The notion of intelligence is relevant to several fields of research, including cognitive and comparative psychology, neuroscience, artificial intelligence, and philosophy, among others. However, there is little agreement within and across these fields on how to characterise and explain intelligence. I put forward a behavioural, operational characterisation of intelligence that can play an integrative role in the sciences of intelligence, as well as preserve the distinctive explanatory value of the notion, setting it apart from the related concepts of cognition and (...) rationality. Finally, I examine a popular hypothesis about the underpinnings of intelligence: the capacity to manipulate internal representations of the environment. I argue that the hypothesis needs refinement, and that so refined, it applies only to some forms of intelligence. (shrink)

From artistic performances in the visual arts and in music to motor control in gymnastics, from tool use to chess and language, humans excel in a variety of skills. On the plausible assumption that skillful behavior is a visible manifestation of intelligence, a theory of intelligence—whether human or not—should be informed by a theory of skills. More controversial is the question as to whether, in order to theorize about intelligence, we should study certain skills in particular. My target is the (...) view that only a particular class/kind of skill (i.e., ‘theoretical’, or ‘intellectual’ skills, versus ‘practical’, or ‘embodied’ skills) manifests intelligence, or especially does so. I call this view ‘Intelligence Elitism’. Intelligence Elitism is pervasive in popular culture, in academic culture, as well as in psychometrics. It has, arguably, a long pedigree in philosophy (though explicit arguments for it are rare). But it is a substantive view which, if false, can promote unfairness. In this article, I defend Intelligence Socialism—the claim that intelligent behavior is everywhere skillful behavior is. I go on to isolate the best case for the further Elitist claim that, even if all skillful behavior is intelligent in some modest sense of ‘intelligence’, there is a principled difference in intelligence between theoretical, or intellectual, skills, on one side, and practical, or embodied skills, on the other, and I show it wanting. (shrink)

The technological innovations of the Fourth Industrial Revolution have facilitated the formulation of strategies to transcend human limitations; strategies that are widely supported by the transhumanist philosophy. The purpose of this article is to explain the relationship between ‘transhumanism’ and ‘technological singularity’, to which end the Fourth Industrial Revolution and transhumanism are also briefly covered. Subsequently, the three main models of technological singularity are evaluated and a definition of this futuristic concept is offered. Finally, the author provides a reflection on (...) the relationship between technological singularity and the transhumanist philosophy, and concludes that they are not synonymous, but rather that the former is an eventuality that will allow the latter. (shrink)

Inspired by recent progress in multi-agent Reinforcement Learning (RL), in this work we examine the collective intelligent behaviour of theoretical universal agents by introducing a weighted mixture operation. Given a weighted set of agents, their weighted mixture is a new agent whose expected total reward in any environment is the corresponding weighted average of the original agents' expected total rewards in that environment. Thus, if RL agent intelligence is quantified in terms of performance across environments, the weighted mixture's intelligence is (...) the weighted average of the original agents' intelligences. This operation enables various interesting new theorems that shed light on the geometry of RL agent intelligence, namely: results about symmetries, convex agent-sets, and local extrema. We also show that any RL agent intelligence measure based on average performance across environments, subject to certain weak technical conditions, is identical (up to a constant factor) to performance within a single environment dependent on said intelligence measure. (shrink)

This is a paper about the general theory of measuring or estimating social intelligence via benchmarks. Hernández-Orallo and Dowe described a problem with certain proposed intelligence measures. The problem suggests that those intelligence measures might not accurately capture social intelligence. We argue that Hernández-Orallo and Dowe's problem is even more general than how they stated it, applying to many subdomains of AGI, not just the one subdomain in which they stated it. We then propose a solution. In our solution, instead (...) of using test-cases within the given AGI subdomain to estimate an AI's intelligence, one would use test-cases in an extended subdomain where test-cases have the ability to simulate the AI being tested. Surprisingly, AIs only designed for the original subdomain can be tested with test-cases in the extended subdomain anyway. By extending the subdomain in this way, we might avoid Hernández-Orallo and Dowe's problem. (shrink)

This paper argues that intelligence can be approximated by the ability to produce accurate predictions. It is further argued that general intelligence can be approximated by context dependent predictive abilities combined with the ability to use working memory to abstract away contextual information. The flexibility associated with general intelligence can be understood as the ability to use selective attention to focus on specific aspects of sensory impressions to identify patterns, which can then be used to predict events in novel situations (...) and environments. The argumentation synthesizes Godfrey-Smith’s environmental complexity theory, adding the notion of niche broadness as well as changes concerning the view of cognition and control, and Hohwy’s predictive mind theory, making explicit the significance of accuracy as a composite of trueness and precision where the nervous system acts as a distributed controller motivating actions that keep the body in homeostasis. (shrink)

