A group of scientists from Stanford University developed an artificial intelligence that "mines" human activities from publishing site Wattpad. The site contains works of fiction that depict the modern world much better than classic literature. To counter the fear that artificial intelligence (AI) could someday become a threat to humanity, a team of scientists last month revealed a new algorithm that will teach robots how to behave appropriately in social situations. The key to this is by allowing them to read and understand children's stories. Now, another group of experts has taken a wider route. Instead of limiting it to children's stories, developers used fiction from Toronto-based Wattpad to help their knowledge-based program called Augur understand the world.

The most common approach to address'Machine Learning' these days is to think of it as just another arduous subject, one which young computer scientists choose in their last two years of college. A huge percentage of people are not able to realise that this'subject' is actually one of the most important strategies that humans use throughout their lifetime. Let me start with a simple example. During childhood, every kid is curious about how the surroundings around him or her work – anyone with any experience of young children will know you have to be very careful about their activities! You can't just let it go near a burning flame, near sharp knives or let it eat just anything off the floor.

Heretofore, we humans have always been inherently narcissistic in our embrace of progress. From our nervous fascination with sexbots to the anthrocentric blinders that guide our search for aliens, our innovations are rarely more than a mirror shining back on our own needs. And why wouldn't it be selfish? The goal is to make the world a better, more interesting place, but those are naturally subjective targets. The big question, then, is if we keep pushing technology and discovery further and further, who is it for? This is what Rose Eveleth asks in her new column, Design Bias: tech is reflective of the people who build and design it, which means that if the builders lack a diverse perspective, so will the end product.

It's kind of confusing because I'm a bigger girl, Dalbesio says. I'm not the biggest girl on the market but I'm definitely bigger than all the girls [Calvin Klein] has ever worked with, so that is really intimidating. She wasn't sure, she said of the shoot, what was expected from her in terms of her size or shape. Refreshingly, what was expected of her was the same thing that was expected of Lara Stone: to take a beautiful picture. So simple, yet so essential, the white shirt is the foundation of any wardrobe.

Speaking at Google Cloud Platform in San Francisco, the chairman of Alphabet and former Google CEO highlighted artificial intelligence as the "next evolution" in computing. This will will spur the creation of new apps and services, he said. "Machine learning and crowdsourcing data will be the basis and fundamentals of every successful, huge IPO win in five years, in the same sense that the transition to [mobile] apps five years ago created the modern corporations of Uber, Snapchat and others." Schmidt highlighted Google AlphaGo - which defeated Go champion Lee Sedol last week - as an example of the power of artificial intelligence. See also: Google DeepMind: What is it, how does it work and should you be scared? He also cited Google's Photos image search software as a practical example of how machine learning can be applied to create new products and services when applied to large crowd-sourced datasets.

At its GCP NEXT 2016 conference, the Google Cloud Platform team announced a new offering designed to mainstream the development of machine learning (ML) apps. Developers can now sign up for a limited preview of Google Cloud Machine Learning, a managed platform that helps build ML models with all kinds of data coming in all sizes. "Cloud Machine Learning will take machine learning mainstream, giving data scientists and developers a way to build a new class of intelligent applications," the company said in a blog post yesterday. "It provides access to the same technologies that power Google Now, Google Photos and voice recognition in Google Search as easy to use REST APIs." Google already has a portfolio of pre-trained ML models, and has added a new one, the Google Cloud Speech API, joining the likes of the Translate API and the Vision API.

A company that helps users to create their own websites now knows what kind of sites their 80 million users are building without pestering them with repeated questions. Wix, a Tel Aviv-based web development company, is using machine learning on Google's cloud platform to learn more about its users so it can help them find the images they need to build interesting and useful websites. That's just the beginning of how machine learning will be used in the cloud, according to industry analysts who say machine learning will be the biggest thing that's ever hit the cloud. David Zuckerman, head of developer experience for Wix, said machine learning in the cloud will be a boon to companies that don't have a major research division. "The cloud has brought this technology to everyone," he said.

K-means clustering is one of the simplest clustering algorithms one can use to find natural groupings of an unlabeled data set. Another way of stating this is that k-means clustering is an unsupervised learning algorithm. "learning the structure of X without being given Y". K-means clustering finds "k" different means (surprise surprise) which represent the centers of k clusters and assigns each data point to one of these clusters. The cluster it is assigned to is the one where the distance (usually Euclidean) from the point to the mean is smallest.

This series of blog posts aims to provide an intuitive and gentle introduction to deep learning that does not rely heavily on math or theoretical constructs. Everything in life depends on time and therefore, represents a sequence. To perform machine learning with sequential data (text, speech, video, etc.) we could use a regular neural network and feed it the entire sequence, but the input size of our data would be fixed, which is quite limiting. Other problems with this approach occur if important events in a sequence lie just outside of the input window. What we need is (1) a network to which we can feed sequences of arbitrary length one element of the sequence per time step (for example a video is just a sequence of images; we feed the network one image at a time); and (2) a network which has some kind of memory to remember important events which happened many time steps in the past.

Here's What We Can Expect from Deep Learning in 2016 and Beyond Can you use robotics to double the size of your house?