How attitudes generated by humanoid robots shape human brain activity

Artificial cognition vs. artificial intelligence for next-generation autonomous robotic agents

The trend in industrial/service robotics is to develop robots that can cooperate with people, interacting with them in an autonomous, safe and purposive way. These are the fundamental elements characterizing the fourth and the fifth industrial revolutions (4IR, 5IR): the crucial innovation is the adoption of intelligent technologies that can allow the development of cyber-physical systems, similar if not superior to humans. The common wisdom is that intelligence might be provided by AI (Artificial Intelligence), a claim that is supported more by media coverage and commercial interests than by solid scientific evidence. AI is currently conceived in a quite broad sense, encompassing LLMs and a lot of other things, without any unifying principle, but self-motivating for the success in various areas. The current view of AI robotics mostly follows a purely disembodied approach that is consistent with the old-fashioned, Cartesian mind-body dualism, reflected in the software-hardware distinction inherent to the von Neumann computing architecture. The working hypothesis of this position paper is that the road to the next generation of autonomous robotic agents with cognitive capabilities requires a fully brain-inspired, embodied cognitive approach that avoids the trap of mind-body dualism and aims at the full integration of Bodyware and Cogniware. We name this approach Artificial Cognition (ACo) and ground it in Cognitive Neuroscience. It is specifically focused on proactive knowledge acquisition based on bidirectional human-robot interaction: the practical advantage is to enhance generalization and explainability. Moreover, we believe that a brain-inspired network of interactions is necessary for allowing humans to cooperate with artificial cognitive agents, building a growing level of personal trust and reciprocal accountability: this is clearly missing, although actively sought, in current AI. The ACo approach is a work in progress that can take advantage of a number of research threads, some of them antecedent the early attempts to define AI concepts and methods. In the rest of the paper we will consider some of the building blocks that need to be re-visited in a unitary framework: the principles of developmental robotics, the methods of action representation with prospection capabilities, and the crucial role of social interaction.

Humanoid facial expressions as a tool to study human behaviour

Besides action vitality forms, facial expressions represent another fundamental social cue which enables to infer the affective state of others. In the present study, we proposed the iCub robot as an interactive and controllable agent to investigate whether and how different facial expressions, associated to different action vitality forms, could modulate the motor behaviour of participants. To this purpose, we carried out a kinematic experiment in which 18 healthy participants observed video-clips of the iCub robot performing a rude or gentle request with a happy or angry facial expression. After this request, they were asked to grasp an object and pass it towards the iCub robot. Results showed that the iCub facial expressions significantly modulated participants motor response. Particularly, the observation of a happy facial expression, associated to a rude action, decreased specific kinematic parameters such as velocity, acceleration and maximum height of movement. In contrast, the observation of an angry facial expression, associated to a gentle action, increased the same kinematic parameters. Moreover, a behavioural study corroborated these findings, showing that the perception of the same action vitality form was modified when associated to a positive or negative facial expression.

Human- or object-like? Cognitive anthropomorphism of humanoid robots

Across three experiments (N = 302), we explored whether people cognitively elaborate humanoid robots as human- or object-like. In doing so, we relied on the inversion paradigm, which is an experimental procedure extensively used by cognitive research to investigate the elaboration of social (vs. non-social) stimuli. Overall, mixed-model analyses revealed that full-bodies of humanoid robots were subjected to the inversion effect (body-inversion effect) and, thus, followed a configural processing similar to that activated for human beings. Such a pattern of finding emerged regardless of the similarity of the considered humanoid robots to human beings. That is, it occurred when considering bodies of humanoid robots with medium (Experiment 1), high and low (Experiment 2) levels of human likeness. Instead, Experiment 3 revealed that only faces of humanoid robots with high (vs. low) levels of human likeness were subjected to the inversion effects and, thus, cognitively anthropomorphized. Theoretical and practical implications of these findings for robotic and psychological research are discussed.

The influence of vitality forms on action perception and motor response

During the interaction with others, action, speech, and touches can communicate positive, neutral, or negative attitudes. Offering an apple can be gentle or rude, a caress can be kind or rushed. These subtle aspects of social communication have been named vitality forms by Daniel Stern. Although they characterize all human interactions, to date it is not clear whether vitality forms expressed by an agent may affect the action perception and the motor response of the receiver. To this purpose, we carried out a psychophysics study aiming to investigate how perceiving different vitality forms can influence cognitive and motor tasks performed by participants. In particular, participants were stimulated with requests made through a physical contact or vocally and conveying rude or gentle vitality forms, and then they were asked to estimate the end of a passing action observed in a monitor (action estimation task) or to perform an action in front of it (action execution task) with the intention to pass an object to the other person presented in the video. Results of the action estimation task indicated that the perception of a gentle request increased the duration of a rude action subsequently observed, while the perception of a rude request decreased the duration of the same action performed gently. Additionally, during the action execution task, accordingly with the perceived vitality form, participants modulated their motor response.