The book’s core argument is that an artificial intelligence that could equal or exceed human intelligence—sometimes called artificial general intelligence (AGI)—is for mathematical reasons impossible. It offers two specific reasons for this claim: Human intelligence is a capability of a complex dynamic system—the human brain and central nervous system. Systems of this sort cannot be modelled mathematically in a way that allows them to operate inside a computer. In supporting their claim, the authors, Jobst Landgrebe and Barry Smith, marshal evidence (...) from mathematics, physics, computer science, philosophy, linguistics, and biology, setting up their book around three central questions: What are the essential marks of human intelligence? What is it that researchers try to do when they attempt to achieve "artificial intelligence" (AI)? And why, after more than 50 years, are our most common interactions with AI, for example with our bank’s computers, still so unsatisfactory? Landgrebe and Smith show how a widespread fear about AI’s potential to bring about radical changes in the nature of human beings and in the human social order is founded on an error. There is still, as they demonstrate in a final chapter, a great deal that AI can achieve which will benefit humanity. But these benefits will be achieved without the aid of systems that are more powerful than humans, which are as impossible as AI systems that are intrinsically "evil" or able to "will" a takeover of human society. (shrink)

Is it OK to lie to Siri? Is it bad to mistreat a robot for our own pleasure? Under what condition should we grant a moral status to an artificial intelligence (AI) system? This paper looks at different arguments for granting moral status to an AI system: the idea of indirect duties, the relational argument, the argument from intelligence, the arguments from life and information, and the argument from sentience. In each but the last case, we find unresolved issues with (...) the particular argument, which leads us to move to a different one. We leave the idea of indirect duties aside since these duties do not imply considering an AI system for its own sake. The paper rejects the relational argument and the argument from intelligence. The argument from life may lead us to grant a moral status to an AI system, but only in a weak sense. Sentience, by contrast, is a strong argument for the moral status of an AI system—based, among other things, on the Aristotelian principle of equality: that same cases should be treated in the same way. The paper points out, however, that no AI system is sentient given the current level of technological development. (shrink)

In their paper 'Reward is enough', Silver et al conjecture that the creation of sufficiently good reinforcement learning (RL) agents is a path to artificial general intelligence (AGI). We consider one aspect of intelligence Silver et al did not consider in their paper, namely, that aspect of intelligence involved in designing RL agents. If that is within human reach, then it should also be within AGI's reach. This raises the question: is there an RL environment which incentivises RL agents to (...) design RL agents? (shrink)

The standard argument to the conclusion that artificial intelligence (AI) constitutes an existential risk for the human species uses two premises: (1) AI may reach superintelligent levels, at which point we humans lose control (the ‘singularity claim’); (2) Any level of intelligence can go along with any goal (the ‘orthogonality thesis’). We find that the singularity claim requires a notion of ‘general intelligence’, while the orthogonality thesis requires a notion of ‘instrumental intelligence’. If this interpretation is correct, they cannot be (...) joined as premises and the argument for the existential risk of AI turns out invalid. If the interpretation is incorrect and both premises use the same notion of intelligence, then at least one of the premises is false and the orthogonality thesis remains itself orthogonal to the argument to existential risk from AI. In either case, the standard argument for existential risk from AI is not sound.—Having said that, there remains a risk of instrumental AI to cause very significant damage if designed or used badly, though this is not due to superintelligence or a singularity. (shrink)

We consider an extended notion of reinforcement learning in which the environment can simulate the agent and base its outputs on the agent's hypothetical behavior. Since good performance usually requires paying attention to whatever things the environment's outputs are based on, we argue that for an agent to achieve on-average good performance across many such extended environments, it is necessary for the agent to self-reflect. Thus weighted-average performance over the space of all suitably well-behaved extended environments could be considered a (...) way of measuring how self-reflective an agent is. We give examples of extended environments and introduce a simple transformation which experimentally seems to increase some standard RL agents' performance in a certain type of extended environment. (shrink)

Is it OK to lie to Siri? Is it bad to mistreat a robot for our own pleasure? Under what condition should we grant a moral status to an artificial intelligence system? This paper looks at different arguments for granting moral status to an AI system: the idea of indirect duties, the relational argument, the argument from intelligence, the arguments from life and information, and the argument from sentience. In each but the last case, we find unresolved issues with the (...) particular argument, which leads us to move to a different one. We leave the idea of indirect duties aside since these duties do not imply considering an AI system for its own sake. The paper rejects the relational argument and the argument from intelligence. The argument from life may lead us to grant a moral status to an AI system, but only in a weak sense. Sentience, by contrast, is a strong argument for the moral status of an AI system—based, among other things, on the Aristotelian principle of equality: that same cases should be treated in the same way. The paper points out, however, that no AI system is sentient given the current level of technological development. (shrink)

