In a TEDx talk entitled "How Computers are Learning to Be Creative", Blaise Agüera y Arcas, Google principal scientist, demonstrated how neural networks recognizing images can run them in reverse-- thus generating them. Of this, he noted that perception and creativity are highly linked together. With Google's neural network models, machine perception and machine creativity are no longer that far-fetched. Arcas identified perception as the process by which simple objects are transformed by the mind into overwhelmingly different concepts. With today's technology, even computers are capable of perception. Creativity, on the other hand, is actually-- as far as Arcas is concerned in his speech-- the "flip side" of the former.

Pathologists have been largely diagnosing disease the same way for the past 100 years, by manually reviewing images under a microscope. But new work suggests that computers can help doctors improve accuracy and significantly change the way cancer and other diseases are diagnosed. A research team from Harvard Medical School and Beth Israel Deaconess Medical Center and recently developed artificial intelligence (AI) methods aimed at training computers to interpret pathology images, with the long-term goal of building AI-powered systems to make pathologic diagnoses more accurate. "Our AI method is based on deep learning, a machine-learning algorithm used for a range of applications including speech recognition and image recognition," explained pathologist Andrew Beck, HMS associate professor of pathology and director of bioinformatics at the Cancer Research Institute at Beth Israel Deaconess. "This approach teaches machines to interpret the complex patterns and structure observed in real-life data by building multi-layer artificial neural networks, in a process which is thought to show similarities with the learning process that occurs in layers of neurons in the brain's neocortex, the region where thinking occurs."

In the construction industry, many projects involve remodeling or refurbishing existing buildings, and such jobs often face delays or cost overruns when hidden problems emerge. "Renovation projects live and die by the quality of information," according to Martin Fischer, a Stanford professor of civil and environmental engineering. New software can analyze this point cloud to automatically extract details to plan a remodel or other purposes. Newer buildings often have computerized blueprints and records, including details such as the number of rooms, doors and windows, and the square footage of floors, ceilings and walls. But such information may not exist for older buildings, necessitating the time-consuming and difficult task of collecting these details manually.

Since taking office, President Obama has sent U.S. troops into action on land or in the skies of seven countries on two continents. Obama's administration has authorized Navy SEALs to kill Osama bin Laden in Pakistan and approved the fatal drone strike on an American cleric in Yemen. Here is a look at targeted killings under the Obama administration. Mansour was killed when a drone strike hit his vehicle as he traveled in Baluchistan, Pakistan. Mansour, known for his mercurial leadership, had been in the U.S. military's crosshairs for years.

The company also believes that autonomous ships will be appearing in service within just 10 years. Speaking about these innovations Oskar Levander, Rolls-Royce, Vice President of Innovation – Marine, said: "This is happening. The technologies needed to make remote and autonomous ships a reality exist." "The AAWA project is testing sensor arrays in a range of operating and climatic conditions in Finland and has created a simulated autonomous ship control system which allows the behaviour of the complete communication system to be explored. We will see a remote controlled ship in commercial use by the end of the decade."

A.I.: Artificial Intelligence--which was released 15 years ago today--was an unprecedented collaboration between two titans of cinema: Stanley Kubrick and Steven Spielberg. The film, based on the 1969 short story Supertoys Last All Summer Long, is set in the late 21st century and tells the story of a robot named David (Haley Joel Osment) who is programmed to feel human love for his parents, Henry and Monica. After Henry and Monica's human son Martin is brought back to life from suspended animation, his jealousy leads him to get David cast off into the wilderness with Teddy, his robotic teddy bear friend. David and Teddy soon befriend a robotic prostitute named Gigolo Joe (Jude Law), and David's quest to become a "real boy" begins in earnest. In 1983, 18 years before A.I. made it into theaters, Stanley Kubrick bought the movie rights to author Brian Aldiss' short story, Supertoys Last All Summer Long.

H2O: A Platform for Big Math From just your laptop to 100's of nodes, H2O gives you a Single System Image - easy aggregation of all the memory and all the cores, and a simple coding style that scales wide at in-memory speeds. H2O is easily 1000x faster than disk based clustering solutions, and often 10x faster than best-of-breed alternative in-memory solutions - and will work directly on your existing Hadoop cluster. H2O ingests a wide variety of formats, parallel and distributed across the cluster, and stores the data highly compressed and then lets you do scale-out math at memory-bandwidth speeds (on compressed data!), making terabyte-scale munging an interactive experience. This is a technical talk on the insides of H2O, specifically focusing on the Single-System-Image aspect: how we write single-threaded code, and have H2O auto-parallelize and auto-scale-out to 100's of nodes and 1000's of cores. Arno is the Chief Architect of H2O, a distributed and scalable open-source machine learning platform.

On the app's home page (above), Google has replaced the single activity dial graph with individual cards and dials for each goal, showing what you've done and how days you have left to meet it. Scrolling down will reveal a chart of your weight, recent workouts (complete with a map), and more (below). Hitting the floating action " " button lets you set new goals, log your weight, add an activity and more. As for the new goals, you can now get a lot more specific than before. It has more activities and more specific metrics, like steps, duration or times per week, day or month.

The simulator is an R package that streamlines the process of performing simulations by creating a common infrastructure that can be easily used and reused across projects. Methodological statisticians routinely write simulations to compare their methods to preexisting ones. While developing ideas, there is a temptation to write "quick and dirty" simulations to try out ideas. This approach of rapid prototyping is useful but can sometimes backfire if bugs are introduced. Using the simulator allows one to remove the "dirty" without sacrificing the "quick." Coding is quick because the statistician focuses exclusively on those aspects of the simulation that are specific to the particular paper being written. Code written with the simulator is succinct, highly readable, and easily shared with others. The modular nature of simulations written with the simulator promotes code reusability, which saves time and facilitates reproducibility. The syntax of the simulator leads to simulation code that is easily human-readable. Other benefits of using the simulator include the ability to "step in" to a simulation and change one aspect without having to rerun the entire simulation from scratch, the straightforward integration of parallel computing into simulations, and the ability to rapidly generate plots, tables, and reports with minimal effort.

Neural Machine Translation (MT) has reached state-of-the-art results. However, one of the main challenges that neural MT still faces is dealing with very large vocabularies and morphologically rich languages. In this paper, we propose a neural MT system using character-based embeddings in combination with convolutional and highway layers to replace the standard lookup-based word representations. The resulting unlimited-vocabulary and affix-aware source word embeddings are tested in a state-of-the-art neural MT based on an attention-based bidirectional recurrent neural network. The proposed MT scheme provides improved results even when the source language is not morphologically rich. Improvements up to 3 BLEU points are obtained in the German-English WMT task.