Technology created by an Australian startup is going to bring the conservation of orang-utans in a Sumatran jungle into the digital age. Adelaide-based Jemsoft's Monocular API is a computer vision machine learning technology, which can be taught to search for items in the real world. Under a new agreement the company will licence the software to conservationists to be applied to footage taken via drone above the canopies to identify nests of the endangered mammal. Rather than conservationists sifting through photos and video manually, Jemsoft's intelligent software can be trained to identify the orang-utan nests. Jemsoft, founded in 2013, is already being used by a number of large enterprise clients, including Fortune 500 companies, for other purposes.

The purpose of this blog post is to demonstrate how to install the Keras library for deep learning. Let me start by saying that Keras is my favorite deep learning Python library. It's a minimalist, modular neural network library that can use either Theano or TensorFlow as a backend. Furthermore, the primary motivation behind Keras really resonates with me: you should be able to experiment super quickly -- going from idea to result, as fast as possible. Coming from a world that mixes both academia and entrepreneurship, the ability to iterate quickly is extremely valuable, especially in the deep learning world where it can take days to weeks to train just a single model.

It's not easy to be healthy. And it's even harder to be healthy at work, where chances are--despite the vogue for standing desks and the like--you're parked in a chair for most of the day, focused on a screen. The average workweek, by one recent measure, is now 47 hours and counting. By and large, more time spent at the office means more time hunched over a computer, probably eating lunch at a desk. Stress--which has been linked to increased risk of stroke, heart disease, and other health issues--is just about inevitable as a result.

Turing test or no Turing test, Intelligent Virtual Assistants are here to stay. As more and more of our daily interactions are between people and artificial intelligence systems, the demands for meaningful and effective interactions is increasing. Expectations from users of intelligent agents need to be the source of great attention, with the appropriate planning, modeling and measurement in place to guarantee the desired performance and results. Join us as we explore the topic of "Training the Robots" where our experts will discuss the elements needed to build virtual assistant capabilities including identification of target processes for automation/AI enhancement, capturing and organizing process knowledge, and contextualizing virtual assistant /human interactions.

Artificial intelligence (AI) and a world in which machines threaten humanity's status quo has been the preserve of science fiction for decades. In the Eighties, Terminator was set in a post-apocalyptic world in which cyborgs rule, RoboCop's protagonist was part-man, part-machine and Short Circuit toyed with the idea of robots developing human-like minds, with rather more endearing results. The reality is the bot has bolted. AI is walking and talking among us. In 2016, we use voice-recognition systems, driverless cars are being trialled and robotic hotel receptionists work in Japan.

There's no doubt Stephen Hawking is a smart guy. But the world-famous theoretical physicist recently declared that women leave him stumped. "Women should remain a mystery," Hawking wrote in response to a Reddit user's question about the realm of the unknown that intrigued him most. While Hawking's remark was meant to be light-hearted, he sounded quite serious discussing the potential dangers of artificial intelligence during Reddit's online Q&A session: A superintelligent AI will be extremely good at accomplishing its goals, and if those goals aren't aligned with ours, we're in trouble. Hawking's comments might seem unrelated.

San Francisco: Google announced a deal to buy Moodstocks, a French startup behind technology that helps smartphones recognise whatever they are aimed at. Moodstocks caught the US technology giant's eye for its work in computer vision and machine learning, as well for accomplishments in enabling smartphones or other mobile devices to recognise images and objects. Google is among Silicon Valley titans investing in ways to get computers to see and understand the world around them the way people do. Machine learning has been woven into an array of Google offerings, such as its free language translation and photo services. "There's a lot more to be done to improve machine vision," Google France tech site lead Vincent Simonet said in a blog post.

Adaptive systems are a special class of open systems that are continuously evolving. They feature dynamic networks of agents that interact with each other and their environments. Examples of complex, adaptive systems include ant and bee colonies, the stock market, and biological ecosystems, as well as human organizations such as political parties, companies, and cities. And because of its ability to shift and change, the behavior of the whole system cannot be predicted by just analyzing its parts in isolation. So what happens when we reimagine a company more like an adaptive organism, rather than a stable machine?

The discovery of major software flaws could render thousands of fMRI brain studies inaccurate. The use of fMRI is a common method for scanning the brain in neuroscience and psychology experiments. To make sense of the data produced, researchers sometimes use a technique called spatial autocorrelation to identify areas of the brain that appear to "light up" during particular tasks or experiences. But some software flaws in the popular fMRI data analysis packages SPM, FSL and AFNI meant this technique routinely produced false positives, resulting in errors 50 per cent of the time or more. Anders Eklund and Hans Knutsson at Linköping University in Sweden and Thomas Nichols at the University of Warwick, UK, calculated this by analysing brain data from a collaborative open fMRI project called 1000 Functional Connectomes.

This is a guest post. The views expressed here are solely those of the author and do not represent positions of IEEE Spectrum or the IEEE. The simple task of picking something up is not as easy as it seems. Roboticists aim to develop a robot that can pick up anything--but today most robots perform "blind grasping," where they're dedicated to picking up an object from the same location every time. If anything changes, such as the shape, texture, or location of the object, the robot won't know how to respond, and the grasp attempt will most likely fail.