research indicating fluid cognitive deficits associated with early hippocampal pathology and with dysregulation of the hypothalamic-pituitary-adrenal axis stress-response system. Findings are seen to be consistent with the idea of an independent fluid cognitive construct and to assist with the interpretation of findings from the study of early compensatory education for children facing psychosocial adversity and from behavior genetic research on intelligence. It is concluded that ongoing development of neurobiologically grounded measures of fluid cognitive skills appropriate for young children will play a key role in understanding early mental development and the adaptive success to which it is related, particularly for young children facing social and economic disadvantage. Specifically, in the evaluation of the efficacy of compensatory education efforts such as Head Start and the readiness for school of children from diverse backgrounds, it is important to distinguish fluid cognition from psychometrically defined general intelligence. (Published Online April 5 2006) Key Words: cognition; cognition-emotion reciprocity; developmental disorders; emotion; fluid cognition; Flynn effect; general intelligence; limbic system; neuroscience; phenylketonuria; prefrontal cortex; psychometrics; schizophrenia. (shrink)

Although brain size and the concept of intelligence have been extensively used in comparative neuroscience to study cognition and its evolution, such coarse-grained traits may not be informative enough about important aspects of neurocognitive systems. By taking into account the different evolutionary trajectories and the selection pressures on neurophysiology across species, Logan and colleagues suggest that the cognitive abilities of an organism should be investigated by considering the fine-grained and species-specific phenotypic traits that characterize it. In such a way, (...) we would avoid adopting human-oriented, coarse-grained traits, typical of the standard approach in cognitive neuroscience. We argue that this standard approach can fail in some cases, but can, however, work in others, by discussing two major topics in contemporary neuroscience as examples: general intelligence and brain asymmetries. (shrink)

Acquisition of word meanings, or vocabulary, reflects general mental ability (psychometric g) more than than do most abilities measured in test batteries. Among diverse subtests, vocabulary is especially high on indices of genetic influences. Bloom's exposition of the psychological complexities of understanding words, involving the primacy of concepts, the theory of mind, and other processes, explains vocabulary's predominant g saturation.

We propose that, for the purpose of studying theoretical properties of the knowledge of an agent with Artificial General Intelligence (that is, the knowledge of an AGI), a pragmatic way to define such an agent’s knowledge (restricted to the language of Epistemic Arithmetic, or EA) is as follows. We declare an AGI to know an EA-statement φ if and only if that AGI would include φ in the resulting enumeration if that AGI were commanded: “Enumerate all the EA-sentences (...) which you know.” This definition is non-circular because an AGI, being capable of practical English communication, is capable of understanding the everyday English word “know” independently of how any philosopher formally defines knowledge; we elaborate further on the non-circularity of this circular-looking definition. This elegantly solves the problem that different AGIs may have different internal knowledge definitions and yet we want to study knowledge of AGIs in general, without having to study different AGIs separately just because they have separate internal knowledge definitions. Finally, we suggest how this definition of AGI knowledge can be used as a bridge which could allow the AGI research community to import certain abstract results about mechanical knowing agents from mathematical logic. (shrink)

Nowadays, new directions for the development of artificial intelligence (AI) have emerged, the task has been set to develop artificial general intelligence (AGI), which is able to go beyond the narrow AI, gain a high degree of autonomy, independently solve problems in different environmental conditions and thus have the ability to perform the functions of natural intelligence. In this regard, important philosophical, theoretical, and methodological questions arise concerning the definition and evaluation of the social significance of (...) new AI achievements, especially regarding the correlation of their socio-humanitarian and technological aspects. It is necessary to overcome a purely technocratic approach, which usually ignores the negative consequences of digitalization, the possible risks and threats of AI development. These difficulties are conditioned by the paradigm gap between the system of concepts used to describe technologies and the systems of concepts specific to socio-humanitarian descriptions and explanations. The so-called ontological paradox arises during digital transformation and the introduction of AI into social systems. An ontological paradox is also an epistemological paradox, since any statement of an ontological type presupposes its epistemological justification. To overcome this gap, V.E. Lepskiy proposes the conception of self-developing polysubjective environments – cybernetics of the third order. This conception allows to create a conceptual “bridge” between two systems of definitions that do not have direct logical connections between them. A productive tool for this can be the information approach, which is widely used to solve such interdisciplinary problems. This makes possible to theoretically correctly connect technological and socio-humanitarian descriptions in a single conceptual structure and to analyze main socio-humanitarian criteria for evaluating digital innovations. The article discusses in detail the main provisions of the information approach and its application for the development of AGI systems, the socio-humanitarian assessment of AI autonomous development, and the resolution of security issues. (shrink)

This article interrogates the challenge artificial general intelligence (AGI) poses to religion and human societies, in general. More specifically, it seeks to respond to “Singularity”—when machines reach a level of intelligence that would put into question the privileged position humanity enjoys as imago Dei . Employing the Bemba notion of mystico‐relationality in dialogue with the concepts of the “created co‐creator” and Christ the Key, it argues for the possibility of AI participating in imago Dei . The (...) findings show that imaging is a fluid, participatory activity that aims at likeness, but also social harmony. It also argues that God is the only original creator, humans are created creators, and that every aspect of visible existence, including AI, is inherently divine imaging. However, strong imaging is only attainable based on the only One and True Image—Christ, whose union of the material and the divine means that all creation can image, excluding nothing, even AI. (shrink)

A simple argument suggests that we can fruitfully model advanced AI systems using expected utility theory. According to this argument, an agent will need to act as if maximising expected utility if they’re to avoid exploitation. Insofar as we should expect advanced AI to avoid exploitation, it follows that we should expected advanced AI to act as if maximising expected utility. I spell out this argument more carefully and demonstrate that it fails, but show that the manner of its failure (...) is instructive: in exploring the argument, we gain insight into how to model advanced AI systems. (shrink)

Psychology has benefited from an enormous wealth of knowledge about processes of cognition in relation to how the brain organizes information. Within the categorization literature, this behavior is often explained through theories of memory construction called exemplar theory and prototype theory which are typically based on similarity or rule functions as explanations of how categories emerge. Although these theories work well at modeling highly controlled stimuli in laboratory settings, they often perform less well outside of these settings, such as explaining (...) the emergence of background knowledge processes. In order to explain background knowledge, we present a non-similarity-based post-Skinnerian theory of human language called Relational Frame Theory which is rooted in a philosophical world view called functional contextualism. This theory offers a very different interpretation of how categories emerge through the functions of behavior and through contextual cues, which may be of some benefit to existing categorization theories. Specifically, RFT may be able to offer a novel explanation of how background knowledge arises, and we provide some mathematical considerations in order to identify a formal model. Finally, we discuss much of this work within the broader context of general semantic knowledge and artificial intelligence research. (shrink)

In recent years, a trend in AI research has started to pursue human-level, general artificial intelli-gence (AGI). Although the AGI framework is characterised by different viewpoints on what intelligence is and how to implement it in artificial systems, it conceptualises intelligence as flexible, general-purposed, and capable of self-adapting to different contexts and tasks. Two important ques-tions remain open: a) should AGI projects simu-late the biological, neural, and cognitive mecha-nisms realising the human intelligent behaviour? and b) what (...) is the relationship, if any, between the concept of general intelligence adopted by AGI and that adopted by psychometricians, i.e., the g factor? In this paper, we address these ques-tions and invite researchers in AI to open a dis-cussion on the theoretical conceptions and practi-cal purposes of the AGI approach. (shrink)