Is imitation learning the route to humanoid robots?

This review investigates two recent developments in artificial intelligence and neural computation: learning from imitation and the development of humanoid robots. It is postulated that the study of imitation learning offers a promising route to gain new insights into mechanisms of perceptual motor control that could ultimately lead to the creation of autonomous humanoid robots. Imitation learning focuses on three important issues: efficient motor learning, the connection between action and perception, and modular motor control in the form of movement primitives. It is reviewed here how research on representations of, and functional connections between, action and perception have contributed to our understanding of motor acts of other beings. The recent discovery that some areas in the primate brain are active during both movement perception and execution has provided a hypothetical neural basis of imitation. Computational approaches to imitation learning are also described, initially from the perspective of traditional AI and robotics, but also from the perspective of neural network models and statistical-learning research. Parallels and differences between biological and computational approaches to imitation are highlighted and an overview of current projects that actually employ imitation learning for humanoid robots is given.     

From embodied to socially embedded agents – Implications for interaction-aware robots

First, this article proposes a minimal definition of embodiment that can be applied across animals and artefacts. We discuss the potential contributions of this operational definition with respect to assessing and measuring the degree of embodiment in different biological and artificial systems. Second, we outline how this definition can be extended to lead to the particular notion of social embeddedness. Socially embedded agents are structurally coupled with their social environment, in that their sensorimotor activity is grounded in the social environment that the agent is surrounded by. Lastly, based on research in the social sciences on human–human interaction, we discuss perceptual requirements for interaction-aware robotic agents—agents whose identification and interpretation of the (social) environment is facilitated by awareness of the structure of agent–agent interactions (including humans ‘in the loop’). We suggest relevant concepts and heuristics that can contribute to studies of degrees of embodiment of robots that interact with social environments. Manipulating and systematically investigating these heuristics permits variation of the degree of embodiment of such interaction-aware robots.     

Animals as cost‐based robots

The frame problem is a problem that arises when an agent attempts to assess the consequences of future behaviour. Strictly, it is a problem of modelling that arises during planning. The problem arises because many of the possible consequences of a planned action are not really relevant to the decision whether to perform the action. The frame problem is typical of the classical approach to artificial intelligence, but it is evident that animals do not suffer from this problem. In this paper it is suggested that animals can circumvent the frame problem because their decision‐making architecture is very different from that traditionally used in artificial intelligence.     

Could robots be phenomenally conscious?

In a recent book (Tye 2017), Michael Tye argues that we have reason to attribute phenomenal consciousness to functionally similar robots like commander Data of Star Trek. He relies on a kind of inference to the best explanation – ‘Newton’s Rule’, as he calls it. I will argue that Tye’s liberal view of consciousness attribution fails for two reasons. First, it leads into an inconsistency in consciousness attributions. Second, and even more importantly, it fails because ceteris is not paribus. The big, categorical difference in history between Data-like robots on the one hand and human beings on the other hand defeats the ceteris paribus assumption, which can be seen by various considerations. So the inference rule cannot be applied. We should not attribute phenomenal consciousness to robots like Data.     

If robots R–US,who am I: Online ‘Christian’ responses to artificial intelligence

In this paper, I present responses to GRIN technologies drawn from Christian posts on the Internet, to answer the following research questions: 1. Which approach to imago Dei informs the online Christian response to artificial intelligence and artificial life (AI/AL)? and 2. To what extent does the preference for a particular approach emerge from a desire to construct the Self? Drawing upon Herzfeld (2002) discussion of AI and theological approaches to understanding the meaning of imago Dei, or humankind made in the image of God, it is shown that the Christian response to AI and AL is elaborated mostly through the functionalist and substantive approaches. It also largely constructs the GRN future as a Hell scenario. With reference to Bhabha's postcolonial model of ambivalence, it is argued that this response is motivated by a desire to regain control over the discursive construction of the Self, which for some, is challenged by AI and AL.     

Can robots make good models of biological behaviour?

How should biological behaviour be modelled? A relatively new approach is to investigate problems in neuroethology by building physical robot models of biological sensorimotor systems. The explication and justification of this approach are here placed within a framework for describing and comparing models in the behavioural and biological sciences. First, simulation models--the representation of a hypothesis about a target system--are distinguished from several other relationships also termed "modelling" in discussions of scientific explanation. Seven dimensions on which simulation models can differ are defined and distinctions between them discussed: 1. Relevance: whether the model tests and generates hypotheses applicable to biology. 2. Level: the elemental units of the model in the hierarchy from atoms to societies. 3. Generality: the range of biological systems the model can represent. 4. Abstraction: the complexity, relative to the target, or amount of detail included in the model. 5. Structural accuracy: how well the model represents the actual mechanisms underlying the behaviour. 6. Performance match: to what extent the model behaviour matches the target behaviour. 7. Medium: the physical basis by which the model is implemented. No specific position in the space of models thus defined is the only correct one, but a good modelling methodology should be explicit about its position and the justification for that position. It is argued that in building robot models biological relevance is more effective than loose biological inspiration; multiple levels can be integrated; that generality cannot be assumed but might emerge from studying specific instances; abstraction is better done by simplification than idealisation; accuracy can be approached through iterations of complete systems; that the model should be able to match and predict target behaviour; and that a physical medium can have significant advantages. These arguments reflect the view that biological behaviour needs to be studied and modelled in context, that is, in terms of the real problems faced by real animals in real environments.     

