Game against the machine : interacting with artificial economic agents

Nowadays artificial agents (robots, artificial intelligences, etc.) are part of our societies and we interact with them daily. Hybrid societies, in which human and artificial agents interact are already reality and may become the norm not the exception. Little research has been done in general with regard to hybrid societies and even less so by economists. With the help of the experimental method we try to shed light on one aspect that could make hybrid societies different from human-only societies: The expectations that humans have with regard to the different types of agents (human or artificial) they may interact with. The experiments presented in this dissertation all compare the behaviour of humans in treatments where they expect to be relying on other humans, with treatments where they rely on "non-human" mechanisms. All experiments indicate that humans condition their behaviour on the type of agent they rely on and that behaviour differs.