Artificial intelligence can do a lot of amazing things. It's helping medical researchers, aiding in just about every computational process, and beating people in lots of games The AIs Are Winning: 5 Times When Computers Beat Humans The AIs Are Winning: 5 Times When Computers Beat Humans Artificial intelligence is getting good. In fact, computers are now beating the best and brightest minds that humanity can offer. What does that mean for us? It's even making your phone smarter 7 Siri Alternatives for Android: Google Now, Cortana, & More 7 Siri Alternatives for Android: Google Now, Cortana, & More There are a lot of personal assistant apps in the Play Store, so we help you sort through the best of them.

Google's DeepMind AI is learning how to talk. And learning how to do it like a person, not a computer. DeepMind has many learning projects going on right now, but the newest one to catch our ears seems to be an increasingly realistic voice and speech pattern system that eliminates more and more of the inhuman, robotic patterns we use to identify computers. Imagine if Siri, Cortana, or Alexa started having inflection, variances, and realistic breathing patterns. Instead of sounding like this, it might sound like this.

On Thursday, Amazon announced that it is adding its voice assistant Alexa to its Fire tablets for the first time. The voice assistant has previously been in Amazon's Echo, Fire TV and Fire Stick devices, but will launch with the new, low-priced 90 Fire HD 8 tablet that ships on Sept. 21. Alexa is also coming to the Fire HD 10, Fire and last year's Fire HD 8 tablets via a free over-the-air software update in the coming months. This move is significant, as it puts Alexa more at odds with Siri, Apple's voice assistant, and pitches up the simmering competition in the voice-assistant space. The end game of all these voice assistants is to become an integral part of users' lives.

Twitter has announced a new "skill" for Echo devices called Twitter Reader. The company tweeted that you can now "use the Twitter Reader to hear trends, top Tweets, notifications & more!" A report yesterday revealed that Amazon might open up its Alexa ecosystem to allow push notifications, so that it could interrupt to warn you about a traffic jam, for instance. However, Twitter also added "Just #AskAlexa" hashtag, so the new feature apparently doesn't support push notifications just yet. It does give you a lot of options -- you can ask Alexa to read back recent tweets, trending topics and even your own tweets. The device is surprisingly prudish and censors profanity, as PC Mag discovered, so if your feed is like mine, that could be a lot of beeping.

Welcome to DeveloperEconomics' weekly news roundup. In this edition, Google cancels its modular smartphone project, Apple plans to delete abandoned apps from the App Store and Epic releases a new version of Unreal Engine 4. Read on for the full news rundown. Google has expanded its Early Access beta testing, allowing any developer apply for the program. Previously, Early Access was only available to 29 hand-picked developers, who got to test out their apps with early-adopters (although other devs were given different testing options). Google says the Early Access program generated over 1 million beta installs since launching in May and "demand is only growing."

As it pulls the plug on iPhone jacks, Apple is making what appears to be a concerted move into machine learning as it seeks to upgrade its next operating system to improve its Siri virtual assistant along with emerging machine learning applications expected to find their way into other Apple devices. Apple (NASDAQ:AAPL) watchers point to last month's acquisition of machine learning startup Turi for about 200 million as another sign that the consumer electronics giants is laying the groundwork for a machine learning push. The Seattle-based startup's machine learning platform focuses on rapid development of real-time services and applications based on embedded machine learning models. Turi is seen as a natural fit for Apple's machine learning push since its application toolkits based on the Python programming language are designed to simplify development of machine learning models that can be embedded into applications and quickly scaled.

When will iOS 10 be released in the UK? What are iOS 10's best new features, and what are the differences between iOS 10 and iOS 9? Which iPhones, iPads and iPod touch models can run iOS 10, and is it possible to install iOS 10 right now? It showcased the software as it had done at WWDC in June. We've got all the details about iOS 10 that you need to know: all of its best and most interesting features, how to install it, and a list of the iPads and iPhones that can run it. Since then Apple has released a series of beta versions of iOS 10, and the final public version will roll out on 13 September to iPhone 5 and later and the iPad mini 2, iPad 4, iPad Air and later generations. In this article you'll find: the 10 best new features in iOS 10; when iOS 10 will be released to the public, and how to download it ahead of the official launch; a list of iPads and iPhones that can run iOS 10; and everything else you need to know about iOS 10 and its imminent launch. Updated, 9 September 2016, with the news that the iOS 10 final release candidate is being rolled out; on 7 September, following 13 September release date; on 31 August with news of Apple's newly announced press event on 7 September; on 24 August with news of iOS 10 beta 7; and on 23 August with iOS 10 release date rumours courtesy of Barbra Streisand Update, 9 September 2016: Apple has notified all members of its iOS beta program that the final release candidate is being rolled out today. "We are pleased to give you access to the iOS 10 and macOS Sierra final release candidates," says Apple, in an email. "We will be shipping the final versions to customers worldwide in the coming weeks and appreciate you trying out the releases and providing feedback. You can install the iOS 10 GM candidate from Software Update."

Fintech start-up Clinc says it has created "the world's first" open source platform for building artificial intelligence (AI) and deep learning-based personal assistants. Clinc calls it "the financial genie" or "Finie". It is pitched to banks, financial institutions and personal finance managers. The solution was developed by a group of professors at the University of Michigan, Jason Mars, Lingjia Tang and Michael Laurenzano. They set up Clinc in 2015, with Mars as CEO, Tang as chief scientific officer, and Laurenzano as CTO.

Tom Goodwin is senior vice president of strategy and innovation at Havas Media. With smartphones in all of our pockets, it's easy to laugh at Thomas Watson's reputed remark that "there is a world market for about five computers." But what if he was off by only four computers instead of four billion? In the early 2000s, we lived in an era of expensive storage and slow connectivity. Those of us tasked with predicting the future faced a dilemma: Would processing and storage get so cheap that each and every device could store everything ever made, or would connectivity become so robust and far-reaching that everything would be stored remotely.

In a variety of application domains the content to be recommended to users is associated with text. This includes research papers, movies with associated plot summaries, news articles, blog posts, etc. Recommendation approaches based on latent factor models can be extended naturally to leverage text by employing an explicit mapping from text to factors. This enables recommendations for new, unseen content, and may generalize better, since the factors for all items are produced by a compactly-parametrized model. Previous work has used topic models or averages of word embeddings for this mapping. In this paper we present a method leveraging deep recurrent neural networks to encode the text sequence into a latent vector, specifically gated recurrent units (GRUs) trained end-to-end on the collaborative filtering task. For the task of scientific paper recommendation, this yields models with significantly higher accuracy. In cold-start scenarios, we beat the previous state-of-the-art, all of which ignore word order. Performance is further improved by multi-task learning, where the text encoder network is trained for a combination of content recommendation and item metadata prediction. This regularizes the collaborative filtering model, ameliorating the problem of sparsity of the observed rating matrix.