This paper looks at philosophical questions that arise in the context of AI alignment. It defends three propositions. First, normative and technical aspects of the AI alignment problem are interrelated, creating space for productive engagement between people working in both domains. Second, it is important to be clear about the goal of alignment. There are significant differences between AI that aligns with instructions, intentions, revealed preferences, ideal preferences, interests and values. A principle-based approach to AI alignment, which combines these elements (...) in a systematic way, has considerable advantages in this context. Third, the central challenge for theorists is not to identify ‘true’ moral principles for AI; rather, it is to identify fair principles for alignment that receive reflective endorsement despite widespread variation in people’s moral beliefs. The final part of the paper explores three ways in which fair principles for AI alignment could potentially be identified. (shrink)

In ‘Computing Machinery and Intelligence’, Turing, sceptical of the question ‘Can machines think?’, quickly replaces it with an experimentally verifiable test: the imitation game. I suggest that for such a move to be successful the test needs to be relevant, expansive, solvable by exemplars, unpredictable, and lead to actionable research. The Imitation Game is only partially successful in this regard and its reliance on language, whilst insightful for partially solving the problem, has put AI progress on the wrong foot, prescribing (...) a top-down approach for building thinking machines. I argue that to fix shortcomings with modern AI systems a nonverbal operationalisation is required. This is provided by the recent Animal-AI Testbed, which translates animal cognition tests for AI and provides a bottom-up research pathway for building thinking machines that create predictive models of their environment from sensory input. (shrink)

After generalizing the Archimedean property of real numbers in such a way as to make it adaptable to non-numeric structures, we demonstrate that the real numbers cannot be used to accurately measure non-Archimedean structures. We argue that, since an agent with Artificial General Intelligence (AGI) should have no problem engaging in tasks that inherently involve non-Archimedean rewards, and since traditional reinforcement learning rewards are real numbers, therefore traditional reinforcement learning probably will not lead to AGI. We indicate two possible ways (...) traditional reinforcement learning could be altered to remove this roadblock. (shrink)

What kinds of fundamental limits are there in how capable artificial intelligence (AI) systems might become? Two questions in particular are of interest: (1) How much more capable could AI become relative to humans, and (2) how easily could superhuman capability be acquired? To answer these questions, we will consider the literature on human expertise and intelligence, discuss its relevance for AI, and consider how AI could improve on humans in two major aspects of thought and expertise, namely simulation and (...) pattern recognition. We find that although there are very real limits to prediction, it seems like AI could still substantially improve on human intelligence. (shrink)

The main goal of my paper is to argue that data compression is a necessary condition for intelligence. One key motivation for this proposal stems from a paradox about intuition and intelligence. For the purposes of this paper, it will be useful to consider playing board games—such as chess and Go—as a paradigm of problem solving and cognition, and computer programs as a model of human cognition. I first describe the basic components of computer programs that play board games, namely (...) value functions and search functions. I then argue that value functions both play the same role as intuition in humans and work in essentially the same way. However, as will become apparent, using an ordinary value function is just a simpler and less accurate form of relying on a database or lookup table. This raises our paradox, since reliance on intuition is usually considered to manifest intelligence, whereas usage of a lookup table is not. I therefore introduce another condition for intelligence that is related to data compression. This proposal allows that even reliance on a perfectly accurate lookup table can be nonintelligent, while retaining the claim that reliance on intuition can be highly intelligent. My account is not just theoretically plausible, but it also captures a crucial empirical constraint. This is because all systems with limited resources that solve complex problems—and hence, all cognitive systems—need to compress data. (shrink)

I want to suggest that the major influence of classical arguments for embodiment like "The Embodied Mind" by Varela, Thomson & Rosch (1991) has been a changing of positions rather than a refutation: Cognitivism has found ways to retreat and regroup at positions that have better fortification, especially when it concerns theses about artificial intelligence or artificial cognitive systems. For example: a) Agent-based cognitivism' that understands humans as taking in representations of the world, doing rule-based processing and then acting on (...) them (sense-plan-act) is often limited to conscious decision processes; and b) Purely syntactic cognition is compatible with embodiment, or supplemented by embodiment (e.g. for 'grounding'). While the empirical thesis of embodied cognition ('embodied cognitive science') is true and the practical engineering thesis ('morphological computation', 'cheap design') is often true, the conceptual thesis ('embodiment is necessary for cognition') is likely false - syntax is often enough for cognition, unless grounding is really necessary. I conclude that it has become more sensible to integrate embodiment with traditional approaches rather than "fight for embodiment" or "against cognitivism". (shrink)

