After decades of extravagant promises and frustrating disappointments, artificial intelligence (AI) is finally starting to deliver real-life benefits to early-adopting companies. Retailers on the digital frontier rely on AI-powered robots to run their warehouses--and even to automatically order stock when inventory runs low. Utilities use AI to forecast electricity demand. A confluence of developments is driving this new wave of AI development. Computer power is growing, algorithms and AI models are becoming more sophisticated, and, perhaps most important of all, the world is generating once-unimaginable volumes of the fuel that powers AI--data.

London's Metropolitan Police force has said it will "carefully consider the contents" of an ethics panel's report before deciding on any continued use of Live Facial Recognition (LFR) technology in public places. The controversial technology was trialled across London for several years, and the London Policing Ethics Panel -- a group commissioned by the London Mayor's office, independent of the police -- issued its report on those trials on Wednesday. The Ethics Panel's conclusions, based on consultations with members of the public, were that most people don't have a big problem with their faces being scanned in public places and cross-checked against lists of suspected criminals -- but there are caveats. According to the report, "57% of respondents thought that in general terms, police use of LFR was acceptable. We will see, however, that the purposes for which it is used makes a significant difference to people's support for its deployment."

For most people, balloons may connote birthday parties, weddings, parades, or on a less celebratory note, meteorology. But, if one new startup has its way, sweeping surveillance may soon make that list too. World View Enterprises Inc., based in Arizona, is working to build what it's calling Stratollites -- balloon mounted-surveillance systems that the company claims can be remotely controlled and adjusted using its own proprietary technology. In an test of unprecedented length, a World View balloon safely completed a 16-day mission, navigating above states in the Western U.S. The feat, says the company, is a major mile marker in the goal of keeping the devices afloat for months at a time. Balloons could be the new method of surveillance according to one Arizona startup, World View.

My 4-year-old daughter and I were deep into a game of checkers one day about three years ago when her eye drifted to a nearby table. There, a black and white board bristled with far more interesting figures, like horses and castles. There was just one problem: I didn't know how. I dimly remembered having learned the basic moves in elementary school, but it never stuck. This fact vaguely haunted me through my life; idle chessboards in hotel lobbies or puzzles in weekend newspaper supplements teased me like reproachful riddles. And so I decided I would learn, if only so I could teach my daughter. The basic moves were easy enough to pick up--a few hours hunched over my smartphone at kids' birthday parties or waiting in line at the grocery store.

The 19th-century physicist Hermann von Helmholtz compared his progress in solving a problem to that of a mountain climber "compelled to retrace his steps because his progress stopped." A mountain climber, von Helmholtz said, "hits upon traces of a fresh path, which again leads him a little further." The physicist's introspection provokes the question: How do creative minds overcome valleys to get to the next higher peak? Because thinking minds are different from evolving organisms and self-assembling molecules, we cannot expect them to use the same means--mechanisms like genetic drift and thermal vibrations--to overcome deep valleys in the landscapes they explore. But they must have some way to achieve the same purpose.

Recently, several studies have proven the global convergence and generalization abilities of the gradient descent method for two-layer ReLU networks by making a positivity assumption of the Gram-matrix of the neural tangent kernel. However, the performance of gradient descent on classification problems has not been well studied, and further investigation of the problem structure is possible. In this work, we present a partially stronger but reasonable assumption for binary classification problems compared to the positivity assumption of the Gram-matrix, where a data distribution can be perfectly classifiable by a tangent model, and we provide a refined generalization analysis of the gradient descent method for two-layer networks with smooth activations. A remarkable point of this study is that our generalization bound has much better dependence on the network width compared to existing results. As a result, our theory significantly enlarges a class of over-parameterized networks having provable generalization ability, with respect to network width, while most studies require much higher over-parameterization.

Multi-hop Reading Comprehension (RC) requires reasoning and aggregation across several paragraphs. We propose a system for multi-hop RC that decomposes a compositional question into simpler sub-questions that can be answered by off-the-shelf single-hop RC models. Since annotations for such decomposition are expensive, we recast sub-question generation as a span prediction problem and show that our method, trained using only 400 labeled examples, generates sub-questions that are as effective as human-authored sub-questions. We also introduce a new global rescoring approach that considers each decomposition (i.e. the sub-questions and their answers) to select the best final answer, greatly improving overall performance. Our experiments on HotpotQA show that this approach achieves the state-of-the-art results, while providing explainable evidence for its decision making in the form of sub-questions.

The timeliness of detection of a sepsis event in progress is a crucial factor in the outcome for the patient. Machine learning models built from data in electronic health records can be used as an effective tool for improving this timeliness, but so far the potential for clinical implementations has been largely limited to studies in intensive care units. This study will employ a richer data set that will expand the applicability of these models beyond intensive care units. Furthermore, we will circumvent several important limitations that have been found in the literature: 1) Models are evaluated shortly before sepsis onset without considering interventions already initiated. 2) Machine learning models are built on a restricted set of clinical parameters, which are not necessarily measured in all departments. 3) Model performance is limited by current knowledge of sepsis, as feature interactions and time dependencies are hardcoded into the model. In this study, we present a model to overcome these shortcomings using a deep learning approach on a diverse multicenter data set. We used retrospective data from multiple Danish hospitals over a seven-year period. Our sepsis detection system is constructed as a combination of a convolutional neural network and a long short-term memory network. We suggest a retrospective assessment of interventions by looking at intravenous antibiotics and blood cultures preceding the prediction time. Results show performance ranging from AUROC 0.856 (3 hours before sepsis onset) to AUROC 0.756 (24 hours before sepsis onset). We present a deep learning system for early detection of sepsis that is able to learn characteristics of the key factors and interactions from the raw event sequence data itself, without relying on a labor-intensive feature extraction work.

One of the biggest challenges in robotics systems is interacting under uncertainty. Unlike robots, humans learn, adapt and perceive their body as a unity when interacting with the world. We hypothesize that the nervous system counteracts sensor and motor uncertainties by unconscious processes that robustly fuse the available information for approximating their body and the world state. Being able to unite perception and action under a common principle has been sought for decades and active inference is one of the potential unification theories. In this work, we present a humanoid robot interacting with the world by means of a human brain-like inspired perception and control algorithm based on the free-energy principle. Until now, active inference was only tested in simulated examples. Their application on a real robot shows the advantages of such an algorithm for real world applications. The humanoid robot iCub was capable of performing robust reaching behaviors with both arms and active head object tracking in the visual field, despite the visual noise, the artificially introduced noise in the joint encoders (up to 40 degrees deviation), the differences between the model and the real robot and the misdetections of the hand.

Accurate estimation of the run time of computational codes has a number of significant advantages for scientific computing. It is required information for optimal resource allocation, improving turnaround times and utilization of science gateways. Furthermore, it allows users to better plan and schedule their research, streamlining workflows and improving the overall productivity of cyberinfrastructure. Predicting run time is challenging, however. The inputs to scientific codes can be complex and high dimensional. Their relationship to the run time may be highly non-linear, and, in the most general case is completely arbitrary and thus unpredictable (i.e., simply a random mapping from inputs to run time). Most codes are not so arbitrary, however, and there has been significant prior research on predicting the run time of applications and workloads. Such predictions are generally application-specific, however. In this paper, we focus on the Gaussian computational chemistry code. We characterize a data set of runs from the SEAGrid science gateway with a number of different studies. We also explore a number of different potential regression methods and present promising future directions.