The middle cingulate cortex and dorso-central insula: A mirror circuit encoding observation and execution of vitality forms

Vitality forms represent the different ways in which actions are performed (e.g., gentle, rude). They express the agent’s attitudes toward others. Previous data indicated that vitality forms of hand actions depend on the dorso-central insula. In the present study, we show that in addition to the insula, the middle cingulate cortex is also involved in hand action modulation. A voxel-based analysis highlighted that voxels showing a similar BOLD signal trend in both action observation and execution are present in both regions. Using a multifiber tractography investigation, we demonstrated that the dorso-central insula and middle cingulate cortex are anatomically connected. These data indicate that the modulation of the parieto-frontal circuit controlling hand actions relies on both the insula and cingulate sectors.

Actions with identical goals can be executed in different ways (gentle, rude, vigorous, etc.), which D. N. Stern called vitality forms [D. N. Stern, Forms of Vitality Exploring Dynamic Experience in Psychology, Arts, Psychotherapy, and Development (2010)]. Vitality forms express the agent’s attitudes toward others. In a series of fMRI studies, we found that the dorso-central insula (DCI) is the region that is selectively active during both vitality form observation and execution. In one previous experiment, however, the middle cingulate gyrus also exhibited activation. In the present study, in order to assess the role of the cingulate cortex in vitality form processing, we adopted a classical vitality form paradigm, but making the control condition devoid of vitality forms using jerky movements. Participants performed two different tasks: Observation of actions performed gently or rudely and execution of the same actions. The results showed that in addition to the insula, the middle cingulate cortex (MCC) was strongly activated during both action observation and execution. Using a voxel-based analysis, voxels showing a similar trend of the blood-oxygen-level-dependent (BOLD) signal in both action observation and execution were found in the DCI and in the MCC. Finally, using a multifiber tractography analysis, we showed that the active sites in MCC and DCI are reciprocally connected.

Affective Contagion: How Attitudes Expressed by Others Influence Our Perception of Actions

Vitality forms represent a fundamental aspect of social interactions by characterizing how actions are performed and how words are pronounced on the basis of the attitude of the agent. Same action, such as a handshake, may have a different impact on the receiver when it is performed kindly or vigorously, and similarly, a gentle or rude tone of voice may have a different impact on the listener. In the present study, we carried out two experiments that aimed to investigate whether and how vocal requests conveying different vitality forms can influence the perception of goal-directed actions and to measure the duration of this effect over time. More specifically, participants were asked to listen to the voice of an actor pronouncing "give me" in a rude or gentle way. Then, they were asked to observe the initial part of a rude or a gentle passing action, continue it mentally, and estimate the time of its completion. Results showed that the perception of different vitality forms expressed by vocal requests influenced the estimation of action duration. Moreover, we found that this effect was limited to a certain time interval (800 ms), after which it started to decay.

Communicative And Affective Components in Processing Auditory Vitality Forms: An fMRI Study

In previous studies on auditory vitality forms, we found that listening to action verbs pronounced gently or rudely, produced, relative to a neutral robotic voice, activation of the dorso-central insula. One might wonder whether this insular activation depends on the conjunction of action verbs and auditory vitality forms, or whether auditory vitality forms are sufficient per se to activate the insula. To solve this issue, we presented words not related to actions such as concrete nouns (e.g.,“ball”), pronounced gently or rudely. No activation of the dorso-central insula was found. As a further step, we examined whether interjections, i.e., speech stimuli conveying communicative intention (e.g., “hello”), pronounced with different vitality forms, would be able to activate, relative to control, the insula. The results showed that stimuli conveying a communicative intention, pronounced with different auditory vitality forms activate the dorsal-central insula. These data deepen our understanding of the vitality forms processing, showing that insular activation is not specific to action verbs, but can be also activated by speech acts conveying communicative intention such as interjections. These findings also show the intrinsic social nature of vitality forms because activation of the insula was not observed in the absence of a communicative intention.

The neural bases of tactile vitality forms and their modulation by social context

People communicate using speech, gestures, and, less frequently, touches. An example of tactile communication is represented by handshake. Customs surrounding handshake vary in different cultures. In Western societies is mostly used when meeting, parting, as a sign of congratulations or at the end of a successful business. Despite its importance in social life, the neural mechanism underlying the affective components conveyed by handshake ("tactile vitality forms") is unknown. Here we combined functional magnetic resonance imaging (fMRI) and electromyography (EMG), to investigate the neural affective activations during handshakes. We demonstrated that handshake conveying gentle or aggressive tactile vitality forms produces a stronger activation of the dorso-central insula. The simultaneous presence of emotional facial expressions modulates the activation of this insular sector. Finally, we provide evidence that the cingulate cortex is involved in the processing of facial expressions conveying different vitality forms.