Proceedings of the papers presented at the Symposium on "Revisiting Turing and his Test: Comprehensiveness, Qualia, and the Real World" at the 2012 AISB and IACAP Symposium that was held in the Turing year 2012, 2–6 July at the University of Birmingham, UK. Ten papers. - http://www.pt-ai.org/turing-test --- Daniel Devatman Hromada: From Taxonomy of Turing Test-Consistent Scenarios Towards Attribution of Legal Status to Meta-modular Artificial Autonomous Agents - Michael Zillich: My Robot is Smarter than Your Robot: On the Need for (...) a Total Turing Test for Robots - Adam Linson, Chris Dobbyn and Robin Laney: Interactive Intelligence: Behaviour-based AI, Musical HCI and the Turing Test - Javier Insa, Jose Hernandez-Orallo, Sergio España - David Dowe and M.Victoria Hernandez-Lloreda: The anYnt Project Intelligence Test (Demo) - Jose Hernandez-Orallo, Javier Insa, David Dowe and Bill Hibbard: Turing Machines and Recursive Turing Tests — Francesco Bianchini and Domenica Bruni: What Language for Turing Test in the Age of Qualia? - Paul Schweizer: Could there be a Turing Test for Qualia? - Antonio Chella and Riccardo Manzotti: Jazz and Machine Consciousness: Towards a New Turing Test - William York and Jerry Swan: Taking Turing Seriously (But Not Literally) - Hajo Greif: Laws of Form and the Force of Function: Variations on the Turing Test. (shrink)

A key distinction in ethics is between members and nonmembers of the moral community. Over time, our notion of this community has expanded as we have moved from a rationality criterion to a sentience criterion for membership. I argue that a sentience criterion is insufficient to accommodate all members of the moral community; the true underlying criterion can be understood in terms of whether a being has interests. This may be extended to conscious, self-aware machines, as well as to any (...) autonomous intelligent machines. Such machines exhibit an ability to formulate desires for the course of their own existence; this gives them basic moral standing. While not all machines display autonomy, those which do must be treated as moral patients; to ignore their claims to moral recognition is to repeat past errors. I thus urge moral generosity with respect to the ethical claims of intelligent machines. (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)

I survey four categories of factors that might give a digital mind, such as an upload or an artificial general intelligence, an advantage over humans. Hardware advantages include greater serial speeds and greater parallel speeds. Self-improvement advantages include improvement of algorithms, design of new mental modules, and modification of motivational system. Co-operative advantages include copyability, perfect co-operation, improved communication, and transfer of skills. Human handicaps include computational limitations and faulty heuristics, human-centric biases, and socially motivated cognition. The shape of hardware (...) growth curves, as well as the ease of modifying minds, are found to have a major impact on how quickly a digital mind may take advantage of these factors. (shrink)

Intending to have a constructive dialogue with the combination of evolutionary theory (E) and metaphysical naturalism (N), Alvin Plantinga’s evolutionary argument against naturalism (EAAN) takes the reliability of human cognition (in normal environments) as a purported explanandum and E&N as a purported explanans. Then, he considers whether E&N can offer a good explanans for this explanandum, and his answer is negative (an answer employed by him to produce a defeater for N). But I will argue that the whole EAAN goes (...) wrong by assuming that R is a qualified explanandum crying out for scientific explanations, since it cannot meet either of the two criteria for any scientifically qualified explanandum: Realizability Criterion and Informativeness Criterion. Hence, EAAN is simply setting a task that E&N, as a scientific theory, will not care at all. Therefore, EAAN cannot substantially shake E&N. (shrink)

This paper revisits the often debated question Can machines think? It is argued that the usual identification of machines with the notion of algorithm has been both counter-intuitive and counter-productive. This is based on the fact that the notion of algorithm just requires an algorithm to contain a finite but arbitrary number of rules. It is argued that intuitively people tend to think of an algorithm to have a rather limited number of rules. The paper will further propose a modification (...) of the above mentioned explication of the notion of machines by quantifying the length of an algorithm. Based on that it appears possible to reconcile the opposing views on the topic, which people have been arguing about for more than half a century. (shrink)

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. (shrink)

