Oxford philosopher and author Nick Bostrom (left) and DeepMind CEO Demis Hassabis (right). Machines will quickly become significantly smarter than humans when they achieve human level intelligence, according to a high-profile panel of artificial intelligence (AI) leaders. A YouTube video released by the Future of Humanity Institute this week shows Elon Musk, the billionaire cofounder of Tesla, SpaceX and PayPal, talking on a panel earlier this month alongside the likes of DeepMind CEO Demis Hassabis, who sold his company to Google for £400 million in 2014, and Oxford philosopher Nick Bostrom. "Once we get to human level-AI, how long before we get to where things start taking off?" asked MIT professor and panel moderator Max Tegmark, citing an "intelligence explosion." Tegmark added: "Some people say days or hours. Others envision it will happen but it might take thousands of years or decades."

Neural networks are all the rage in Silicon Valley, infusing so many internet services with so many forms of artificial intelligence. But as good as they may be at recognizing cats in your online photos, AI researchers know that neural networks are still quite flawed, so much so that some wonder whether these pattern recognition systems are a viable path to more advanced--and more reliable--forms of AI. Able to learn tasks by analyzing vast amounts of data, neural networks power everything from face recognition at Facebook to translation at Microsoft to internet search at Google. They're beginning to help chatbots learn the art of conversation. But because they can't make sense of the world without help from such large amounts of carefully labelled data, they aren't suited to everything.

The Deep Learning market is growing at an annual rate of 65 percent annually and is expected to reach $1.8 billion by 2022 according to a report by Research and Markets. Deep Learning is a kind of machine learning, but it is more super charged than just machine learning. Deep learning software algorithms derive insight by processing massive amounts of data, like images, video, audio or text. With standard machine learning, an analyst or programmer would first create a model that explicitly tells the computer what to look for in the data, but with deep learning, instead the software sifts through the data and tries to identify patterns on its own. Businesses are beginning to apply deep learning techniques to areas like customer churn prediction, financial fraud detection and product recommendation.

Carnegie Mellon's No-Limit Texas Hold'em software made short work of four of the world's best professional poker players in Pittsburgh at the grueling "Brains vs. Artificial Intelligence" poker tournament. Poker now joins chess, Jeopardy, go, and many other games at which programs outplay people. But poker is different from all the others in one big way: players have to guess based on partial, or "imperfect" information. One of the most annoying things about Android has long been the custom skins that manufacturers would slap on top of the operating system. Things have gotten better in recent years, but plenty of users would be happier using Android as Google intended.

Stanford researchers are leveraging GPU-based machines in the Amazon EC2 cloud to run deep learning workloads with the goal of improving diagnostics for a chronic eye disease, called diabetic retinopathy. The disease is a complication of diabetes that can lead to blindness if blood sugar is poorly controlled. It affects about 45 percent of diabetics and 100 million people worldwide, many in developing nations. Final-year Stanford PhD students Apaar Sadhwani and Jason Su got involved in developing the diagnostic solution as part of a class project and corresponding Kaggle competition that was held last year. Sponsor Amazon provided AWS cloud credits in support of the research.

In this work, we are interested in generalizing convolutional neural networks (CNNs) from low-dimensional regular grids, where image, video and speech are represented, to high-dimensional irregular domains, such as social networks, brain connectomes or words' embedding, represented by graphs. We present a formulation of CNNs in the context of spectral graph theory, which provides the necessary mathematical background and efficient numerical schemes to design fast localized convolutional filters on graphs. Importantly, the proposed technique offers the same linear computational complexity and constant learning complexity as classical CNNs, while being universal to any graph structure. Experiments on MNIST and 20NEWS demonstrate the ability of this novel deep learning system to learn local, stationary, and compositional features on graphs.

Could AI-powered bots fight hate speech by flooding the internet with love? As social media platforms have become ever more intrinsic to how we live our lives and begun to evolve into the primary medium through which we communicate and listen to the rest of the world, their rise has handed a megaphone to the world's hate and vitriol. In fact, it was Twitter who initially stepped forward to staunchly defend the rights of terrorists and their sympathizers to communicate via its platform before abruptly reversing itself in the face of fierce public criticism. Yet, despite myriad programs and policies designed on paper to fight abuse, in reality the platforms have done very little to curb the spread of hate speech, harassment and violent threats. This raises the question of whether the rise of deep learning-powered "bots" could offer a powerful solution to online hate speech, by deploying them en masse to report, counter and overwhelm hateful posts in realtime. Over the last few years deep learning algorithms have made enormous advances in their ability to process human text and imagery at levels of sophistication and accuracy that approach human levels at times, while even simple ELIZA bots have managed to carry on fairly convincing chats for more than half a century.

Many of the discussions on automation surround how it will impact manual labor. However, while it's easy to see how robotics can take over precise and repetitive manufacturing tasks, the rise of artificial intelligence (AI) is allowing for more cognition-based tasks to be taken over by computers as well. Chui stated, "In about 60 percent of occupations, over 30 percent of the things that people do could be automated -- either using robots or artificial intelligence, machine learning, deep learning, all of these technologies that we're hearing more and more about."

Intel (NASDAQ:INTC) has made a number of acquisitions as well. Saffron specialized in cognitive computing, combining data analytics with deep learning; this relates directly to Intel's efforts in the same area. Indisys provided natural-language recognition, gesture recognition, and virtual-assistant technologies, but also created user interfaces for unmanned drones. Itseez focused on software for the Internet of Things (IoT), cameras, drones, and autonomous driving. Movidius brought computer-vision hardware for drones and cameras; it also provided system-on-a-chip (SoC) technology for accelerating computer vision, as well as deep-learning capability.

DeepMind is a British AI startup which was relatively unknown until it was bought by Google for around $600 million in 2014. Since then DeepMind has continued to refine its neural-network driven technology which has broken new frontiers with machine learning, particularly deep learning. Perhaps DeepMind's most famous accomplishment so far is being the brains behind AlphaGo, the first computer program to beat a professional human player of the board game Go. AlphaGo was developed by feeding DeepMind's machine learning algorithms with 30 million moves from historical tournament data, and then having it play against itself and learn from each defeat or victory. DeepMind's work is based on a solid grounding in neuroscience.