China is planning to upgrade its naval power with unmanned AI submarines that aim to provide an edge over the fleets of their global counterparts. A report by the South China Post on Sunday revealed Beijing's plans to build the automated subs by the early 2020s in response to unmanned weapons being developed in the US. The subs will be able to patrol areas in the South China Sea and Pacific Ocean that are home to disputed military bases. While the expected cost of the submarines has not been disclosed, they're likely to be cheaper than conventional submarines as they do not require life-supporting apparatus for humans. However, without a human crew, they'll also need to be resilient enough to be at sea without onboard repairs possible. The XLUUVs (Extra-Large Unmanned Underwater Vehicles) are much bigger than current underwater vehicles, will be able to dock as any other conventional submarine, and will carry a large amount of weaponry and equipment.

Wired editor Maria Streshinsky speaks to computer and data science experts Kai-Fu Lee and Fei-Fei Li. We're told constantly that artificial Intelligence (AI) is ever-rising in its ubiquity, seeping into every industry, finding its place in all aspects of the business-- enabling us to work in different ways; in some cases, threatening to take over our roles entirely. Stats such as recruitment firm Robert Walters', which predicts AI will give rise to 133 million new jobs across the globe in the future, can sound vague and far off in a distant future, while things probably haven't seemed to have changed much at our desks. But rest assured, hype aside, the'age of AI' is drawing closer, and the evidence lies in businesses' eagerness to invest in the talent to make it happen. The AI specialist now represents the fastest-growing role in the United States over the last four years.

That role, A.I. specialist, is the fastest growing U.S. job in terms of number of hires, at least according to LinkedIn, which published its annual emerging jobs report on Tuesday. Hirings for A.I. specialists on the career networking service have grown 74% annually over the past four years, LinkedIn said. But it didn't reveal how many jobs that represents, only that demand for that job role is growing faster than other emerging jobs. What's noteworthy about this year's survey is that last year's top job role, blockchain developer, is absent from the latest list. It highlights how the recent craze over cryptocurrencies and blockchain created a brief demand for blockchain-related jobs, but as the hype died down, so too did demand for people with blockchain skills.

It's never a bad time to be an engineer--or to have people skills. LinkedIn's third annual US emerging jobs report has identified the 15 fastest-growing jobs, as well as the skills and cities most associated with them. This year the company found that the number of artificial intelligence and data science roles continue to expand across nearly every industry. For the first time, robotics has made an appearance on the list, and at least five roles in the ranking include the word "engineer" in the title. But it's not just high-tech roles that have seen a lot more hiring action in the past five years, which is how far back LinkedIn looks to measure the emergence of roles based on user profile data and hiring growth trends.

Multivariate time series forecasting is an important yet challenging problem in machine learning. Most existing approaches only forecast the series value of one future moment, ignoring the interactions between predictions of future moments with different temporal distance. Such a deficiency probably prevents the model from getting enough information about the future, thus limiting the forecasting accuracy. To address this problem, we propose Multi-Level Construal Neural Network (MLCNN), a novel multi-task deep learning framework. Inspired by the Construal Level Theory of psychology, this model aims to improve the predictive performance by fusing forecasting information (i.e., future visions) of different future time. We first use the Convolution Neural Network to extract multi-level abstract representations of the raw data for near and distant future predictions. We then model the interplay between multiple predictive tasks and fuse their future visions through a modified Encoder-Decoder architecture. Finally, we combine traditional Autoregression model with the neural network to solve the scale insensitive problem. Experiments on three real-world datasets show that our method achieves statistically significant improvements compared to the most state-of-the-art baseline methods, with average 4.59% reduction on RMSE metric and average 6.87% reduction on MAE metric.

SpaceX launched a 3-ton cargo payload to the International Space Station (ISS) on Thursday, which included barley grains for a beer experiment, mice for muscle-building research and a robot designed to show empathy. The Falcon 9 rocket carrying the recycled Dragon capsule filled with the goodies lifted off from Cape Canaveral, Fla., around 12:30 p.m. The capsule is expected to arrive at the station housing six astronauts -- three Americans, two Russians and one Italian -- on Sunday. SpaceX recovered the new booster on a barge just off the coast in the Atlantic several minutes following liftoff so that it could be reused. SpaceX employees in Southern California cheered when the booster landed, and again a few minutes later when the capsule reached orbit.

Prototype of the Force 12 Xplorer being tested near Victoria, British Columbia. It uses a rigid ... [ ] wingsail for propulsion. It's been a great year for Open Ocean Robotics, a British Columbia-based startup that makes solar-powered drones that can gather environmental data in real time and help address a multitude of issues. During 2019, Open Ocean Robotics won a most-promising startup award from the National Community for Angels, Incubators, and Accelerators; $100,000 in a Spring Impact Investor Challenge; and was a finalist in a New Ventures BC Competition, to name a few. So how do you follow that up for 2020?

If you hadn't heard, Elon Musk is worried about the machines. Though that may seem a quixotic stance for the head of multiple tech companies to take, it seems that his proximity to the bleeding edge of technological development has given him the heebie-jeebies when it comes to artificial intelligence. He's shared his fears of AI running amok before, likening it to "summoning the demon," and Musk doubled down on his stance at a meeting of the National Governors Association this weekend, telling state leaders that AI poses an existential threat to humanity. "Until people see robots going down the street killing people, they don't know how to react because it seems so ethereal. AI is a rare case where I think we need to be proactive in regulation instead of reactive. Because I think by the time we are reactive in AI regulation, it's too late," according to the MIT Tech Review.

Neural networks have become increasingly prevalent within the geosciences for applications ranging from numerical model parameterizations to the prediction of extreme weather. A common limitation of neural networks has been the lack of methods to interpret what the networks learn and how they make decisions. As such, neural networks have typically been used within the geosciences to accurately identify a desired output given a set of inputs, with the interpretation of what the network learns being used - if used at all - as a secondary metric to ensure the network is making the right decision for the right reason. Network interpretation techniques have become more advanced in recent years, however, and we therefore propose that the ultimate objective of using a neural network can also be the interpretation of what the network has learned rather than the output itself. We show that the interpretation of a neural network can enable the discovery of scientifically meaningful connections within geoscientific data. By training neural networks to use one or more components of the earth system to identify another, interpretation methods can be used to gain scientific insights into how and why the two components are related. In particular, we use two methods for neural network interpretation. These methods project the decision pathways of a network back onto the original input dimensions, and are called "optimal input" and layerwise relevance propagation (LRP). We then show how these interpretation techniques can be used to reliably infer scientifically meaningful information from neural networks by applying them to common climate patterns. These results suggest that combining interpretable neural networks with novel scientific hypotheses will open the door to many new avenues in neural network-related geoscience research.

NEW DELHI – Japan and India agreed Saturday to conduct their first joint fighter aircraft exercise in Japan as part of efforts to promote bilateral security cooperation in the face of China's military buildup and regional assertiveness. In inaugural "two-plus-two" security talks, the nations' foreign and defense ministers also welcomed the significant progress in negotiations for a pact that would allow the sharing of defense capabilities and supplies including fuel and ammunition. They called for a speedy conclusion to the acquisition and cross-servicing agreement (ACSA), according to a joint statement issued after the talks in New Delhi. The two governments are planning to sign the deal when Prime Minister Shinzo Abe visits India for talks with Prime Minister Narendra Modi in mid-December, according to Japanese officials. Tokyo and New Delhi aim to have a joint exercise involving fighter jets from the Air Self-Defense Force and the Indian Air Force next year, the officials said.