An intelligent machine surpassing human intelligence across a wide set of skills has been proposed as a possible existential catastrophe. Among those concerned about existential risk related to artificial intelligence, it is common to assume that AI will not only be very intelligent, but also be a general agent. This article explores the characteristics of machine agency, and what it would mean for a machine to become a general agent. In particular, it does so by articulating some important differences between (...) belief and desire in the context of machine agency. One such difference is that while an agent can by itself acquire new beliefs through learning, desires need to be derived from preexisting desires or acquired with the help of an external influence. Such influence could be a human programmer or natural selection. We argue that to become a general agent, a machine needs productive desires, or desires that can direct behavior across multiple contexts. However, productive desires cannot sui generis be derived from non-productive desires. Thus, even though general agency in AI could in principle be created by human agents, general agency cannot be spontaneously produced by a non-general AI agent through an endogenous process. In conclusion, we argue that a common AI scenario, where general agency suddenly emerges in a non-general agent AI, such as DeepMind’s superintelligent board game AI AlphaZero, is not plausible. (shrink)

V nasledujúcom príspevku sa snažíme zdôvodniť vyčlenenie troch typov procesov v pozorovanom vesmíre - procesov slabo, stredne a silne usmernených, a to na základe rôznej miery autonómnosti ich štruktúrnych prvkov a rôznej miery či intenzity zákonov, ktorými sú usmerňované alebo riadené. Individuálne a konkrétne procesy sú tak v podstate totožné s individuálnymi a konkrétnymi systémami, cez ktoré, v ktorých a prostredníctvom ktorých sa úplne realizujú, disponujúc tak vždy a všade vlastným substanciálnym obsahom. Na tomto základe potom vyčleňujeme slabo usmernené procesy (...) v neživej prírode s málo vyčlenenými alebo autonómnymi štruktúrnymi prvkami, stredne usmernené procesy živej prírody s oveľa väčšími, vyčlenenejšími a autonómnejšími štruktúrnymi prvkami vo forme živých buniek a postupne rastúcich a zdokonaľujúcich sa organizmov a tried organizmov a nakoniec silne usmernené procesy tzv. sociálnej prírody alebo ľudskej spoločnosti, pozostávajúcej z ľudí ako svojich základných alebo štruktúrnych prvkov a tiež z nimi generovaného technického a/lebo civilizačného "obalu", ktoré sú podobne ako živé procesy a systémy usmerňované a riadené čoraz väčším počtom pre nich špecifických vedeckých zákonov, princípov alebo "pravidiel". (shrink)

From an object-oriented perspective, this paper investigates the interdisciplinary aspects of problem representation as well the differences between representation of problems in the mind and that in the machine. By defining an object as a combination of a symbol-structure and its associated operations, it shows how the representation of problems can become related to control, which conducts the search in finding a solution. Different types of representation of problems in the machine are classified into four categories, and in a similar (...) way four distinct models are distinguished for the representation of problems in the mind. The concept of layered hierarchies, as the main theme of the object-oriented paradigm, is used to examine the implications of problem representation in the mind for improving the representation of problems in the machine. (shrink)

Legg and Hutter, as well as subsequent authors, considered intelligent agents through the lens of interaction with reward-giving environments, attempting to assign numeric intelligence measures to such agents, with the guiding principle that a more intelligent agent should gain higher rewards from environments in some aggregate sense. In this paper, we consider a related question: rather than measure numeric intelligence of one Legg- Hutter agent, how can we compare the relative intelligence of two Legg-Hutter agents? We propose an elegant answer (...) based on the following insight: we can view Legg-Hutter agents as candidates in an election, whose voters are environments, letting each environment vote (via its rewards) which agent (if either) is more intelligent. This leads to an abstract family of comparators simple enough that we can prove some structural theorems about them. It is an open question whether these structural theorems apply to more practical intelligence measures. (shrink)

Since Turing proposed the first test of intelligence, several modifications have been proposed with the aim of making Turing’s proposal more realistic and applicable in the search for artificial intelligence. In the modern context, it turns out that some of these definitions of intelligence and the corresponding tests merely measure computational power. Furthermore, in the framework of the original Turing test, for a system to prove itself to be intelligent, a certain amount of deceit is implicitly required which can have (...) serious security implications for future human societies. In this article, we propose a unified framework for developing intelligence tests which takes care of important ethical and practical issues. Our proposed framework has several important consequences. Firstly, it results in the suggestion that it is not possible to construct a single, context independent, intelligence test. Secondly, any measure of intelligence must have access to the process by which a problem is solved by the system under consideration and not merely the final solution. Finally, it requires an intelligent agent to be evolutionary in nature with the flexibility to explore new algorithms on its own. (shrink)

