The advances in communication technology have allowed us to create fully automated markets that can function twenty-four hours a day using software agents as proxies to decrease transaction costs and increase efficiency. In this paper I examine the short-term and long-term behavioral effects of these agents on human market participants using experiments with economically-motivated human subjects. Findings indicate that humans try to outsmart the less flexible agents, which leads to lower efficiency levels initially. This behavior disappears towards the end of the experiment, and superior efficiency levels are reached.