A  n  t  h  r  o  p  o  s 
Social Robot Research @ Media Lab Europe

The Project Vision

Realising the notion of a robot capable of "amicability," or as a social interface between people in the physical environment and the digital world, is not a trivial task. It evokes fundamental questions both in AI and social cognitive research.

Fundamental questions

For example, it raises questions of the foundations of AI such as

The Anthropos project does not directly aim to provide a definition of "intelligence" but rather seeks to understand it indirectly through human-robot interaction.

Early AI research to date has concentrated on solving localised problems and providing tools (e.g. neural networks, fuzzy logic, genetic algorithms, multiagent systems). This has led to a wealth of strategies for solving local issues with varying degrees of success. The next step in AI research 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 a narrow take on intelligence.

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 a socially-capable robot is to investigate human social intelligence and behaviour. From a human-machine interaction perspective, a socially-capable robot also facilitates the maximal use of the digital world to improve or provide alternative approaches in education, information dissemination, and other aspects of daily life.

The Project

Designing such an interface then results in questions such as

In order to begin answering these questions, current trends in the Anthropos project are