Spoke Overview

Current  AI techniques address success-specific tasks. We need to overcome the barrier of complexity that prevents addressing many real-world situations involving autonomous situated agents, their interaction with the environment, and the cooperation among them and with humans. The integration of heterogeneous computational representation and reasoning techniques, such as symbolic and sub-symbolic representations, different learning paradigms, and models of human-machine cooperation and interaction: 

• Embodied AI able to work in physical environments by integrating model based reasoning and learning from data and human behaviour;

• Interactive AI able to interact with people like humans do and proactively operate to meet people's needs;

• Cooperative AI able to understand the dynamics of society, to work in a community of cooperative humans and artificial agents.