Many scenarios involve a tension between individual interest and the interests of others. Such situations are called social dilemmas. Because of their ubiquity in economic and social interactions constructing agents that can solve social dilemmas is of prime importance to researchers interested in multi-agent systems. We discuss why social dilemmas are particularly difficult, propose a way to measure the 'success' of a strategy, and review recent work on using deep reinforcement learning to construct agents that can do well in both perfect and imperfect information bilateral social dilemmas.

We present a brief overview of learning dynamics for anti-coordination in ad-hoc scenarios. Specifically, we consider multi-armed bandit algorithms, reinforcement learning, and symmetric strategies for the repeated resource allocation game. In a multi-agent system with dynamic population where every agent is able to learn, the anti-coordination problem exhibits unique challenges. Thus, it is essential for the success of a joint plan that the agents can quickly and robustly learn their optimal behavior. In this work we will focus on convergence rate, efficiency, and fairness in the final outcome.

This paper describes the architecture that integrates DALI MASs (Multi-Agent Systems) and ASP (Answer Set Programming) modules for reaching goals in a flexible and timely way, where DALI is a computational-logic-based fully implemented agent-oriented logic programming language and ASP modules includes solvers that allow affordable and flexible planning capabilities. The proposed DALI MAS architecture exploits such modules for flexible goal decomposition and planning, with the possibility to select plans according to a suite of possible preferences and to re-plan upon need. We present an abstract case-study concerning DALI agents which cooperate for exploring an unknown territory under changing circumstances in an optimal or at least sub-optimal fashion. The architecture can be exploited not only by DALI agents, but rather by any kind of logical agent.

Internet of things (IoT), from heart monitoring implants to home heating control systems, are becoming an integral part of our daily lives. We expect these technologies to become smarter, able to autonomously reason, act, and communicate with other entities in the environment and act to achieve shared goals. To realize the full potential of these systems, we need to understand the mechanisms that allow multiple agents to effectively operate in changing and uncertain environments. This paper presents a framework that postulates that optimal multi-agent systems achieve adaptive behaviors by minimizing the team’s free energy, where energy minimization process consists of incremental perception (inference) and control (action) phases. We discuss instantiation of this mechanism for a problem of joint distributed decision making, provide the concomitant abstractions and computational mechanisms, and present experimental evidence that energy-based agent teams significantly outperform utility-based teams. We discuss different adaptation mechanisms and scales, explain agent interdependencies produced by energy-based modeling, and look at the role of learning in the adaptation process. We hypothesize that to efficiently operate in uncertain and changing environments, IoT devices must not only maintain enough intelligence to perceive and act locally, but also possess team-level adaptation primitives. We posit that such primitives must embody energy-minimizing mechanisms but can be locally defined without the need for agents to possess global team-level objectives or constraints.

For the Internet of Everything (IoE), from an AI perspective, we discuss the meaning, value and effect that the internet of things (IoT) is expected to have on ordinary life, in industry (IIoT), on the battlefield (IoBT), in the medical field (IoMT) and with intelligent-agent feedback in the form of constructive and destructive interference (IoIT). We consider the topic open-ended but with an AI perspective that addresses how the IoE affects sensing, perception, cognition and behavior, or causal relations whether the context is clear or uncertain for mundane decisions, complex decisions on the battlefield, life and death decisions in the medical arena, or decisions affected by intelligent agents and machines. We pay attention to theoretical perspectives for how these “things” may affect individuals, teams and society; and in turn how they may affect these “things”. We are most interested in what may happen when these “things” begin to think. Our ultimate goal is to use AI to advance autonomy and autonomous characteristics to improve the performance of individual agents and hybrid teams of humans, machines, and robots for the betterment of society.

We examine moral decision making in autonomous systems as inspired by a central question posed by Rossi with respect to moral preferences: can AI systems based on statistical machine learning (which do not provide a natural way to explain or justify their decisions) be used for embedding morality into a machine in a way that allows us to prove that nothing morally wrong will happen? We argue for an evaluation which is held to the same standards as a human agent, removing the demand that ethical behavior is always achieved. We introduce four key meta-qualities desired for our moral standards, and then proceed to clarify how we can prove that an agent will correctly learn to perform moral actions given a set of samples within certain error bounds. Our group-dynamic approach enables us to demonstrate that the learned models converge to a common function to achieve stability. We further explain a valuable intrinsic consistency check made possible through the derivation of logical statements from the machine learning model. In all, this work proposes an approach for building ethical AI systems, coming from the perspective of artificial intelligence research, and sheds important light on understanding how much learning is required in order for an intelligent agent to behave morally with negligible error.

Implementing Machine Ethics in Intelligent Agents involves trustworthiness and safety, meaning that agents should do what is expected they should do (at least, even in case of malfunctioning of any kind, concerning high-priority goals) and should not behave in unexpected potentially harmful ways. This topics are strongly related with "assurance", i.e., to ensuring that system users can rely upon the system. This paper deals with assurance of logical agent systems via temporal-logic-based runtime self-monitoring and checking.

We propose and analyze a new stochastic gradient method, which we call Stochastic Unbiased Curvature-aided Gra- dient (SUCAG), for finite sum optimization problems. SUCAG constitutes an unbiased total gradient tracking technique that uses Hessian information to accelerate convergence. We an- alyze our method under the general asynchronous model of computation, in which functions are selected infinitely often, but with delays that can grow sublinearly. For strongly convex problems, we establish linear convergence for the SUCAG method. When the initialization point is sufficiently close to the optimal solution, the established convergence rate is only dependent on the condition number of the problem, making it strictly faster than the known rate for the SAGA method. Furthermore, we describe a Markov-driven approach of implementing the SUCAG method in a distributed asynchronous multi-agent setting, via gossiping along a random walk on the communication graph. We show that our analysis applies as long as the undirected graph is connected and, notably, establishes an asymptotic linear convergence rate that is robust to the graph topology. Numerical results demonstrate the merit of our algorithm over existing methods.

These combinations of real-time biological systems can blend knowledge, exploration, and exploitation to unify intelligence and solve problems more efficiently. These simple agents interact locally, within their environment, and new behaviors emerge from the group as a whole. In the world of evolutionary alogirthms one such inspired method is particle swarm optimization (PSO). It is a swarm intelligence based computational technique that can be used to find an approximate solution to a problem by iteratively trying to search candidate solutions (called particles) with regard to a given measure of quality around a global optimum. The movements of the particles are guided by their own best known position in the search-space as well as the entire swarm's best known position.

The Arcade Learning Environment (ALE) is an evaluation platform that poses the challenge of building AI agents with general competency across dozens of Atari 2600 games. It supports a variety of different problem settings and it has been receiving increasing attention from the scientific community, leading to some high-profile success stories such as the much publicized Deep Q-Networks (DQN). In this article we take a big picture look at how the ALE is being used by the research community. We show how diverse the evaluation methodologies in the ALE have become with time, and highlight some key concerns when evaluating agents in the ALE. We use this discussion to present some methodological best practices and provide new benchmark results using these best practices. To further the progress in the field, we introduce a new version of the ALE that supports multiple game modes and provides a form of stochasticity we call sticky actions. We conclude this big picture look by revisiting challenges posed when the ALE was introduced, summarizing the state-of-the-art in various problems and highlighting problems that remain open.