Man and Machine

by Brian R. Duffy

The primary research challenges of the Anthropos project are to understand the relationship and to establish a bond between man and machine. Can the illusion of life and intelligence emerge through simply engaging people in social interaction with a robot?

AI research to date has led to a wealth of strategies for solving local problems with varying degrees of success. The next step has been to integrate these tools into whole systems, such as robotics, to realise a system that displays a breadth of 'intelligent' behaviour. This integration proved to be nontrivial and highlighted issues with classical AI such as its strong representation approaches and often narrow take on intelligence. In looking to extend beyond traditional paradigms in developing an artificial form of 'intelligence', the field of social robotics has recently emerged.

An impressive display of human intelligence is in our ability to adapt in social scenarios. In fact, social intelligence has been theorised as the mechanism from which the capability for the complexity of human languages evolved. Thus, one of the purposes of Media Lab Europe's research on socially-capable robots is to investigate human social intelligence and behaviour. From a human-machine interaction perspective, a socially-capable robot facilitates our access to the digital world through intuitive social mechanisms to improve or provide alternative approaches in education, information dissemination, and our future daily interactions with machines. Once domestic robots become more than just a washing machine in the home, our social interaction with them becomes inevitable. The issue is how do we realise such a socially capable robot?

Media Lab Europe's 'JoeRobot' at the Flutterfugue performance with SmartLab and NYU CATLab in London 2002. Photo courtesy of Brent Jones

In order to begin answering these questions, the Anthropos Project at Media Lab Europe seeks to decompose the interaction issues between man and machine. The current research areas are:

• Balancing function and form for social robots. Anthropos and JoeRobot (Figure 1) are prototypes built to explore the development of socially capable robots. Key to the notion of expandability and rapid prototyping, a modular nervous systems strategy employs standardised interface protocols (Firewire and USB) for actuator and perceptor components. Research on integrating a socially capable robot into performance spaces has demonstrated the power of the form as an interface to the digital information domain (figure 1). People's willingness to engage with a machine that judiciously employs anthropomorphic features we are familiar with in social contexts facilitates man-machine interaction.
• Strength and degree of minimal expression and communication. The Emotion Robots work is a series of experiments to investigate how minimal the set of humanlike features can be for a social robot. Data illustrates people's propensity to attribute such concepts as emotions and intelligence to machines performing computationally simple behaviours.
• The seamless integration of real physical worlds and digital information space. The Agent Chameleons project strives to develop digital 'minds' that can seamlessly migrate, mutate and evolve on their journey between and within physical and digital information spaces. This challenges the traditional boundaries between the physical and the virtual through the empowerment of mobile agents. Three key attributes mutation, migration and evolution underpin the Agent Chameleons concept where digital personal assistants are developed that opportunistically migrate and choose a 'body' (whether a robot, an avatar in virtual reality, an animated character on a PDA, or a web agent) to facilitate its intentions.

The Anthropos Project draws this work together to look at developing a different perspective on what an artificially intelligent entity could become. Machines have intrinsic properties that are often seen as hindrances when the reference is either humans or other biological entities. The objective is to embrace those aspects that are constructive and integrate these with a machine's inherent advantages, ie being a machine. While arguments have prevailed for many years over the nature of intelligence and whether it can be realised in a machine, this work aims to demonstrate the power of perceived intelligence and people's willingness to interpret a social robots interactions according to human-like social references. The key issue becomes a balance between function and form.

Links:
Anthropos Project at Media Lab Europe: http://anthropos.mle.ie

Please contact:
Brian R. Duffy, Media Lab Europe, Dublin, Ireland
Tel: +353 1 474 2823
E-mail: brd@media.mit.edu