The early ontogeny of infants’ imitation of on screen humans and robots

Traditionally, infants have learned how to interact with objects in their environment through direct observations of adults and peers. In recent decades these models have been available over different media, and this has introduced non-human agents to infants’ learning environments. Humanoid robots are increasingly portrayed as social agents in on screen, but the degree to which infants are capable of observational learning from screen-based robots is unknown. The current study thus investigated how well 1- to 3-year-olds (N = 230) could imitate on-screen robots relative to on-screen and live humans. Participants exhibited an imitation deficit for robots that varied with age. Furthermore, the well-known video deficit did not replicate as expected, and was weak and transient relative to past research. Together, the findings documented here suggest that infants are learning from media in ways that differ from past generations, but that this new learning is nuanced when novel technologies are involved.     

Is age kinder to females than to males?

To investigate possible gender differences in the patterns of age-cognition relations, a meta-analysis was conducted on the data from 25 studies (5,201 participants) by the second author and his collaborators. Gender and age differences consistent with prior reports were found, but there were few measures on which the age × gender interactions were significant. Although the majority of neuroimaging studies investigating gender differences in age-related atrophy and functional decline report greater age-related differences in males, the only significant interactions on the cognitive measures in this study (on measures of speed and reasoning) were in the direction of lesser age-related declines for males than for females.     

Do we have to be Realists about Colour in order to be able to attribute Colour Perceptions to Other Persons?

One of the main targets of Barry Stroud’s criticism in his recent book ‚The Quest for Reality. Subjectivism and the Metaphysics of Colour’ are eliminativist theories of colour which he regards as a version of the metaphysical project of the unmasking of colours (Stroud, 2000). According to this view, no physical objects have any of the colours we see them or believe them to have. However, although this error theory describes all our colour perceptions as illusory, and all our colour beliefs as false, it cannot deny that we actually perceive colours and that we do believe that physical objects are coloured. Therefore, it has to account for these psychological facts without relying on any assumptions about the colours of things. Thus, the central question for the unmasking project is whether it is possible to acknowledge someone’s perceiving a certain colour or having beliefs about the colours of things without holding that anything anywhere has any colour at all. Contrary to Stroud, this paper defends the view that we can acknowledge that people believe in colours without having ourselves to accept their existence.     

What do humans feel with mistreated humans,animals, robots,and objects? Exploring the role of cognitive empathy

Motivation and Emotion - The aim of this paper is to present a study in which we compare the degree of empathy that a convenience sample of university students expressed with humans, animals,...     

A Policy-Capturing Approach to Individuals′ Decisions to Be Absent

This study provided a within-subjects assessment of the factors associated with an individual′s decision to be absent and examined whether there were differences between individuals in their decisions. A sample of maintenance and clerical employees at a large Midwest university responded to hypothetical scenarios describing factors that might contribute to their decisions to be absent on a particular day. Illness explained mare variance than any other factor in individual′s absence decisions. Several other within-subject and between-subject influences were identified. Results from a cluster analysis suggested that the importance of these factors differed between subgroups of individuals, lending support to Johns and Nicholson′s (1982) argument that the meaning of absence is not the same for all individuals.     

Get healthy to marry or marry to get healthy?

This study tested selection and causal effects on the marriage‐health link by analyzing predictors of mortality through body mass index (BMI) and chronic illnesses (CI) on emerging adults (EA; N = 302) who were followed biannually for 8 years from the Panel Study of Income Dynamics. Results showed EA with unhealthy BMI's are more likely to marry and improve their BMI after marrying, substantiating selection and causal effects. Only the selection effect was substantiated for CI, whereby White EAs with CI were less likely to marry. Marriage appears to benefit overall health for EA, as findings suggest improved health behaviors after marriage. Future research should assess what health behaviors change during marriage for EA and if they help CI management.     

Emotional Desensitization to Violence Contributes to Adolescents’ Violent Behavior

Many adolescents are exposed to violence in their schools, communities and homes. Exposure to violence at high levels or across multiple contexts has been linked with emotional desensitization, indicated by low levels of internalizing symptoms. However, the long-term consequences of such desensitization are unknown. This study examined emotional desensitization to violence, together with externalizing problems, as mediators of the relationship between exposure to violence in pre-adolescence and violent behavior in late adolescence. A community sample of youth (N?=?704; 48 % female; 76 % African American, 22 % Caucasian) reported on their exposure to violence in multiple settings at ages 11, 13 and 18. Internalizing and externalizing problems were assessed at ages 11 and 13; violent behavior was measured at age 18. Structural Equation Modeling showed that exposure to high levels of violence at age 11 was associated with lower levels of internalizing problems (quadratic effect) at age 13, as was exposure to violence across multiple contexts (linear effect). In turn, fewer internalizing problems and more externalizing problems at age 13 predicted more violent behavior at age 18. The results suggest that emotional desensitization to violence in early adolescence contributes to serious violence in late adolescence.