Want to know why mail drones aren't ready for prime time? The Siberian town of Ulan-Ude was expecting to beam with pride as organizer Rudron/Expeditor 3M tested a postal drone in the area for the first time, but they left red-faced after the inaugural flight went spectacularly wrong. The hexacopter courier went haywire moments after takeoff, smacking into the side of a building at high speed -- as you can see in the video below, it went from technological triumph to an embarrassing pile of scrap metal in a matter of seconds. It was a costly crash, too, as the drone reportedly cost about $20,000. It's not certain what went wrong, although regional leader Alexei Tsydenov speculated to Reuters that the 100-plus WiFi signals in the area might have played a part.

Recently, news broke that SF Holding Co. received a government license to operate drones in China. This would open up the possibility for package delivery via drone in the second largest economy in the world. "More than 40 billion packages were shipped around China in 2017 and the number is expected to rise to 49 billion in 2018," according to the State Post Bureau, China's delivery-industry watchdog. The studies also stated that the drones equipped with battery with an average 2.5mile action radius consumed less energy per package, than traditional trucks. Moreover, package delivery via drone could also help reduce air pollution.

Malcolm Gladwell, among accomplishments including being appointed to the Order of Canada, is the author of numerous New York Times best sellers such as Blink, Outliers and The Tipping Point. Gladwell is not only a journalist, author, public figure and speaker, he is also a staple in influencing and observing social sciences while leading the way for thousands of entrepreneurs to think critically regarding their business endeavors. I was invited to speak this year at the summit in Dubai and had an opportunity to talk with Gladwell about artificial intelligence, the future of technology and, of course, whether or not he was a dork in high school. As an entrepreneur, we often have a feeling of being different or separate from others and their way of thinking. So, of course, I wanted to hear that Gladwell felt that same way.

Microsoft is betting big on'democratising' artificial intelligence (AI) and plans to make it'available to all' to improve lives and transform organizations. Reducing the risk of accidents, predicting the behavior of users, detecting the disease and helping the farms to increase the crop fields are some of the advantages of AI-based innovations. Is it something that would be harnessed by some of the biggest companies in the world? Who is AI for?" said Anant Maheshwari, president, Microsoft India, in his keynote address at the company's'AI for All' conference here. "And one of the core perspectives we have had as a company is to democratize AI and bring it for everyone," he said. Microsoft is helping more than 650 Indian-based partners to build innovative solutions with the help of its cognitive services, AI, machine learning and IoT solutions. The company and its partner firms have deployed AI solutions in different fields such as agriculture, retail, e-commerce, manufacturing, healthcare, education, and BFSI. For instance, Flipkart is using AI and analytical capabilities in Microsoft cloud computing Azure to deliver a customized experience to its users. Ola uses Microsoft's IoT and AI solutions to improve the experience of its passengers and drivers. The telematics platform transforms the car into a high-performance intelligent vehicle, improving engine and driver performance. It also enables to predict breakdowns, make smart navigation and enhance security and safety. Microsoft has signed an agreement with a Bengaluru based healthcare startup Forus, on retinal imaging devices. It would use AI and IoT for early detections of eye diseases such as glaucoma, macular degeneration, diabetic retinopathy and help in reducing blindness. "AI is going to be very important because when you can give the response to somebody [patient] immediately, the chances of he taking the whole thing more seriously is very high," said K. Chandrasekhar, co-founder, and CEO of Forus. He said that India had only 20,000 ophthalmologists for the growing population of 1.3 billion. Microsoft also said that as we are entering into the world of AI with all the ingredients like big data, computing power, and cloud technologies together, the company is taking the consequences of AI very seriously. "We don't want to unleash it, we want to be very mindful of it," said Peggy Johnson, executive vice president, business development at Microsoft. "As AI is introduced in our daily lives, we want people to know what is going on with their data," said Ms. Johnson. "If anything should go wrong, you can't say'oh the machine did it' that just is not an acceptable answer.

Samsung Electronics has announced a new laptop called Notebook Odyssey Z. The device features a powerful processor that's designed for an unparalleled gaming experience. Samsung launched the new portable computer at Intel's Worldwide Launch Event for new processors in Beijing, China on Tuesday and introduced it to the rest of the world via its online newsroom on the same day. In its announcement, the South Korean tech giant proudly stated that the Notebook Odyssey Z is a "strikingly thin and light laptop" made with state-of-the-art engineering as well as advanced thermal management to ensure that it provides a gaming experience like no other. According to Samsung, the Notebook Odyssey Z has an entirely new approach to gaming that takes into consideration the needs of consumers who love to play video games for long periods.

Russia's test launch of mail delivery service by drone ends in spectacular failure in the Siberian city Ulan-Ude.

The US Trade Representative has published the list of Chinese products that would be subject to its proposed tech tariffs, and there are a few clear themes. The move would hike the costs of about 1,300 products, including industrial robots, communication satellites, spacecraft and a slew of semiconductors.The aim, as before, is to punish China for allegedly goading American companies into transferring their patents and technology to Chinese firms for the sake of claiming economic superiority. The USTR claimed the proposed tariffs would stymie Chinese plans will "minimizing the impact" on the American economy. The tariffs are still subject to a 60-day notice process that would include public comments until May 11th and a public hearing on May 15th. Whether or not it plays out as claimed is another matter.

WASHINGTON – The Trump administration on Tuesday escalated its aggressive actions on trade by proposing 25 percent tariffs on $50 billion in Chinese imports to protest Beijing's alleged theft of American technology. The Office of the U.S. Trade Representative issued a list targeting 1,300 Chinese products, including industrial robots and telecommunications equipment. The suggested tariffs wouldn't take effect right away: A public comment period will last until May 11, and a hearing on the tariffs is set for May 15. Companies and consumers will have the opportunity to lobby to have some products taken off the list or have others added. The latest U.S. move risks heightening trade tensions with China, which on Monday had slapped taxes on $3 billion in U.S. products in response to earlier U.S. tariffs on steel and aluminum imports.

With deep convolutional neural networks (CNN) introduced by [1] and [2], computer vision tasks and more specifically image classification have made huge improvements in the years following [3]. CNN performances benefit a lot from big collections of annotated images like [4] or [5]. They are trained by optimizing a loss function with gradient descents computed on random mini-batches according to [6]. The method called stochastic gradient descent (SGD) has proved to be very efficient to train neural networks in general. However current CNN structures are extremely deep like the 100 layers ResNet of [7] and contains a lot of parameters (around 60M for Alexnet [3] and 130M for vgg [8]). Those structures involve heavy gradient computation times making the training on big data-sets very slow. Computation on GPU accelerates the training but requires huge local memory caches. Nevertheless the mini-batch optimization seems suitable for distributing the training over several threads. Many methods have been proposed like 1 [9, 10], which propose to distribute the batches over different threads called workers that periodically exchange information via a central thread to synchronize their models.

In recent years, China, the United States and other countries, Google and other high-tech companies have increased investment in artificial intelligence. Deep learning is one of the current artificial intelligence research's key areas. This paper analyzes and summarizes the latest progress and future research directions of deep learning. Firstly, three basic models of deep learning are outlined, including multilayer perceptrons, convolutional neural networks, and recurrent neural networks. On this basis, we further analyze the emerging new models of convolution neural networks and recurrent neural networks. This paper then summarizes deep learning's applications in many areas of artificial intelligence, including voice, computer vision, natural language processing and so on. Finally, this paper discusses the existing problems of deep learning and gives the corresponding possible solutions.