How can we build a learner that can capture the essence of what makes a hard problem more complex than a simple one, break the hard problem along characteristic lines into smaller problems it knows how to solve, and sequentially solve the smaller problems until the larger one is solved? To work towards this goal, we focus on learning to generalize in a particular family of problems that exhibit compositional and recursive structure: their solutions can be found by composing in sequence a set of reusable partial solutions. Our key idea is to recast the problem of generalization as a problem of learning algorithmic procedures: we can formulate a solution to this family as a sequential decision-making process over transformations between representations. Our formulation enables the learner to learn the structure and parameters of its own computation graph with sparse supervision, make analogies between problems by transforming one problem representation to another, and exploit modularity and reuse to scale to problems of varying complexity. Experiments on solving a variety of multilingual arithmetic problems demonstrate that our method discovers the hierarchical decomposition of a problem into its subproblems, generalizes out of distribution to unseen problem classes, and extrapolates to harder versions of the same problem, yielding a 10-fold reduction in sample complexity compared to a monolithic recurrent neural network.

Games generalize the optimization paradigm by introducing different objective functions for different optimizing agents, known as players. Generative Adversarial Networks (GANs) are arguably the most popular game formulation in recent machine learning literature. GANs achieve great results on generating realistic natural images, however they are known for being difficult to train. Training them involves finding a Nash equilibrium, typically performed using gradient descent on the two players' objectives. Game dynamics can induce rotations that slow down convergence to a Nash equilibrium, or prevent it altogether. We provide a theoretical analysis of the game dynamics. Our analysis, supported by experiments, shows that gradient descent with a negative momentum term can improve the convergence properties of some GANs.

We study the statistical properties of an estimator derived by applying a gradient ascent method with multiple initializations to a multi-modal likelihood function. We derive the population quantity that is the target of this estimator and study the properties of confidence intervals (CIs) constructed from asymptotic normality and the bootstrap approach. In particular, we analyze the coverage deficiency due to finite number of random initializations. We also investigate the CIs by inverting the likelihood ratio test, the score test, and the Wald test, and we show that the resulting CIs may be very different. We provide a summary of the uncertainties that we need to consider while making inference about the population. Note that we do not provide a solution to the problem of multiple local maxima; instead, our goal is to investigate the effect from local maxima on the behavior of our estimator. In addition, we analyze the performance of the EM algorithm under random initializations and derive the coverage of a CI with a finite number of initializations. Finally, we extend our analysis to a nonparametric mode hunting problem.

Imagine a state-of-the-art driverless car zipping along a road with a disabled, 90-year-old passenger. The car must make a decision: drive into the mother and child and kill them, or swerve into a wall and kill the passenger. This is a variation of the trolley problem, a thought experiment which dominates academic and popular thinking about the ethics of driverless cars. The problem is that such debates not only dismiss the complexity of the system in which driverless cars will exist, but are also moral red herrings. The real ethical issues lie in the politics and power concerns with driverless cars.

In the medium to long term, AI expertise must not reside in only a small number of countries – or solely within narrow segments of the population. Governments worldwide must invest in developing and retaining home-grown talent and expertise in AI if their countries are to be independent of the dominant AI expertise that is now typically concentrated in the US and China. And they should work to ensure that engineering talent is nurtured across a broad base in order to mitigate inherent bias issues. Corporations, foundations and governments should allocate funding to develop and deploy AI systems with humanitarian goals. The humanitarian sector could derive significant benefit from such systems, which might for example decrease response times in emergencies.

An ex-Apple engineer has been charged with stealing secret blueprints for the tech giant's automated car project before trying to flee the US for China. Xiaolang Zhang was arrested by FBI agents at San Jose airport in California on Saturday when he passed through a security checkpoint. He is accused of downloading the plan for a circuit board for the automated car just days before he quit to go to a Chinese self-driving car startup. The charge is punishable by 10 years in prison and a $250,000 fine. A criminal complaint filed on Monday said Zhang was hired by Apple in December of 2015 to develop software and hardware for the company's autonomous vehicle project, where he designed and tested circuit boards to analyze sensor data.

After graduating from IIT-Madras, the first thing that Vivek Rajkumar did was to buy four acres near Thiruvananthapuram and start farming, according to an article on India's TheHinduBusinessLine.com. He was trying to cultivate paddy and grow banana. His neighbours were small farmers and all of them consulted a self-styled local expert on all matters agriculture. That expert, says Vivek, had no clue of what he was talking about. Yields were bad and profitability was poor; it didn't make any sense at all, he adds.

Many social critics and observers have argued in recent decades that the Orwellian notion of a dystopian, "Big Brother is watching you" future was a false alarm. A New York Times report details how Chinese authorities are using cutting-edge technology to increasingly keep tabs on their citizens, in order to nab drug smugglers and murder suspects and even to publicly shame inconsiderate jay-walkers. "With millions of cameras and billions of lines of code, China is building a high-tech authoritarian future," the Times' technology correspondent Paul Mozur wrote. "Beijing is embracing technologies like facial recognition and artificial intelligence to identify and track 1.4 billion people. It wants to assemble a vast and unprecedented national surveillance system, with crucial help from its thriving technology industry."

The U.S. navy has said that from January 2016 to August 2017 there was an average of 2.5 "unsafe" or "unprofessional" interactions per month between U.S. Navy and Iranian maritime forces, including an Iranian drone flying near a U.S. Navy warplane and Iranian military boat sailing close to a US Navy vessel. Tehran has accused U.S. forces of provocation.

The Institute of Textiles and Clothing (ITC) of The Hong Kong Polytechnic University (PolyU) organised the world's first ever academic conference titled "Artificial Intelligence on Fashion and Textile Conference 2018" (AIFT). The AIFT conference is in partnership with the Vision and Beauty Team at Alibaba, an online e-commerce multinational corporation, and The Textile Institute, UK, an establishment that promotes professionalism in all areas associated with the textile industries worldwide. AIFT 2018 aims to explore the integration of the fashion and textile supply chain, and Artificial Intelligence (AI). A diverse group of researchers, engineers, and practitioners will gather to exchange their ideas on the use of Artificial Intelligence in the fashion and textile industry. Furthermore, the conference seeks to foster discussion and exploration of the most promising theories and applied intelligence topics.