Trusted, Safe, and Distributed Artificial Intelligence in Lifelong Machine Learning Tasks
Smart machines that continues to learn forever run the risk that they may learn undesirable behaviours. The aim of this project is to ensure this does not happen. This requires the design of safety nets in place to ensure that the artificial intelligence will continue to operate in a trusted manner and continues to be safe. Adding to the challenge is the need to design these systems in a distributed manner. The candidate will work closely with Prof. Hussein Abbass (www.husseinabbass.net) team at UNSW-Canberra and will contribute safety assurance algorithms to the operations of swarm intelligence systems that continue to learn for life.