The scientific method is grounded on the concept of causation and intelligent agency. In this project, we aim to give a philosophically coherent, theoretically justified and experimentally validated defence of the thesis that to be a cause, information must be embodied and acquires causal power when intelligent agents use embodied information to optimise their performance in a physical setting. The project will address big questions including: Does information have causal power? How does one understand causal relations in a quantum world? Are there new quantum architectures for control, based on automated intelligent agents (IA) with access to the full power of quantum information processing?

The project is structured around three integrated investigations, philosophical, theoretical and experimental. The theoretical investigation will use quantum information theory to find communication tasks that lead to novel causal structures, determine if a quantum intelligent agent (IA) can learn more efficiently than a classical IA, and design linear optical schemes for the experiment. The philosophical component will explore the extent to which quantum information throws light on how we should understand causation and will integrate the theory and experimental results to provide a coherent defence of the thesis that information has causal power in quantum physics. At the end of the project we will have discovered—and experimentally implemented—a set of statistical inequalities for testing quantum causality and a new class of quantum control protocols incorporating quantum computation.