Twitter users may soon notice a great improvement in the visual content of their apps, if their latest acquisition is any indicator. Twitter recently acquired a new-ish startup based in London specializing in artificial intelligence (AI). The company, Magic Pony, uses machine learning to build improved systems for visual processing. More specifically, according to Twitter's blog post, Magic Pony's technology is, "based on research by the team to create algorithms that can understand the features of imagery and will be used to enhance Twitter's strength in live and video." Twitter plans to add this technology to other recent acquisitions including Madbits and Whetlab.

Machine learning is set to go mainstream in the next two years, according to a new survey by software and application development vendor SoftServe Inc. The new report, based on an April poll of 300 U.S. and U.K.-based medium and large enterprises, shows that 62 percent of firms expect to roll out machine learning-based tolls for business analytics within the next two years. The majority of those companies said the most promising opportunity for machine learning lays in real-time data analysis. According to SoftServe, the survey is evidence that machine learning is moving past the "hype cycle", with enterprises looking to automate analytics processes in areas like business intelligence and cyber security. In the latter area, further evidence of machine learning's progress comes from the Defense Advanced Research Projects Agency (DARPA), which is sponsoring an "all-machine" hackathon at the DEF CON hacking conference in Las Vegas next month.

Reuters – A former Google engineer, whose speech recognition software is used in more than a billion Android smartphones, has launched a company that uses blockchain technology to build a new operating system for banks. Paul Taylor, a Cambridge University academic with an expertise in artificial intelligence, speech synthesis and machine learning, started working on the system, called Vault OS, two years ago in a basement in London's Shoreditch district, known for being a tech startup hub. Blockchain technology has captured the imagination of the financial industry, with advocates saying it has the potential to shake up how markets operate. The technology, which underpins the digital currency bitcoin, creates a shared database in which participants can trace every transaction ever made. The ledger is tamper-proof and transparent, meaning that transactions can be processed without the need for third-party verification.

Microsoft Azure Machine Learning simplifies data analysis and empowers you to find the answers your business needs. The question isn't whether you can find the answers.

So the question I have is: what does the frontier of the space of optimal networks look like, what are the inherent limits of depth vs expressivity of these models, and are there dimensional scaling laws that can describe all this in an information theoretic way? This recent paper gives a great treatment on the expressivity of convolution networks by using a deep layered architecture that generalizes convolutional neural networks called sim-nets. As a simple first step I wanted to see what could be done to visualize the operations a deep neural net performs. So I constructed a standard network that takes vector inputs of size 2 and produces vector outputs of size 3 which we can think of as a mapping of the cartesian plane into RGB color space. Taking many copies of this net and randomly initializing them, (with normally distributed weights and biases) we can plot them in a grid and see the networks' outputs as a set of images.

On Monday, I outlined my view of what makes "an analytic" and the jargon that goes along with it. I ended the post wondering whether "an analytic" is just another name for "business rules" that organizations tend to follow today. In my view, the main difference between the two, which might be no difference at all for a given situation, is that "analytics" are generally more complex mathematically and operate on a more general "feature space". This generalization allows rigorously developed techniques from statistics and applied mathematics to have a chance at being applied to a messy real world problem. Business rules tend to be more human understandable and more directly embedded into the specific data formats or information processing software used by a business.

Baidu has made its second investment in a U.S. fintech company inside a month after it put an undisclosed sum of money into ZestFinance, a big data firm specializing in credit scoring. Baidu, which operates China's dominant search platform, took part in a 60 million round for payments firm Circle in June. The deal is part of an agreement that will see Baidu use ZestFinance's technology to develop a credit scoring platform that is based on its search data. That's important in a market like China because traditional credit systems are broken there. There's precious little formalized credit history data while many people don't use banks heavily or are unbanked.

Monday 18th July 2016, Whitley: Sir Ben Ainslie has hailed Land Rover's artificial intelligence (AI) as a'game changer' ahead of the British America's Cup team's home event in Portsmouth (22-24 July). Land Rover, Title and Exclusive Innovation Partner to Land Rover BAR, is applying its big data processing power and machine learning expertise to help co-engineer the fastest boat in America's Cup history and bring the world's oldest sporting trophy to the UK for the first time. Land Rover engineers, embedded into the team for over a year, are using artificial intelligence to explore and find patterns in sailing performance data to help'make the boat go faster'. When testing, the sailing team receives over 16 GB of uncompressed data per day from sensors on the boat – the equivalent of filling an iPhone's memory. The ability to process and make sense of this volume of data is unprecedented in sailing.

She joined as senior reporter in April 2014 having previously worked as assistant editor at Government Computing. There has been an unusually large focus devoted to chatbots within the tech community in 2016, thanks to news like Facebook's announcement it will integrate chatbot capabilities into Messenger, and Microsoft's controversial'teen girl' Twitter bot'Tay'. However how much of this has actually translated into real-world applications? It's early days, but there are a few companies and organisations that have started to try out chatbots as part of their operations. Here are a few of the most notable examples.

Artificial Intelligence and Cognitive Computing has an increasing importance in the interconnected digital world. Data is being generated at a pace like never before. Abundance of Data generated by machines, social media, and others, along with advanced algorithms are allowing us to overcome the barriers of machine knowledge and opening new vistas of intelligence and insights being generated from data. We will look into some of the latest applications of Artificial Intelligence and Cognitive computing and how various industries are getting transformed by this phenomenon. We will also explore some of the cutting edge algorithms that are playing a role in this digital revolution.