Alphabet's Google Inc unveiled its answer to Amazon's Alexa virtual assistant along with new messaging and virtual reality products at its annual I/O developer conference on Wednesday, doubling down on artificial intelligence and machine learning as the keys to its future. Google Chief Executive Sundar Pichai introduced Google Assistant, a virtual personal assistant, along with the tabletop speaker appliance Google Home. He also unveiled Allo, a new messaging service that will compete with Facebook's WhatsApp and Messenger products and feature a chatbot powered by the Google Assistant. Allo, like WhatsApp, will also have end-to-end encryption when it is rolled out this summer. Amazon's Echo, a surprise hit that has other tech giants racing to match it, uses a virtual assistant called Alexa, a cloud-based system that controls the Echo speaker and responds to voice-controlled commands by users.

Researchers from MIT have designed a new ingestible "robot" that could one day be used to patch internal wounds, deliver medicine, or remove accidentally swallowed objects from the stomach. The design consists of a specially folded sheet of dried pig intestine (usually used in sausage casing) and a tiny magnet. Folded up, this capsule can be swallowed by a patient. It then hits the stomach and unfolds in the acidic juices, where it can be guided to complete certain tasks using external magnets. The design is very much a work-in-progress, but its creators think it offers a promising model for the future.

Toyota and Volkswagen are the latest in a string of auto industry giants to partner with ride-hailing companies, as the future of "mobility services" steers toward less ownership and more self-driving cars. Toyota and Uber announced a partnership Tuesday in which Toyota, valued at 177 billion, will invest an undisclosed amount to collaborate with Uber to develop autonomous cars. Volkswagon has invested 300 million in Gett, a taxi-hailing service that operates in 60 cities in Israel, Russia, the United States, and Britain, allowing the Israeli startup to expand in Europe, The Wall Street Journal reported. Tuesday's announcements are the latest in a series of major investments by the automakers to compete with Apple (whose 500 billion value is as much as seven major automakers combined) and Google, as the everyday commute appears poised to undergo a massive transformation. The focus on autonomous cars comes as Tesla, another tech startup, is attempting revolutionize the electric car and battery technology.

A former McDonald's chief executive has warned that raising the minimum wage will spur unemployment as companies will instead employ robots that work for less. "I guarantee you if a 15 minimum wage goes across the country you're going to see a job loss like you can't believe," said Edward Rensi in an appearance on Fox Business Network Tuesday. "It's cheaper to buy a 35,000 robotic arm than it is to hire an employee who's inefficient making 15 an hour bagging French fries." The minimum wage has been a hot topic this spring, with some states and employers deciding to up their minimum wage to 15 an hour in the coming years. California will raise its minimum wage to 15 an hour by 2022.

China's labor-intensive industries have been rapidly shifting toward jobs-killing automation in recent years, with a world-leading 16 percent rise in industrial-robot sales last year alone. Now, a new official estimate has put a spotlight on some of the effects of these investments in China's electronics-manufacturing center. Taiwan's Hon Hai Precision Industry Co., more commonly known as Foxconn, has shed 60,000 workers at one factory alone, according to a recent report in the South China Morning Post. Foxconn is most famously known as Apple's largest contract manufacturer, which has faced criticism over the treatment of its workers and for a string of worker suicides. "The Foxconn factory has reduced its employee strength from 110,000 to 50,000, thanks to the introduction of robots," Xu Yulian, spokesman for Kunshan city, told the Post.

We learn the structure of a Markov Network between two groups of random variables from joint observations. Since modelling and learning the full MN structure may be hard, learning the links between two groups directly may be a preferable option. We introduce a novel concept called the \emph{partitioned ratio} whose factorization directly associates with the Markovian properties of random variables across two groups. A simple one-shot convex optimization procedure is proposed for learning the \emph{sparse} factorizations of the partitioned ratio and it is theoretically guaranteed to recover the correct inter-group structure under mild conditions. The performance of the proposed method is experimentally compared with the state of the art MN structure learning methods using ROC curves. Real applications on analyzing bipartisanship in US congress and pairwise DNA/time-series alignments are also reported.

Any-angle pathfinding is a fundamental problem in robotics and computer games. The goal is to find a shortest path between a pair of points on a grid map such that the path is not artificially constrained to the points of the grid. Prior research has focused on approximate online solutions. A number of exact methods exist but they all require super-linear space and pre-processing time. In this study, we describe Anya: a new and optimal any-angle pathfinding algorithm. Where other works find approximate any-angle paths by searching over individual points from the grid, Anya finds optimal paths by searching over sets of states represented as intervals. Each interval is identified on-the-fly. From each interval Anya selects a single representative point that it uses to compute an admissible cost estimate for the entire set. Anya always returns an optimal path if one exists. Moreover it does so without any offline pre-processing or the introduction of additional memory overheads. In a range of empirical comparisons we show that Anya is competitive with several recent (sub-optimal) online and pre-processing based techniques and is up to an order of magnitude faster than the most common benchmark algorithm, a grid-based implementation of A*.

The goal of subspace learning is to find a $k$-dimensional subspace of $\mathbb{R}^d$, such that the expected squared distance between instance vectors and the subspace is as small as possible. In this paper we study subspace learning in a partial information setting, in which the learner can only observe $r \le d$ attributes from each instance vector. We propose several efficient algorithms for this task, and analyze their sample complexity

Recently, there has been considerable progress on designing algorithms with provable guarantees -- typically using linear algebraic methods -- for parameter learning in latent variable models. But designing provable algorithms for inference has proven to be more challenging. Here we take a first step towards provable inference in topic models. We leverage a property of topic models that enables us to construct simple linear estimators for the unknown topic proportions that have small variance, and consequently can work with short documents. Our estimators also correspond to finding an estimate around which the posterior is well-concentrated. We show lower bounds that for shorter documents it can be information theoretically impossible to find the hidden topics. Finally, we give empirical results that demonstrate that our algorithm works on realistic topic models. It yields good solutions on synthetic data and runs in time comparable to a {\em single} iteration of Gibbs sampling.

We introduce a new representation and feature extraction method for biological sequences. Named bio-vectors (BioVec) to refer to biological sequences in general with protein-vectors (ProtVec) for proteins (amino-acid sequences) and gene-vectors (GeneVec) for gene sequences, this representation can be widely used in applications of deep learning in proteomics and genomics. In the present paper, we focus on protein-vectors that can be utilized in a wide array of bioinformatics investigations such as family classification, protein visualization, structure prediction, disordered protein identification, and protein-protein interaction prediction. In this method, we adopt artificial neural network approaches and represent a protein sequence with a single dense n-dimensional vector. To evaluate this method, we apply it in classification of 324,018 protein sequences obtained from Swiss-Prot belonging to 7,027 protein families, where an average family classification accuracy of 93%+-0.06% is obtained, outperforming existing family classification methods. In addition, we use ProtVec representation to predict disordered proteins from structured proteins. Two databases of disordered sequences are used: the DisProt database as well as a database featuring the disordered regions of nucleoporins rich with phenylalanine-glycine repeats (FG-Nups). Using support vector machine classifiers, FG-Nup sequences are distinguished from structured protein sequences found in Protein Data Bank (PDB) with a 99.8% accuracy, and unstructured DisProt sequences are differentiated from structured DisProt sequences with 100.0% accuracy. These results indicate that by only providing sequence data for various proteins into this model, accurate information about protein structure can be determined.