Animals, including humans, are usually judged on what they could become, rather than what they are. Many physical and cognitive abilities in the ‘animal kingdom’ are only acquired (to a given degree) when the subject reaches a certain stage of development, which can be accelerated or spoilt depending on how the environment, training or education is. The term ‘potential ability’ usually refers to how quick and likely the process of attaining the ability is. In principle, things should not be different (...) for the ‘machine kingdom’. While machines can be characterised by a set of cognitive abilities, and measuring them is already a big challenge, known as ‘universal psychometrics’, a more informative, and yet more challenging, goal would be to also determine the potential cognitive abilities of a machine. In this paper we investigate the notion of potential cognitive ability for machines, focussing especially on universality and intelligence. We consider several machine characterisations (non-interactive and interactive) and give definitions for each case, considering permanent and temporal potentials. From these definitions, we analyse the relation between some potential abilities, we bring out the dependency on the environment distribution and we suggest some ideas about how potential abilities can be measured. Finally, we also analyse the potential of environments at different levels and briefly discuss whether machines should be designed to be intelligent or potentially intelligent. (shrink)

Intelligence can be understood as a form of rationality, in the sense that an intelligent system does its best when its knowledge and resources are insufficient with respect to the problems to be solved. The traditional models of rationality typically assume some form of sufficiency of knowledge and resources, so cannot solve many theoretical and practical problems in Artificial Intelligence (AI). New models based on the Assumption of Insufficient Knowledge and Resources (AIKR) cannot be obtained by minor revisions or extensions (...) of the traditional models, and have to be established fully according to the restrictions and freedoms provided by AIKR. The practice of NARS, an AI project, shows that such new models are feasible and promising in providing a new theoretical foundation for the study of rationality, intelligence, consciousness, and mind. (shrink)

The problem of generating generally capable agents is an important frontier in artificial intelligence (AI) research. Such agents may demonstrate open-ended, versatile, and diverse modes of expression, similar to humans. We interpret the work of Heintz & Scott-Phillips as a minimal sufficient set of socio-cognitive biases for the emergence of generally expressive AI, separate yet complementary to existing algorithms.

This paper argues that intelligence can be approximated by the ability to produce accurate predictions. It is further argued that general intelligence can be approximated by context dependent predictive abilities combined with the ability to use working memory to abstract away contextual information. The flexibility associated with general intelligence can be understood as the ability to use selective attention to focus on specific aspects of sensory impressions to identify patterns, which can then be used to predict events in novel situations (...) and environments. The argumentation synthesizes Godfrey-Smith’s environmental complexity theory, adding the notion of niche broadness as well as changes concerning the view of cognition and control, and Hohwy’s predictive mind theory, making explicit the significance of accuracy as a composite of trueness and precision where the nervous system acts as a distributed controller motivating actions that keep the body in homeostasis. (shrink)

Humans routinely make inferences about both the contents and the workings of other minds based on observed actions. People consider what others want or know, but also how intelligent, rational, or attentive they might be. Here, we introduce a new methodology for quantitatively studying the mechanisms people use to attribute intelligence to others based on their behavior. We focus on two key judgments previously proposed in the literature: judgments based on observed outcomes (you're smart if you won the game) and (...) judgments based on evaluating the quality of an agent's planning that led to their outcomes (you're smart if you made the right choice, even if you didn't succeed). We present a novel task, the maze search task (MST), in which participants rate the intelligence of agents searching a maze for a hidden goal. We model outcome‐based attributions based on the observed utility of the agent upon achieving a goal, with higher utilities indicating higher intelligence, and model planning‐based attributions by measuring the proximity of the observed actions to an ideal planner, such that agents who produce closer approximations of optimal plans are seen as more intelligent. We examine human attributions of intelligence in three experiments that use MST and find that participants used both outcome and planning as indicators of intelligence. However, observing the outcome was not necessary, and participants still made planning‐based attributions of intelligence when the outcome was not observed. We also found that the weights individuals placed on plans and on outcome correlated with an individual's ability to engage in cognitive reflection. Our results suggest that people attribute intelligence based on plans given sufficient context and cognitive resources and rely on the outcome when computational resources or context are limited. (shrink)

This paper explores aspects of GPT-3 that have been discussed as harbingers of artificial general intelligence and, in particular, linguistic intelligence. After introducing key features of GPT-3 and assessing its performance in the light of the conversational standards set by Alan Turing in his seminal paper from 1950, the paper elucidates the difference between clever automation and genuine linguistic intelligence. A central theme of this discussion on genuine conversational intelligence is that members of a linguistic community never merely respond “algorithmically” (...) to queries through a selective kind of pattern recognition, because they must also jointly attend and act with other speakers in order to count as genuinely intelligent and trustworthy. This presents a challenge for systems like GPT-3, because representing the world in a way that makes conversational common ground salient is an essentially collective task that we can only achieve jointly with other speakers. Thus, the main difficulty for any artificially intelligent model of conversation is to account for the communicational intentions and motivations of a speaker through joint attention. These joint motivations and intentions seem to be completely absent from the standard way in which systems like GPT-3 and other artificial intelligent systems work. This is not merely a theoretical issue. Since GPT-3 and future iterations of similar systems will likely be available for commercial use through application programming interfaces, caution is needed regarding the risks created by these systems, which pass for “intelligent” but have no genuine communicational intentions, and can thereby produce fake and unreliable linguistic exchanges. (shrink)

