Marc Carrel-Billiard, Managing Director Global Technology R&D, AccentureBangalore: AI can double annual economic growth rates by 2035 by changing the nature of work and spawning a new relationship between man and machine, according to Accenture Research. The impact of AI technologies on business is projected to boost labour productivity by up to 40 percent by fundamentally changing the way work is done and reinforcing people's role to drive growth in business. In an interview with ETCIO, Accenture's Managing Director Global Technology R&D, Marc Carrel-Billiard talks about company's technology labs and its focus areas, new technologies and its impact and key tech trends that CIOs and businesses need to look for and much more. Marc is with Accenture for the past 18 years and currently oversees the Accenture Technology Labs, Accenture Open Innovation, Accenture's global technology R&D organization which explores new and emerging technologies, across seven locations around the world. Which are the key technology domains that you are trying to focus on?

The World Chess Championship is nearing its close, with the final tie-breaking match between Norway's Magnus Carlsen and Russia's Sergey Karjakin set for Monday. The match, a series of 12 games that began on November 11th, has attracted celebrities, tech leaders, and high-profile media coverage. In part, that's thanks to its New York location, where chess has enjoyed a decade-long surge in popularity. The continuing popularity of chess might have been hard to predict in 1997, after IBM's Deep Blue defeated human World Champion Gary Kasparov (also in New York). Before the match, commentators thought a loss by Kasparov would diminish chess as a pursuit.

Web Summit, the massive technology conference in Lisbon, Portugal, polled 224 investors face-to-face at Venture, one of the subconferences within the Web Summit framework. A smaller percentage, only 53%, said that it is "inevitable that Artificial Intelligence will destroy millions of jobs." Arguably, we're already seeing this happen. Self-driving cars and trucks are closer and closer every day. Earlier this year, iPhone manufacturer Foxconn replaced 60,000 factory workers with robots.

The omnipresence of deep learning architectures such as deep convolutional neural networks (CNN)s is fueled by the synergistic combination of ever-increasing labeled datasets and specialized hardware. Despite the indisputable success, the reliance on huge amounts of labeled data and specialized hardware can be a limiting factor when approaching new applications. To help alleviating these limitations, we propose an efficient learning strategy for layer-wise unsupervised training of deep CNNs on conventional hardware in acceptable time. Our proposed strategy consists of randomly convexifying the reconstruction contractive auto-encoding (RCAE) learning objective and solving the resulting large-scale convex minimization problem in the frequency domain via coordinate descent (CD). The main advantages of our proposed learning strategy are: (1) single tunable optimization parameter; (2) fast and guaranteed convergence; (3) possibilities for full parallelization. Numerical experiments show that our proposed learning strategy scales (in the worst case) linearly with image size, number of filters and filter size.

Developers of text-to-speech synthesizers (TTS) often make use of human raters to assess the quality of synthesized speech. We demonstrate that we can model human raters' mean opinion scores (MOS) of synthesized speech using a deep recurrent neural network whose inputs consist solely of a raw waveform. Our best models provide utterance-level estimates of MOS only moderately inferior to sampled human ratings, as shown by Pearson and Spearman correlations. When multiple utterances are scored and averaged, a scenario common in synthesizer quality assessment, AutoMOS achieves correlations approaching those of human raters. The AutoMOS model has a number of applications, such as the ability to explore the parameter space of a speech synthesizer without requiring a human-in-the-loop.

This is the Proceedings of NIPS 2016 Workshop on Interpretable Machine Learning for Complex Systems, held in Barcelona, Spain on December 9, 2016

Variational inference is a powerful tool for approximate inference. However, it mainly focuses on the evidence lower bound as variational objective and the development of other measures for variational inference is a promising area of research. This paper proposes a robust modification of evidence and a lower bound for the evidence, which is applicable when the majority of the training set samples are random noise objects. We provide experiments for variational autoencoders to show advantage of the objective over the evidence lower bound on synthetic datasets obtained by adding uninformative noise objects to MNIST and OMNIGLOT. Additionally, for the original MNIST and OMNIGLOT datasets we observe a small improvement over the non-robust evidence lower bound.

Machine learning is a powerful way to access information about your customers in order to personalize the experience to meet their needs. James Staten, chief strategy officer for Microsoft Cloud, works with customers around the world and knows the importance of having a complete picture of how and when customers interact (or don't interact) with a brand. Instead of simply sorting customers into basic groups, machine learning can access huge data sets through the cloud, including data your company might not collect itself, such as social media analytics and information from retailers. The cloud allows users to aggregate huge amounts of data to give instant insights and predictive analysis. European soccer team Real Madrid uses these tools to create an amazing customer experience.

Chris Nicholson co-founded Skymind and Deeplearning4j, the most popular deep-learning framework for Java. Quitting Twitter is easy -- I've done it a hundred times. Someone called it "a clown car that drove into a gold mine," and like all clown cars, Twitter makes the passengers get out once in awhile. If I go back, it's because I'm addicted. For an information junkie, that little bubble is hard to resist.

When I were a lad, watching the news on the telly, waiting to be allowed to use the set to plug in my ZX Spectrum, I'd be told to concentrate on the stories from the nearby towns: car workers being laid off as robots took their jobs. Stay in school, son, and get into a profession. A degree and a place in a management trainee scheme was the preferred route. Don't make things, be a knowledge worker, I'd be told. Information is the new oil.