Cast your mind back to April, and the image of a United passenger being ejected from an overbooked plane, wrestled to the ground by security guards and then forcibly dragged, kicking and screaming, from the flight he was about to take home. Such an unsavoury incident was deeply damaging for the United brand, but it also cast a shadow over the customer service profession as a whole. How could an established business like United misread a situation so badly? Had the humanity of customer service become so far-removed that rigid procedure replaced common sense? In many ways, 2017 has been a year defined by the customer service profession questioning its own humanity. The proliferation of chatbots, artificial intelligence and voice assistants has increasingly raised the question of where humans fit into the service function in their business.

We always keep hearing that robots or machines will replace human, and our workplaces will change dramatically. The fundamental truth is that it will, and we have an opportunity to start planning for that, but, like anything else, it's unclear exactly when the tipping point will be. Artificial Intelligence is one of the most trending topics in today's date. It is the way of enabling a computer software to think intelligently in a manner similar to that of humans. In technical terms, it is an integrated solution of Machine Learning, Data Science, Data Mining, Predictive Analytics, Multi-agent Systems, and fast & reliable computation.

Before an agent takes an action, it observes the communications from other agents from the previous time step as well as the locations of all entities and objects in the world. It stores that communication in a private recurrent neural network, giving it a memory for the words it hears. We use discrete communication actions (messages formed of separate, word-like symbols) sent over a differentiable communication channel. A communication channel is differentiable if it allows agents to directly inform each other about what message they should have sent at each time step, by slightly altering their messages to make a positive change in the reward both agents expect to receive. Agents accomplish this by calculating the gradient of future reward with respect to changes in the sent messages (i.e.

When Gilbert Serrano opened the door of his potential dream house, a modern, two-bedroom rental, he was surprised to be greeted not by a real estate agent, but by a robot. An iPad mounted on the machine displayed an agent's smiling face. These robots, rolled out last summer in the Bay Area by high-tech property management startup Zenplace, are intended to take the hassle out of coordinating showing times between agents and prospective renters. They're just one piece of the new wave of technology that's changing the way houses are bought, sold and rented, as platforms such as Zillow, Redfin and a host of smaller startups have eroded the real estate agent's importance. These days, clients can use artificial intelligence to comb property data without a human real estate agent, take virtual tours of promising houses from their couches, and even apply for their favorite apartments online.

We study the cost sharing problem for cooperative games in situations where the cost function C is not available via oracle queries, but must instead be learned from samples drawn from a distribution, represented as tuples (S, C(S)), for different subsets S of players. We formalize this approach, which we call statistical cost sharing, and consider the computation of the core and the Shapley value. Expanding on the work by Balcan et al, we give precise sample complexity bounds for computing cost shares that satisfy the core property with high probability for any function with a non-empty core. For the Shapley value, which has never been studied in this setting, we show that for submodular cost functions with curvature bounded curvature kappa it can be approximated from samples from the uniform distribution to a sqrt{1 - kappa} factor, and that the bound is tight. We then define statistical analogues of the Shapley axioms, and derive a notion of statistical Shapley value and that these can be approximated arbitrarily well from samples from any distribution and for any function.

We consider a two-player sequential game in which agents have the same reward function but may disagree on the transition probabilities of an underlying Markovian model of the world. By committing to play a specific policy, the agent with the correct model can steer the behavior of the other agent, and seek to improve utility. We model this setting as a multi-view decision process, which we use to formally analyze the positive effect of steering policies. Furthermore, we develop an algorithm for computing the agents' achievable joint policy, and we experimentally show that it can lead to a large utility increase when the agents' models diverge.

We study online reinforcement learning in average-reward stochastic games (SGs). An SG models a two-player zero-sum game in a Markov environment, where state transitions and one-step payoffs are determined simultaneously by a learner and an adversary. We propose the \textsc{UCSG} algorithm that achieves a sublinear regret compared to the game value when competing with an arbitrary opponent. This result improves previous ones under the same setting. The regret bound has a dependency on the \textit{diameter}, which is an intrinsic value related to the mixing property of SGs. Slightly extended, \textsc{UCSG} finds an $\varepsilon$-maximin stationary policy with a sample complexity of $\tilde{\mathcal{O}}\left(\text{poly}(1/\varepsilon)\right)$, where $\varepsilon$ is the error parameter. To the best of our knowledge, this extended result is the first in the average-reward setting. In the analysis, we develop Markov chain's perturbation bounds for mean first passage times and techniques to deal with non-stationary opponents, which may be of interest in their own right.

Autonomous agents optimize the reward function we give them. What they don't know is how hard it is for us to design a reward function that actually captures what we want. When designing the reward, we might think of some specific training scenarios, and make sure that the reward will lead to the right behavior in those scenarios. Inevitably, agents encounter new scenarios (e.g., new types of terrain) where optimizing that same reward may lead to undesired behavior. Our insight is that reward functions are merely observations about what the designer actually wants, and that they should be interpreted in the context in which they were designed. We introduce inverse reward design (IRD) as the problem of inferring the true objective based on the designed reward and the training MDP. We introduce approximate methods for solving IRD problems, and use their solution to plan risk-averse behavior in test MDPs. Empirical results suggest that this approach can help alleviate negative side effects of misspecified reward functions and mitigate reward hacking.

We explore deep reinforcement learning methods for multi-agent domains. We begin by analyzing the difficulty of traditional algorithms in the multi-agent case: Q-learning is challenged by an inherent non-stationarity of the environment, while policy gradient suffers from a variance that increases as the number of agents grows. We then present an adaptation of actor-critic methods that considers action policies of other agents and is able to successfully learn policies that require complex multi-agent coordination. Additionally, we introduce a training regimen utilizing an ensemble of policies for each agent that leads to more robust multi-agent policies. We show the strength of our approach compared to existing methods in cooperative as well as competitive scenarios, where agent populations are able to discover various physical and informational coordination strategies.

We consider a model of game-theoretic learning based on online mirror descent (OMD) with asynchronous and delayed feedback information. Instead of focusing on specific games, we consider a broad class of continuous games defined by the general equilibrium stability notion, which we call λ-variational stability. Our first contribution is that, in this class of games, the actual sequence of play induced by OMD-based learning converges to Nash equilibria provided that the feedback delays faced by the players are synchronous and bounded. Subsequently, to tackle fully decentralized, asynchronous environments with (possibly) unbounded delays between actions and feedback, we propose a variant of OMD which we call delayed mirror descent (DMD), and which relies on the repeated leveraging of past information. With this modification, the algorithm converges to Nash equilibria with no feedback synchronicity assumptions and even when the delays grow superlinearly relative to the horizon of play.