From autistic to social agents

By F. Dignum, Rui Prada, G. Hofstede. 2014

52 citations recorded by the source corpus.

The theory, models and architectures of intelligent agents are based loosely on the theory of intentions from Bratman resulting in the so-called BDI agents. Although this functions well for single agents it has been long recognized that this approach falls short for multi-agent systems. It lacks appropriate social aspects to make natural interaction possible. The original concept for intelligent agents was based on a (simple) idea of how people reason about actions. We propose that we go back to the foundation and acknowledge that people are in the core social beings. I.e. we don't function as rational agents with the addition of some "sociality" modules to make us aware of other people. Rather we are social at the base and this sociality pervades all our reasoning, motivation, and any other aspect of our behavior. In this paper we propose a new set of core cognitive elements to replace the BDI approach and discuss the paradigm of a social landscape. However, although we aim for a radical change in the way the community creates social agents we also believe that the new approach incorporates previous work, such as BDI. Our claim is that deliberation about actions and BDI are certainly a part of how agents cope with a dynamic world, but are not the core part of social agents that are part of a social world interacting with other agents and humans in a natural way.

Subfields: Artificial Intelligence, Cognitive Neuroscience, Social Psychology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Cognitive Neuroscience

In Social Psychology

Citations in the corpus

Explanation in Artificial Intelligence: Insights from the Social Sciences (2017, background): “While the idea of Theory of Mind in AI is not new; see for example [177, 37]; it’s application to explanation has not been adequately explored.”
Practical Specification of Belief Manipulation in Games (2021, background): “…challenge of AI: cognitive scientists have linked it to social intelligence in humans (Baron-Cohen 1997; Wimmer and Perner 1983), a phenomenon that AI researchers have seen as critical for building collaborative systems from autonomous agents (Castelfranchi 1998; Dignum, Prada, and Hofstede 2014).”
Integrating social power into the decision-making of cognitive agents (2016, background): “As part of this progress, in recent years, there has been increasing interest in social concepts that may contribute to new advances concerning social intelligence and believability in agents [17,24].”