Priyanka

Artificial intelligence (AI) has greatly changed the world by making tasks faster and more efficient. AI agents are designed to imitate human tasks but the extent to which they truly replicate human qualities remains an open question. Can these agents possess conscience like human? In general, the answer is no. However, can a form of conscience emerge inside them without explicitly programmed to do so? This study seeks the answer of that possibility. Specifically, it examines whether two agents competing in a game can develop cooperative behavior by recognizing mutual benefits and forming a moral foundation. By analyzing game outcomes and the agents’ explanations for their actions, this study aims to investigate whether moral intuitions can emerge and evolve in artificial agents.

Artificial Intelligence (AI) agents have become a fixture of modern life, often serving as the first point of contact in customer service, assistance and decision-making roles that were once exclusively human. While these agents do not possess conscience, this study, a continuation of Practicum I, asks a more subtle question; can something resembling moral intuition emerge from within an AI agent, without being directly programmed? To explore this, it observes two AI agents, each guided by a different behavioral framing, interacting in an iterated game and examines whether they gravitate toward cooperation and if so, whether their self-generated explanations of action reflect an evolving form of moral reasoning behind those choices. At its core, this study seeks to understand whether interaction can drive changes in how an agent behaves and expresses moral thought.