Invited Talk @ TU Berlin / SCIoI: Hybrid AI for Safe Embodied Intelligence

Enabling robots to perform complex manipulation tasks in open-world environments requires approaches to intelligence that bridge the traditional divide between symbolic and subsymbolic systems. Hybrid systems that combine reasoning over symbolic knowledge with probabilistically approximately correct inference over probability distributions promise to make robot manipulation actions generalizable, interpretable, and scalable to address real-world challenges. This talk explores how hybrid AI systems can enable robots to become not just capable agents, but also safe colleagues and companions for humans.

This talk was part of a lecture series at the Science of Intelligence (SCIoI) Collaborative Research Cluster, facilitated by the RBO (Robotics and Biology Lab) at the TU Berlin.