Computer companions for elderly people are'emotionally dangerous' and will never be good enough to take the place of real human contact, an artificial intelligence expert has warned. In December the University of Singapore introduced'Nadine' the world's most lifelike robot whose creators say will eventually provide childcare and offer friendship to lonely pensioners. More basic'care-bots' which simply show an image of a computerised face are already on the market for as little as 3,000. But Maggie Boden, professor of Cognitive Science at the University of Sussex, warned that machines would never be able to understand abstract ideas such as loyalty or hurt. "Computer companions worry me very much," Prof Boden told an audience at The Hay Literary Festival at Hay-on-Wye, Wales.

Classical collaborative filtering, and content-based filtering methods try to learn a static recommendation model given training data. These approaches are far from ideal in highly dynamic recommendation domains such as news recommendation and computational advertisement, where the set of items and users is very fluid. In this work, we investigate an adaptive clustering technique for content recommendation based on exploration-exploitation strategies in contextual multi-armed bandit settings. Our algorithm takes into account the collaborative effects that arise due to the interaction of the users with the items, by dynamically grouping users based on the items under consideration and, at the same time, grouping items based on the similarity of the clusterings induced over the users. The resulting algorithm thus takes advantage of preference patterns in the data in a way akin to collaborative filtering methods. We provide an empirical analysis on medium-size real-world datasets, showing scalability and increased prediction performance (as measured by click-through rate) over state-of-the-art methods for clustering bandits. We also provide a regret analysis within a standard linear stochastic noise setting.

Reinforcement learning for embodied agents is a challenging problem. The accumulated reward to be optimized is often a very rugged function, and gradient methods are impaired by many local optimizers. We demonstrate, in an experimental setting, that incorporating an intrinsic reward can smoothen the optimization landscape while preserving the global optimizers of interest. We show that policy gradient optimization for locomotion in a complex morphology is significantly improved when supplementing the extrinsic reward by an intrinsic reward defined in terms of the mutual information of time consecutive sensor readings.

By analyzing Rembrandt's oeuvre, machine learning has simulated a new work by the celebrated artist Rembrandt, who died 347 years ago, has'created' a new work all thanks to machine learning. By studying how the major Impressionist painter created his trademark spotlight effect in many of his portraits, the team at Microsoft were able to render a painting that Rembrandt, in flesh, could have painted himself. The process began by building a 150 gigabyte-pile of data on Rembrandt's'artistic DNA.' Scans, including 2D and 3D, were upscaled with deep learning algorithms. The'Next Rembrandt' is a portrait of a man with white collar. After studying his portraits made between 1632 and 1642, the following attributes for the likely subject were discovered: Caucasian male with facial hair, in his thirties, with a white collar on black clothing, a hat and with the head tilted to the right. The team then used algorithms to compute Rembrandt's ideal proportions in his paintings.

Can a machine be creative? Google thinks so, and it has an entire team dedicated to teaching machines how to view the world a little more like us emotional humans. Think about computers as if they were children and it's simple to understand how coders can teach them to learn. Artificial intelligence is, at the beginning, very basic and simple. Human moderators instruct computers, showing them how to think and thus teach themselves.

This is a post based on the talk given at Warsaw Data Science Meetup on April 12th 2016. It's difficult to make predictions, especially about the future. Nowadays, whenever we hear about a recent breakthrough in science, technology or business, it is more than likely that the so-called'big-data' is part of the success. Whether it's a new medical algorithm to discover early stages of cancer or a new method for reducing traffic in big cities, massive amounts of digital data are very often a'secret ingredient'. As a matter of fact, everyone of us has probably heard at least a couple of the estimates on how much data human mankind creates every second, every minute or every hour.

The assistant wars are in full swing, with Google Assistant and Viv entering the fray and existing players such as Apple's Siri, Microsoft's Cortana, Hound and Google's Now all stepping up fight for consumers' hearts and minds. Apple's rumored Amazon Echo competitor, which VentureBeat believes is a next-gen Apple TV, could blow all the assistants out of the water when it comes to deciphering complex natural language queries. And powering it--VocalIQ, a sophisticate technology Apple acquired back in October 2015. Tech Insider provided an in-depth overview of how VocalIQ could make Siri a lot smarter than it is today. Tech Insider's piece, titled "Apple is working on an AI system that wipes the floor with Google and everyone else," notes that Apple had been testing VocalIQ technology against Siri, Google Now and Cortana. "Users asked each AI questions using normal language, not the robotic commands you're used to using with digital assistants," reads the article.

As a natural response to Facebook's push to get more pages to use Messenger as a platform, brands are developing chat bots to handle customer service. How well is this working? At the Socialbakers Engage conference in Prague earlier this month, experts talked about the future of chat bots. The chat bots companies are starting to develop are more like the first step of a customer interaction, allowing them to take care of low-level inquiries, and then provide information to a human to complete the process. We're going to see a lot more intelligent bots out there.

Can a machine be creative? Google thinks so, and it has an entire team dedicated to teaching machines how to view the world a little more like us emotional humans. Think about computers as if they were children and it's simple to understand how coders can teach them to learn. Artificial intelligence is, at the beginning, very basic and simple. Human moderators instruct computers, showing them how to think and thus teach themselves.

Set mostly in the Earth-like world of Azeroth, Warcraft follows the premise set forth by Warcraft: Orcs and Humans, the franchise's first release for PC in late 1994. Having destroyed their own planet with dark magic, the Orcs attempt to take over the world of humans, setting off an epic clash of clans. What follows onscreen is a live action/CGI mashup visually reminiscent of Avatar that opts for a kinetic pace of nonstop battles over deep character study and plodding narrative. Jones wasn't originally attached to this Warcraft adaption -- Sam Raimi (The Evil Dead) was. But, having been a longtime fan, he jumped at the chance to pick up where Raimi left off and build a film that lets audiences decide who the good guys are. In advance of the film's US release June 10th, I sat down with Jones to talk about his love of PC gaming, how spaghetti Westerns informed Warcraft's battles and his early sci-fi and fantasy inspirations.