Enterprise cloud usage has been in the forefront of big players for the past few years. Amazon, IBM, Google and Microsoft are expanding their offerings to serve better the enterprise users and their needs. Google announced a set of machine learning based services focused on enterprise users. Similar to upcoming Amazon EC2's Elastic GPUs and Microsoft's Azure N-Series, powered by NVidia GPUs, Google will soon offer cloud based GPUs with per minute billing focused on Machine Learning tasks. Google slashed pricing for its Cloud Vision API to 1/5, offering face, label, OCR, company logos, explicit content and landmark and image properties recognition through off the shelf algorithms and their API.

In the closing weeks of 2016, Google published an article which quietly sailed under most people's radar. Which is a shame, because the article may just be the most astonishing thing about machine learning that I read last year. Don't feel bad if you missed it. Not only was the article competing with the pre-Christmas rush most of us were navigating, it was also tucked away on Google's Research Blog beneath the geektastic headline Zero-Shot Translation with Google's Multilingual Neural Machine Translation System. It doesn't exactly scream must read, does it?

The capacity of a neural network to absorb information is limited by its number of parameters. Conditional computation, where parts of the network are active on a per-example basis, has been proposed in theory as a way of dramatically increasing model capacity without a proportional increase in computation. In practice, however, there are significant algorithmic and performance challenges. In this work, we address these challenges and finally realize the promise of conditional computation, achieving greater than 1000x improvements in model capacity with only minor losses in computational efficiency on modern GPU clusters. We introduce a Sparsely-Gated Mixture-of-Experts layer (MoE), consisting of up to thousands of feed-forward sub-networks. A trainable gating network determines a sparse combination of these experts to use for each example. We apply the MoE to the tasks of language modeling and machine translation, where model capacity is critical for absorbing the vast quantities of knowledge available in the training corpora. We present model architectures in which a MoE with up to 137 billion parameters is applied convolutionally between stacked LSTM layers. On large language modeling and machine translation benchmarks, these models achieve significantly better results than state-of-the-art at lower computational cost.

Mr. HUANG talked about the challenges facing the industry in the era of AI and how it could translate those challenges into opportunities. He pointed out that the emergence of neural machine translation, the application of speech interactive technology and the rapid development of the Internet Era had a huge impact on traditional human translation. Though the industry has begun to realize the importance of AI and other technologies, the mechanism and platform to promote their application is still lacking. Mr. HUANG believed that governments, language colleges, language and AI service providers, experts and freelance translators should join hands in exploring this topic and advise the industry on development in the new era. In addition to the guidance of government and industrial organizations, an effective cooperative mechanism shall be established to ensure the maximum synergy between colleges, research institutes and language service providers.

Editor's note: This is the first chapter of a book I'm working on called Demystifying Artificial Intelligence. The goal of the book is to demystify what modern AI is and does for a general audience. So something to smooth the transition between AI fiction and highly mathematical descriptions of deep learning. I'm developing the book over time - so if you buy the book on Leanpub know that there is only one chaper in there so far, but I'll be adding more over the next few weeks and you get free updates. The cover of the book was inspired by this amazing tweet by Twitter user @notajf. Feedback is welcome and encouraged!

The potential usefulness of automatic computerized translation was recognized by the very first AI researchers in the 1950s. But it wasn't until new algorithms emerged in the 1980s and 1990s that the field made significant progress. Now, translation tools of great sophistication are playing a growing role in both everyday office use and for specialized fields, as the economy becomes increasingly globalized and companies sell products and services in multiple markets. The poster child for computer translation is Google Translate, the easy-to-use, general purpose Web-based translation engine that can handle nearly 60 languages. Google's Translate has the same 800-pound gorilla status in its world that the company's namesake product does in search.

A renewed international effort is gearing up to design computers and software that smash language barriers and create a borderless global marketplace. A woman sits at a desk in Manhattan, talking to herself in French. The phrases she balances on each breath are musical to American ears. She has postcards of Montreal tacked up on the walls of her cubicle – pastel-painted houses in the snow – so as she sculpts the contours of each syllable, she can remind herself of the place where the sounds she's making are heard every day in the street. Her name is Guylaine Laperrière, and she came to New York City more than a decade ago to study musical theater. One day, a friend asked her if she wanted to make a little cash dubbing a French voice-over for a promotional short about insurance. She took the job, and was surprised how much she enjoyed bringing ideas from one language home into another. This article has been reproduced in a new format and may be missing content or contain faulty links.

Compiled by Carl Zimmer (zimmer@panix.com) Natural Language Laboratory, School of Computing Science, Simon Fraser University Burnaby, British Columbia Researchers are devising "relaxed grammars" to extract the sense of transcribed speech through a method known as "partial parsing," which deciphers chunks of language rather than breaking down the structure of entire sentences. Their work is being integrated into closed-captioning technology for real-time TV translations. This article has been reproduced in a new format and may be missing content or contain faulty links. Contact wiredlabs@wired.com to report an issue.

This article has been reproduced in a new format and may be missing content or contain faulty links. Contact wiredlabs@wired.com to report an issue. The outcome is a halt in federal funding for machine translation R&D. Darpa launches its Spoken Language Systems (SLS) program to develop apps for voice-activated human-machine interaction. Researchers focus on portable systems for face-to-face English-language business negotiations in German and Japanese.

Running software that took four years and millions of dollars to develop, Carbonell's machine – or rather, the server farm it's connected to a few miles away – is attempting a task that has bedeviled computer scien tists for half a century. The message isn't encrypted or scrambled or hidden among thousands of documents. I brought along the text, taken from a Spanish newspaper transcript of a 2004 al Qaeda video claiming responsibility for the Madrid train bombings, to test Meaningful Machines' automated translation software. The brainchild of a quirky former used-car salesman named Eli Abir, the company has been designing the system in secret since just after 9/11. Now the application is ready for public scrutiny, on the heels of a research paper that Carbonell – who is also a professor of computer science at Carnegie Mellon University and head of the school's Language Technologies Institute – presented at a conference this summer. In it, he asserts that the company's software represents not only the most accurate Spanish-to-English translation system ever created but also a major advance in the field of machine translation. This article has been reproduced in a new format and may be missing content or contain faulty links.