Decision theorists disagree about how instrumentally rational agents, i.e., agents trying to achieve some goal, should behave in so-called Newcomb-like problems, with the main contenders being causal and evidential decision theory. Since the main goal of artificial intelligence research is to create machines that make instrumentally rational decisions, the disagreement pertains to this field. In addition to the more philosophical question of what the right decision theory is, the goal of AI poses the question of how to implement any given (...) decision theory in an AI. For example, how would one go about building an AI whose behavior matches evidential decision theory’s recommendations? Conversely, we can ask which decision theories describe the behavior of any existing AI design. In this paper, we study what decision theory an approval-directed agent, i.e., an agent whose goal it is to maximize the score it receives from an overseer, implements. If we assume that the overseer rewards the agent based on the expected value of some von Neumann–Morgenstern utility function, then such an approval-directed agent is guided by two decision theories: the one used by the agent to decide which action to choose in order to maximize the reward and the one used by the overseer to compute the expected utility of a chosen action. We show which of these two decision theories describes the agent’s behavior in which situations. (shrink)

Purpose The purpose of this paper is to present the conceptual model of an innovative methodology (SAT) to assess the social acceptance of technology, especially focusing on artificial intelligence (AI)-based technology. -/- Design/methodology/approach After a review of the literature, this paper presents the main lines by which SAT stands out from current methods, namely, a four-bubble approach and a mix of qualitative and quantitative techniques that offer assessments that look at technology as a socio-technical system. Each bubble determines the social (...) variability of a cluster of values: User-Experience Acceptance, Social Disruptiveness, Value Impact and Trust. -/- Findings The methodology is still in development, requiring further developments, specifications and validation. Accordingly, the findings of this paper refer to the realm of the research discussion, that is, highlighting the importance of preventively assessing and forecasting the acceptance of technology and building the best design strategies to boost sustainable and ethical technology adoption. -/- Social implications Once SAT method will be validated, it could constitute a useful tool, with societal implications, for helping users, markets and institutions to appraise and determine the co-implications of technology and socio-cultural contexts. -/- Originality/value New AI applications flood today’s users and markets, often without a clear understanding of risks and impacts. In the European context, regulations (EU AI Act) and rules (EU Ethics Guidelines for Trustworthy) try to fill this normative gap. The SAT method seeks to integrate the risk-based assessment of AI with an assessment of the perceptive-psychological and socio-behavioural aspects of its social acceptability. (shrink)

According to how AI has defined itself from its beginning, thinking in non-living matter, i.e., without life, is possible. The premise of symbolic AI is that operating on representations of reality machines can understand it. When this assumption did not work as expected, the mathematical model of the neuron became the engine of artificial “brains.” Connectionism followed. Currently, in the context of Machine Learning success, attempts are made at integrating the symbolic and connectionist paths. There is hope that Artificial General (...) Intelligence (AGI) performance can be achieved. As encouraging as neuro-symbolic AI seems to be, it remains unclear whether AGI is actually a moving target as long as AI itself remains ambiguously defined. This paper makes the argument that the intelligence of machines, expressed in their performance, reflects how adequate the means used for achieving it are. Therefore, energy use and the amount of data necessary qualify as a good metric for comparing natural and artificial performance. (shrink)

In the last 20 years the Turing test has been left further behind by new developments in artificial intelligence. At the same time, however, these developments have revived some key elements of the Turing test: imitation and adversarialness. On the one hand, many generative models, such as generative adversarial networks, build imitators under an adversarial setting that strongly resembles the Turing test. The term “Turing learning” has been used for this kind of setting. On the other hand, AI benchmarks are (...) suffering an adversarial situation too, with a ‘challenge-solve-and-replace’ evaluation dynamics whenever human performance is ‘imitated’. The particular AI community rushes to replace the old benchmark by a more challenging benchmark, one for which human performance would still be beyond AI. These two phenomena related to the Turing test are sufficiently distinctive, important and general for a detailed analysis. This is the main goal of this paper. After recognising the abyss that appears beyond superhuman performance, we build on Turing learning to identify two different evaluation schemas: Turing testing and adversarial testing. We revisit some of the key questions surrounding the Turing test, such as ‘understanding’, commonsense reasoning and extracting meaning from the world, and explore how the new testing paradigms should work to unmask the limitations of current and future AI. Finally, we discuss how behavioural similarity metrics could be used to create taxonomies for artificial and natural intelligence. Both testing schemas should complete a transition in which humans should give way to machines—not only as references to be imitated but also as judges—when pursuing and measuring machine intelligence. (shrink)

