Are robots present?

Are Robots Present? New trends from Cognitive Sciences Giuseppe Riva , Ph.D. Department of Psychology Università Cattolica del Sacro Cuore, Milan, Italy ATN-P Lab. Istituto Auxologico Italiano, Milan, Italy E-mail: [email_address] http://www.emergingcommunication.com http://www.e-psychology.net http://www.ambientintelligence.org Applied Technology for Neuro-Psychology Lab .

Summary This presentation will focus on: The new concept of embodied cognition The concept of presence: linking action and cognition How it is born and its technology background; The difficulties experienced; What we already know. A new vision of the concept of presence based on the latest neuropsychological theories and development: Its meaning for the development of VR applications for rehabilitation : A layered sense of presence; The focus on the final application. To deepen this vision : two free books (PDF) http://www.emergingcommunication.com/volume5.html http://www.emergingcommunication.com/volume8.html

The background: Creating Cognitive Robots In their recent positioning paper about the long term requirements for cognitive robots Slogan and colleagues (2006) are aware of the difficulty of the task. The first point they underline is the lack of comprehension of cognitive mechanisms : “ At present no AI systems come close to having this [cognitive] collection of capabilities. There appears to be something deep, evolutionarily old, and largely unexplored which is shared with a subset of other animals and which lies behind many characteristically human capabilities, including the ability to use language, to reason hypothetically, to think about what others are doing, to design new complex physical structures, to think about complex action sequences and compare alternatives, before executing them.” http://www.cs.bham.ac.uk/research/projects/cosy/bib-db/papers/slomanetal06cogrob.pdf

The background: Creating Cognitive Robots This statement is not so different from what McCarty wrote a decade ago (1996): “ The known facts are incomplete, and there is no a priori limitation on what facts are relevant. It may not even be decided in advance what phenomena are to be taken into account. The consequences of actions cannot be fully determined. The common sense informatic situation necessitates the use of approximate concepts that cannot be fully defined and the use of approximate theories involving them.” http://www-formal.stanford.edu/jmc/human/human.html Where is the problem? Apparently there is a problem in how cognition is defined .

A new trend: Embodied Cognition For many researchers in the area cognition is primarily a matter of performing formal operations on abstract symbols and has little or nothing to do with the environment in which it occurs. Further, action, perception, and interpretation are considered separate activities . But is this true?

A new trend: Embodied Cognition Since 2001 Ziemke has been underlying the need of a change in paradigm in this area through the introduction of the concept of embodiment . The Embodied Cognition paradigm takes as its starting point the idea that cognition occurs in specific environments, and for specific ends . Moreover, the Embodied Cognition approach underlines the central role of body in shaping the mind. Specifically, the mind has to be understood in the context of its relationship to a physical body that interacts with the world .

A new trend: Embodied Cognition Hence human cognition, rather than being centralized, abstract, and sharply distinct from peripheral input and output modules, has instead deep roots in sensorimotor processing .

The leading theories: Embodied Cognition At the core of the Embodied Cognition approach there are three of the most recent theories emerged in cognitive science: the Common Coding Theory (Hommel et al, 2001), the Situated Simulation Theory (Barsalou, 2003) and the Covert Imitation Theory (Wilson and Knoblick, 2005). According to them, our conceptual system produces dynamically contextualized representations (simulations) that support situated action in different contexts.

The leading theories: Embodied Cognition This is allowed by a common coding – the motor code – shared by perception, action and concepts (Prinz, 2004). The same code also allows the subject for natively recognizing actions done by other beings (Gallese, 2004, 2005): seeing someone grasping an apple produces a contextualized simulation of the action .

From Embodied Cognition to Presence However, this picture has a big hole in it: if perception, action and concepts share the same language how can we differentiate between them ? Here we suggest that a psychology of “presence” is a possible answer to this question. In particular we argue that the evolutive role of presence is to allow the process of self-identification through the separation between “self” and “other,” and between “internal” and “external”. Following this vision, a cognitive robot has to be “present” in the external world by progressively separating itself by it .

