Weargue that hedging is an activity that human and machine agents should engage in more broadly, even when the agent's value is not necessarily in monetary units.

Understanding how large language model (LLM) agents behave in strategic interactions is essential as these systems increasingly participate autonomously in economically and morally consequential decisions. We evaluate LLM preferences using canonical economic games, finding substantial deviations from human behavior. Models like GPT-4o show excessive cooperation and limited incentive sensitivity, while reasoning models, such as o3-mini, align more consistently with payoff-maximizing strategies. We propose a supervised fine-tuning pipeline that uses synthetic datasets derived from economic reasoning to align LLM agents with economic preferences, focusing on two stylized preference structures. In the first, utility depends only on individual payoffs (homo economicus), while utility also depends on a notion of Kantian universalizability in the second preference structure (homo moralis). We find that fine-tuning based on small datasets shifts LLM agent behavior toward the corresponding economic agent. We further assess the fine-tuned agents' behavior in two applications: Moral dilemmas involving autonomous vehicles and algorithmic pricing in competitive markets. These examples illustrate how different normative objectives embedded via realizations from structured preference structures can influence market and moral outcomes. This work contributes a replicable, cost-efficient, and economically grounded pipeline to align AI preferences using moral-economic principles.

While the philosophical literature has extensively studied how decisions relate to arguments, reasons and justifications, decision theory almost entirely ignores the latter notions and rather focuses on preference and belief. In this article, we argue that decision theory can largely benefit from explicitly taking into account the stance that decision-makers take towards arguments and counter-arguments. To that end, we elaborate a formal framework aiming to integrate the role of arguments and argumentation in decision theory and decision aid. We start from a decision situation, where an individual requests decision support. In this context, we formally define, as a commendable basis for decision-aid, this individual's deliberated judgment, popularized by Rawls. We explain how models of deliberated judgment can be validated empirically. We then identify conditions upon which the existence of a valid model can be taken for granted, and analyze how these conditions can be relaxed. We then explore the significance of our proposed framework for decision aiding practice. We argue that our concept of deliberated judgment owes its normative credentials both to its normative foundations (the idea of rationality based on arguments) and to its reference to empirical reality (the stance that real, empirical individuals hold towards arguments and counter-arguments, on due reflection). We then highlight that our framework opens promising avenues for future research involving both philosophical and decision theoretic approaches, as well as empirical implementations.

One key motivation for using contests in real-life is the substantial evidence reported in empirical contest-design literature for people's tendency to act more competitively in contests than predicted by the Nash Equilibrium. This phenomenon has been traditionally explained by people's eagerness to win and maximize their relative (rather than absolute) payoffs. In this paper we make use of "simple contests," where contestants only need to strategize on whether to participate in the contest or not, as an infrastructure for studying whether indeed more effort is exerted in contests due to competitiveness, or perhaps this can be attributed to other factors that hold also in non-competitive settings. The experimental methodology we use compares contestants' participation decisions in eight contest settings differing in the nature of the contest used, the number of contestants used and the theoretical participation predictions to those obtained (whenever applicable) by subjects facing equivalent non-competitive decision situations in the form of a lottery. We show that indeed people tend to over-participate in contests compared to the theoretical predictions, yet the same phenomenon holds (to a similar extent) also in the equivalent non-competitive settings. Meaning that many of the contests used nowadays as a means for inducing extra human effort, that are often complex to organize and manage, can be replaced by a simpler non-competitive mechanism that uses probabilistic prizes.

This paper studies two suboptimal advice provisioning methods ("advisors") as an alternative to providing optimal advice in repeated advising settings. Providing users with suboptimal advice has been reported to be highly advantageous whenever the optimal advice is non-intuitive, hence might not be accepted by the user. Alas, prior methods that rely on suboptimal advice generation were designed primarily for a single-shot advice provisioning setting, hence their performance in repeated settings is questionable. Our methods, on the other hand, are tailored to the repeated interaction case. Comprehensive evaluation of the proposed methods, involving hundreds of human participants, reveals that both methods meet their primary design goal (either an increased user profit or an increased user satisfaction from the advisor), while performing at least as good with the alternative goal, compared to having people perform with: (a) no advisor at all; (b) an advisor providing the theoretic-optimal advice; and (c) an effective suboptimal-advice-based advisor designed for the non-repeated variant of our experimental framework.

Our hypothesis is that by equipping certain agents in a multi-agent system controlling an intelligent building with automated decision support, two important factors will be increased. The first is energy saving in the building. The second is customer value---how the people in the building experience the effects of the actions of the agents. We give evidence for the truth of this hypothesis through experimental findings related to tools for artificial decision making. A number of assumptions related to agent control, through monitoring and delegation of tasks to other kinds of agents, of rooms at a test site are relaxed. Each assumption controls at least one uncertainty that complicates considerably the procedures for selecting actions part of each such agent. We show that in realistic decision situations, room-controlling agents can make bounded rational decisions even under dynamic real-time constraints. This result can be, and has been, generalized to other domains with even harsher time constraints.