We consider the problem of multiple agents sensing and acting in environments with the goal of maximising their shared utility. In these environments, agents must learn communication protocols in order to share information that is needed to solve the tasks. By embracing deep neural networks, we are able to demonstrate end-to-end learning of protocols in complex environments inspired by communication riddles and multi-agent computer vision problems with partial observability. We propose two approaches for learning in these domains: Reinforced Inter-Agent Learning (RIAL) and Differentiable Inter-Agent Learning (DIAL). The former uses deep Q-learning, while the latter exploits the fact that, during learning, agents can backpropagate error derivatives through (noisy) communication channels.

AI has a long way to go before it sees widespread adoption in the enterprise, but more and more companies are looking at the potential its capabilities have to offer. "It's still not too late to get started, but if you aren't taking steps now to plan and act for 2020, much less 2025, you'll very quickly find yourself fighting a 21st century war with 20th century weapons," according to the study. Recent reports have pointed to automation supported by intelligent software agents as causing a disruption in the workforce. Forrester projects that by 2021, AI and cognitive technologies will cause the elimination of 6% of U.S. jobs. But, according to Cognizant, only one-third of respondents think it is "very likely" that fewer people will be needed in the workforce. Instead, most think that people and the ingenuity they have to offer will still be required.

The literature on judgment aggregation is moving from studying impossibility results regarding aggregation rules towards studying specific judgment aggregation rules. Here we give a structured list of most rules that have been proposed and studied recently in the literature, together with various properties of such rules. We first focus on the majority-preservation property, which generalizes Condorcet-consistency, and identify which of the rules satisfy it. We study the inclusion relationships that hold between the rules. Finally, we consider two forms of unanimity, monotonicity, homogeneity, and reinforcement, and we identify which of the rules satisfy these properties.

We determine the quality of randomized social choice mechanisms in a setting in which the agents have metric preferences: every agent has a cost for each alternative, and these costs form a metric. We assume that these costs are unknown to the mechanisms (and possibly even to the agents themselves), which means we cannot simply select the optimal alternative, i.e. the alternative that minimizes the total agent cost (or median agent cost). However, we do assume that the agents know their ordinal preferences that are induced by the metric space. We examine randomized social choice functions that require only this ordinal information and select an alternative that is good in expectation with respect to the costs from the metric. To quantify how good a randomized social choice function is, we bound the distortion, which is the worst-case ratio between expected cost of the alternative selected and the cost of the optimal alternative. We provide new distortion bounds for a variety of randomized mechanisms, for both general metrics and for important special cases. Our results show a sizable improvement in distortion over deterministic mechanisms.

Business and technology management executives wondered what big data meant, when the cloud would disrupt their companies, and how to engage effectively on social channels. In 2016, Hadoop turned 10, the cloud has been around even longer, and social has become a way of business and life. As a refresh to my 2014 blog and report, here are the next 15 emerging technologies Forrester thinks you need to follow closely. We organize this year's list into three groups -- systems of engagement technologies will help you become customer-led, systems of insight technologies will help you become insights-driven, and supporting technologies will help you become fast and connected. You might have noticed a few glaring omissions.

As autonomous technology continues to evolve, the Pentagon finds itself being pulled in two directions, enticed by the capabilities that autonomous systems could provide while also insistent it always be subservient to humans, and a set of human morals and mindsets. That tension was on full display Aug. 25, when a new report from a key Pentagon advisory group called for an acceleration of autonomous systems within the US military at the same time the country's second highest ranking uniformed officer warned that there will need to be limits on how the technology is used in order to avoid the dreaded killer-robot scenario. Speaking at the Center for Strategic and International Studies, Gen. Paul Selva, vice chairman of the Joint Chiefs of Staff, laid out his concerns with the "Terminator Conundrum," the idea that a fully autonomous system could be created with the capability to make decisions about when and where to inflict violence. While noting that technologists in the Pentagon believe that capability is still a decade away, Selva noted that 15 years ago he was told a digital rendering of the world would be impossible and never happen, before dryly telling the audience" "So I guess Google Earth is an impossibility." He also threw his support behind the idea of a treaty or global convention against the creation of wholly autonomous systems that can operate without a man in the loop controlling it, saying: "I do think we need to examine the bodies of law and convention that might constrain anyone in the world from building that kind of a system.

A big challenge in environmental monitoring is the spatiotemporal variation of the phenomena to be observed. To enable persistent sensing and estimation in such a setting, it is beneficial to have a time-varying underlying environmental model. Here we present a planning and learning method that enables an autonomous marine vehicle to perform persistent ocean monitoring tasks by learning and refining an environmental model. To alleviate the computational bottleneck caused by large-scale data accumulated, we propose a framework that iterates between a planning component aimed at collecting the most information-rich data, and a sparse Gaussian Process learning component where the environmental model and hyperparameters are learned online by taking advantage of only a subset of data that provides the greatest contribution. Our simulations with ground-truth ocean data shows that the proposed method is both accurate and efficient.

We present Probabilistic Doxastic Temporal (PDT) Logic, a formalism to represent and reason about probabilistic beliefs and their temporal evolution in multi-agent systems. This formalism enables the quantification of agents beliefs through probability intervals and incorporates an explicit notion of time. We discuss how over time agents dynamically change their beliefs in facts, temporal rules, and other agents beliefs with respect to any new information they receive. We introduce an appropriate formal semantics for PDT Logic and show that it is decidable. Alternative options of specifying problems in PDT Logic are possible. For these problem specifications, we develop different satisfiability checking algorithms and provide complexity results for the respective decision problems. The use of probability intervals enables a formal representation of probabilistic knowledge without enforcing (possibly incorrect) exact probability values. By incorporating an explicit notion of time, PDT Logic provides enriched possibilities to represent and reason about temporal relations.

Analysts say a handful of technologies are poised to change our lives by 2021. While Forrester Research sees 15 emerging technologies that are important right now (see the full list here), five of them could shake things up in a big way for businesses and the public in general, according to Brian Hopkins, an enterprise architecture analyst with Forrester. "The technologies we selected will have the biggest impact on your ability to win, serve and retain customers whose expectations of service through technology are only going up," Hopkins wrote in the report. "Our list focuses on those technologies that will have the biggest business impact in the next five years." Of Forrester's larger list, which includes the likes of edge computing, security automation and real-time interaction management, the five that Hopkins pulled out to highlight have the greatest potential for disruption.

In this paper we study a model-based approach to calculating approximately optimal policies in Markovian Decision Processes. In particular, we derive novel bounds on the loss of using a policy derived from a factored linear model, a class of models which generalize numerous previous models out of those that come with strong computational guarantees. For the first time in the literature, we derive performance bounds for model-based techniques where the model inaccuracy is measured in weighted norms. Moreover, our bounds show a decreased sensitivity to the discount factor and, unlike similar bounds derived for other approaches, they are insensitive to measure mismatch. Similarly to previous works, our proofs are also based on contraction arguments, but with the main differences that we use carefully constructed norms building on Banach lattices, and the contraction property is only assumed for operators acting on "compressed" spaces, thus weakening previous assumptions, while strengthening previous results.