What is presence? The sense of ‘being there’ (Held & Durlach, Sheridan, Zeltzer: premier issue of PRESENCE, 1992) ‘ A perceptual illusion of nonmediation’ (Lombard and Ditton, 1997) ‘ A mental state in which a user feels physically present within the computer-mediated environment’ ( Draper & Kaber, 1998) ‘ The subjective experience of being in one place or environment, even when one is physically situated in another’ (Witmer & Singer, 1998)

Being There: the efforts Creating the sense of presence is a major challenge and is leading to the development of new interdisciplinary research combining cognitive psychology, haptic (sense of touch) studies, computer graphics and multimedia design, advanced communication theory and socio-cultural issues communication technologies .

The origin of presence Telepresence research (1971-1980) Local Physical Space Remote Physical Space

Being There: The holy grail A final theory of presence - emerging through this interdisciplinary research - should explore the cognitive and affective roots of sensory perception, consciousness and self . This theory could allow the design of innovative systems that offer "richer" experiences than any current media and communication technologies. However, this is not an easy task.

Presence: the efforts Moreover, we have different “presences” (Riva et al., 2004) : Physical presence refers to the sense of being physically located in mediated space; Social presence refers to the feeling of being together, of social interaction with a virtual or remotely located communication partner. At the intersection of these two categories, we can then identify Co-presence or a sense of being together in a shared space, combining significant characteristics of both physical and social presence.

Presence: What we know There is consensus that the experience of presence is a complex, multidimensional perception, formed through an interplay of raw (multi-)sensory data and various cognitive processes – an experience in which attentional factors play a crucial role as well. Two categories of variables can determine a user's presence: (i) media characteristics , and (ii) user characteristics .

Presence: What we know Characteristics of the medium can be subdivided into media form and media content variables. Both of these are known to have a significant impact on the individual’s sense of presence such that, depending on the levels of appropriate, rich, consistent, and captivating sensory stimulation, varying levels of presence can be produced.

Presence: What we know About the characteristics of the users we also know that an important role is played by the context in which the user is in. Presence is always mediated by both physical (our body, technological devices, etc.) and conceptual ( expectations, roles, etc.) tools (ARTIFACTS), that belong to a given culture: we have to develop synthetic environments in which actors may function in an ecologically valid way easily identifying what they have to do.

Presence Summary (EMMA project, 2004) Slater and Usoh (1993) “ Psychology is the physics of VR in the sense that the virtual environment is manufactured towards creating a cognitive state” ( R. Lauria, 1997 ) FACTORS CONTRIBUTING TO PRESENCE EXTERNAL FACTORS INTERNAL FACTORS COMMON PROCESSES INDIVIDUAL DIFFERENCES STRUCTURAL ASPECTS PSYCHOLOGICAL ASPECTS EMOTIONS

Our attempt Recent research in neuroscience has tried to understand human action from two different but converging perspectives: the cognitive and the volitional. On one side, cognitive studies analyze how action is planned and controlled in response to environmental conditions . On the other side, volitional studies analyze how action is planned and controlled by subject’s needs, motives and goals . In our vision (Riva et al., 2006) we suggest that presence is the missing link between these two approaches.

Presence as selection mechanism (1) Specifically we consider presence as a selective neuropsychological phenomenon, evolved from the interplay of our biological and cultural inheritance, whose goal is the enaction of the volition of the self : presence is the non mediated (prereflexive) perception of successful intentions in action . The model makes sense in terms of evolutionary psychology and is beginning to be supported by evidence of the neural and other physical correlates of action , imitation and self-monitoring.

Presence as selection mechanism (2) Presence is described here as a defining feature of the nervous system able to solve a key problem for its survival: how to differentiate between internal and external intentions . In other words, it is presence that transforms intentionality – the ability to recognize purposeful actions – into the ability of producing an intention – the agent’s mental state that drives such actions. The presence process can be described as a sophisticated but unconscious form of monitoring of action and experience , transparent to the self but critical for its existence.

Presence as selection mechanism (3) At the neural level, recent research indicates that the insula and the right inferior parietal cortex may be crucial in distinguishing the self from the other. Activation in the right inferior parietal lobe is found in reciprocal imitation. Similarly, activation of the right inferior parietal lobe correlates with the subjective sense of ownership in action execution. Further, mentally simulating the actions of another person results in activation in the right inferior parietal cortex . The main experiential outcome of this process is the sense of agency: we feel that we are both the author and the owner of our own action .

