Human-Agent Interaction: Challenges for Bringing Humans and Agents Together Rui Prada and Ana Paiva INESC-ID and Instituto Superior T´ ecnico, Universidade de Lisboa Av. Prof. Doutor An´ ıbal Cavaco Silva 2744-016 Porto Salvo, Portugal http://www.gaips.inesc-id.pt Abstract. There has been an increasing interest in the area of Human- Agent Interaction (HAI), as demonstrated by the recent organisation of workshops and conferences dedicated solely to this topic (and this work- ship is a very good example of this). However, as illustrated by the topics in some of these events, there are different visions of what the area of Human-Agent Interaction is. In this paper we argue that while human computer interaction (HCI) studies the interaction between humans and computers, the field of Human-Agent Interaction should be focused on the systems, models, applications and methodologies that allow for hu- mans and “agents” to interact, collaborate, negotiate and even become peers. Differently from HCI in HAI we have “autonomy” on the side of the machines, making the interaction much richer, decentralised and emergent. With this idea in mind, in this paper we try to draw some challenges to the field, which we need to address if we aim at having humans and agents together interacting as peers. 1 Introduction The world is changing and autonomous technological artefacts are now becoming a reality. Self-driving cars, autonomous vacuum cleaners, nurse robots, or virtual sales-people can now act autonomously to support human activities. Yet, as these new technologies enter our social world, they must be able to interact with humans in a natural way. There has been a recent interest in the area of Human-Agent Interaction (HAI), as demonstrated by the recent organisation of workshops, such as this particular one 1 and conferences such as the HAI and the HRI, 2 dedicated solely to this topic. Although, at first glance it may appear that this topic is fairly recent in the Intelligent Agents and Multiagent’s community, this is not the case, and many avenues have been investigated by researchers over the past twenty years. In fact the Autonomous Agents Conference (one of the predecessors of the AAMAS conference) that started in 1997, had several papers in Human- Agent Interaction, devoting a few sessions to that topic (see for example papers 1 See for example http://haidm.wordpress.com/ 2 See for example http://hai-conference.net/ihai2013/ or the well know HRI seris in http://humanrobotinteraction.org/