Deep learning is a kind of machine learning which happens in a certain type of artificial neural networks called deep networks. Artificial deep networks, which exhibit many similarities with biological ones, have consistently shown human-like performance in many intelligent tasks. This poses the question whether this performance is caused by such similarities. After reviewing the structure and learning processes of artificial and biological neural networks, we outline two important reasons for the success of deep learning, namely the extraction of successively (...) higher level features and the multiple layer structure, which are closely related to each other. Then some indications about the framing of this heated debate are given. After that, an assessment of the value of artificial deep networks as models of the human brain is given from the similarity perspective of model representation. Finally, a new version of computational functionalism is proposed which addresses the specificity of deep neural computation better than classic, program based computational functionalism. (shrink)

This paper considers the question: In what ways can artificial intelligence assist with interdisciplinary research for addressing complex societal problems and advancing the social good? Problems such as environmental protection, public health, and emerging technology governance do not fit neatly within traditional academic disciplines and therefore require an interdisciplinary approach. However, interdisciplinary research poses large cognitive challenges for human researchers that go beyond the substantial challenges of narrow disciplinary research. The challenges include epistemic divides between disciplines, the massive bodies of (...) relevant literature, the peer review of work that integrates an eclectic mix of topics, and the transfer of interdisciplinary research insights from one problem to another. Artificial interdisciplinarity already helps with these challenges via search engines, recommendation engines, and automated content analysis. Future “strong artificial interdisciplinarity” based on human-level artificial general intelligence could excel at interdisciplinary research, but it may take a long time to develop and could pose major safety and ethical issues. Therefore, there is an important role for intermediate-term artificial interdisciplinarity systems that could make major contributions to addressing societal problems without the concerns associated with artificial general intelligence. (shrink)

“Evidence” in the form of data collected and analysis thereof is fundamental to medicine, health and science. In this paper, we discuss the “evidence-based” aspect of evidence-based medicine in terms of statistical inference, acknowledging that this latter field of statistical inference often also goes by various near-synonymous names—such as inductive inference (amongst philosophers), econometrics (amongst economists), machine learning (amongst computer scientists) and, in more recent times, data mining (in some circles). Three central issues to this discussion of “evidence-based” are (i) (...) whether or not the statistical analysis can and/or should be objective and/or whether or not (subjective) prior knowledge can and/or should be incorporated, (ii) whether or not the analysis should be invariant to the framing of the problem (e.g. does it matter whether we analyse the ratio of proportions of morbidity to non-morbidity rather than simply the proportion of morbidity?), and (iii) whether or not, as we get more and more data, our analysis should be able to converge arbitrarily closely to the process which is generating our observed data. For many problems of data analysis, it would appear that desiderata (ii) and (iii) above require us to invoke at least some form of subjective (Bayesian) prior knowledge. This sits uncomfortably with the understandable but perhaps impossible desire of many medical publications that at least all the statistical hypothesis testing has to be classical non-Bayesian—i.e. it is not permitted to use any (subjective) prior knowledge. (shrink)

Hopes for biasing the odds towards the development of AGI that is human-friendly depend on finding and employing ethical theories and practices that can be incorporated successfully in the construction; programming and/or developmental growth; education and mature life world of future AGI. Mainstream ethical theories are ill-adapted for this purpose because of their mono-logical decision procedures which aim at “Golden rule” style principles and judgments which are objective in the sense of being universal and absolute. A much more helpful framework (...) for ethics is provided by a dialogical approach using conflict resolution and negotiation methods; a “Rainbow rule” approach to diversity; and a notion of objectivity as emergent impartiality. This conflict resolution approach will also improve our chances in dealing with two other problems related to the “Friendly AI” problem; the difficulty of programming AI to be not merely smarter but genuinely wiser and the dilemmas that arise in considering whether AGIs will be Friendly to humans out of mere partisanship or out of genuine intent to promote the Good. While these issues are challenging; a strategy for pursuing and promoting research on them can be articulated and basic legislation and corporate policies can be adopted to encourage their development as part of the project of biasing the odds in favor of Friendly and Wise AGI. (shrink)