Presence as selection mechanism (4) Here we argue that is the feeling of presence that provides to the self feedback about the status of its activity : the self perceives the variations in the feeling of presence and tunes its activity accordingly. This is achieved through a : during self-produced actions a sensory simulative forward model prediction of the outcome of the action is produced along with the actual motor command. The results of the comparison (which occurs at a sub-personal level) between the sensory prediction and the sensory consequences of the act can then be used to determine both the agent of the action and to track any possible variation in its course . If no variations are perceived, the self is able to concentrate on the action and not on its monitoring .

Why is presence a distinct psychological process ? Recently the observation of mirror neurons has lead to the suggestion that there is not only a rich representation of motor action but also that this representation is used for multiple purposes: action execution, action planning, action imaging, and action recognition. Of particular importance is the observation that one agent can recognise an action plan of another one and that the same neurons are involved.

Why is presence a distinct psychological process ? Mirror neurons and canonical neurons provide neural mechanisms for the centrality of observation and imitation in the ontogeny and phylogeny of the learning and transmission of cultural tool systems. BUT we need Presence to differentiate between: The real act of doing something; Its mental representation. Without this basic distinction imitation and learning is totally impossible

Embodiment and Cognition Maturana & Varela’s (1980, 1987) definition of autopoietic systems: A living system is an autopoietic machine whose function it is to create and maintain the unity that distinguishes it from the medium in which it exists. Lakoff & Johnson (1980, 1999) argued that abstract concepts are based on metaphors grounded in bodily experience/activity .

Embodiment and Cognition These vision support a shift from: the traditional cognitive theories of selfhood assuming that self and personal identity are wholly mind-based , to an embodied approach. This new approach, headed by recent work from Damasio, finds in the evolution of the central nervous system the origin of the self.

Presence as feeling of being in an external world Presence as selection mechanism

A layered self According to Damasio, self is a layered concept :  the proto self : a coherent collection of neural patterns that map, moment by moment, the physical state of the organism;  the core self : a transient entity which is continuously generated through encounters with objects; the autobiographical self : a systematic record of the more invariant properties that the organism has discovered about itself.

Presence as layered process In particular we hypothesize that is possible to associate a specific layer of presence with each of the three levels of self identified by Damasio. Self versus non-self ( proto presence ) Self versus present external world ( core presence ) Self relative to present external world ( extended presence ) Given that each layer of presence solves a peculiar facet of the internal/external world separation, it is characterized by specific properties .

First layer: Proto Presence The evolutionary goal of the proto self is to predict the characteristics of the external world as it is experienced through sensorial inputs. Proto presence is an embodied presence related to the level of perception-action coupling (self vs. non-self) : the more the organism is able to couple correctly perceptions and movements, the more it differentiates itself from the external world, thus increasing its probability of surviving.

Second layer: Core Presence The evolutionary goal of the core self is the integration of specific sensory occurrences into single percepts. We can define the core presence as the activity of selective attention made by the self on perceptions : the more the organism can focus on its sensorial experience by leaving in the background the remaining neural processes, the more it can identify the present moment and its current tasks, rising its probability of surviving.

Third layer: Extended Presence The evolutionary goal of extended self is the shift from meaning-as-comprehensibility to meaning-as-significance. The role of extended presence is to verify the significance (“ Relevance ”, Sperber and Wilson, 1995) of the experience for the self: The more the self is present in significant experiences, the more he will be able to reach its goals, augmenting its possibility of surviving.

Presence: Reading the books The paper version (hardcover) of the books describing this vision: Being There, and From Communication to Presence can be ordered on the IOS Press web site (Emerging Communication Series): http://www.iospress.nl The electronic version of the books can be freely downloaded by the series web site together with other VR/Communication books: http://www.emergingcommunication.com

Are robots present?

More Related Content

What's hot (20)

Similar to Are robots present? (20)

More from Riva Giuseppe (20)

Recently uploaded (20)

Are robots present?