Machine learning, which helps computers do things like understand complex voice commands and improve image search capabilities, can be taxing on traditional hardware. Google should know – over 100 of its products and features use this technology to run and improve themselves constantly. The company has revealed that over the past few years, it quietly developed its own custom processor for such tasks. The Tensor Processing Unit (TPU) is built expressly for running TensorFlow, Google's in-house machine learning system that it open-sourced last year. Our biggest ever edition of TNW Conference is fast approaching!

May 18 (Reuters) - Salesforce.com Inc, a marketing-and-sales software provider for 17 years, may be joining the ranks of the technology old-guard, but the 27 percent first-quarter revenue growth it posted on Wednesday still showed some startup zip. The company also provided a rosy outlook, raising its full-year revenue forecast and predicting that technology like artificial intelligence would drive future growth. Shares of Salesforce, seen as a barometer for the cloud-computing sector, rose 5.9 percent in after-hours trading. Cloud-delivered software, provided online on a subscription basis, has gained popularity with businesses due to its flexibility and cost benefits. Salesforce, founded in 1999, has won market share from more traditional software providers such as Oracle Corp and SAP AG, Chief Executive Marc Benioff reminded analysts on a conference call after the company reported its earnings.

Artificial Intelligence (AI) techniques have been used extensively for understanding how systems and humans interact. Such techniques can be applied in the context of cyber security to enable a better understanding of how human beings interact with cyber systems. Resilience in cyber security systems can be augmented using these techniques, including Machine Learning, Natural Language Processing and Game theory. AI does provide us with the tools that enable greater cyber threat intelligence, as we try to stay one step ahead of the criminals. However, AI itself poses a number of challenges not least around the ethical question of using "human-like" machines.

SoftBank Robotics Corp. said Thursday it has expanded opportunities for app developers to use Google's Android platform for its humanoid robot Pepper. It hopes to encourage more programmers to develop apps for the device. An Android-ready Pepper model will go on sale in July, though only for developers, and on Thursday an Android development toolkit was made available for download, Fumihide Tomizawa, SoftBank Robotics chief, told a news conference at the firm's headquarters in Tokyo. Previously, there was a development platform specially designed for Pepper, but it was not based on Android. Tomizawa said the number of developers could increase 100-fold, as there are only several thousand Pepper app designers but hundreds of thousands of Android coders.

If there's one thing that I've learned in the few short years that I've been a fully-fledged scientist, it's that time is one of the most valuable commodities that you can give a researcher. In all its myriad forms, time is invaluable to the scientific process – time to develop ideas, time to write grants. The time that you need to run an experiment. Critically, the time that participants are willing to give you in the pursuit of knowledge. It's a precious thing, for everyone involved.

'How to be good?' is the pre-eminent question for ethics, although one that philosophers and ethicists seldom address head on. It was the question Plato posed in a slightly different form in The Republic when he said, "We are discussing no trivial subject, but how a man should live." Marcus Aurelius thought he knew the answer. When he unequivocally stated in his Meditations "A King's lot: to do good and be damned." He was himself a king and ruled almost all of the world that was known to him.

Despite plenty of excitement it's still unclear how conversational UIs can be made to work in a practical sense. But opinionated design principles can help us push past the hype, and design something real people will want to use every day. We're not lacking for self-assured sermons on how conversational UIs are the future. Much less is written about the practicalities of actually designing chatbot interactions. Yet it seems like this is precisely what we now need.

Game lovers are likely to be familiar with Nvidia's graphics processors. The US tech firm continues to derive the bulk of its revenue from gaming-related semiconductors, but there is now a new growth segment -- chips for artificial intelligence (AI). Business related to deep learning, a core part of an artificially intelligent machine, is one key factor behind Nvidia's impressive first-quarter results. The advanced machine learning function is being applied to numerous areas, including speech recognition, video search, face detection and self-driving vehicles. Nvidia is already working closely with automaker Tesla in development of driverless car technology, which deploys AI to detect pedestrians and track lanes and to recognize traffic signs.

I've written a number of blog posts about regression analysis and I've collected them here to create a regression tutorial. This tutorial covers many aspects of regression analysis including: choosing the type of regression analysis to use, specifying the model, interpreting the results, determining how well the model fits, making predictions, and checking the assumptions. At the end, I include examples of different types of regression analyses. If you're learning regression analysis right now, you might want to bookmark this tutorial! Before we begin the regression analysis tutorial, there are several important questions to answer.

They continue, "(2) Speed and Scale Matter. In order to analyze, swiftly and accurately, billions of events in real-time, machine learning models require a level of computational power and scalability that cannot be accomplished using old-school on-premise architecture and conventional database methods. Cloud-based architectures can significantly augment the efficacy of machine learning. Algorithms can be infused with the collective knowledge of a crowdsourced community where threat intelligence is aggregated and updated instantly. Identified attacks can then be turned into a new detection and learned by the algorithm, and shared with others within the cloud network to prevent the attack – sending the bad actors back to the drawing board."