DeepMind researchers have trained an AI to solve computer coding challenges as well as the average person. In a paper published last week in the journal Science, the group from Google’s AI division described how AlphaCode performed when pitted against human programmers, ranking in the top 54.3 percent in simulated tests, commensurate with “approximately human-level performance.” “This performance in competitions approximately corresponds to a novice programmer with a few months to a year of training,” according to Science, which says about half the humans who compete in coding contests could outperform the AI. Continue reading Deepmind’s AlphaCode AI Can Program Like Human Coders

Los Altos, CA-based startup Cerebras, dedicated to advancing deep learning, has created a computer chip almost nine inches (22 centimeters) on each side — huge by the standards of today’s chips, which are typically the size of postage stamps or smaller. The company plans to offer this chip to tech companies to help them improve artificial intelligence at a faster clip. The Cerebras Wafer-Scale Engine (WSE), which took three years to develop, has impressive stats: 1.2 trillion transistors, 46,225 square millimeters, 18 gigabytes of on-chip memory and 400,000 processing cores. Continue reading Cerebras Builds Enormous Chip to Advance Deep Learning

Visual effects studios are increasingly using cloud computing technology to crank out film sequences faster and more efficiently. Rendering, the process of finalizing each frame, is an intensive processing process, and the cloud allows studios to use as much processing power as needed to finish the job. When the studio is in between film projects, it can save money because the studio can scale back on its computing power demand, which helps keep VFX costs low for filmmakers. Continue reading VFX Studios Leverage the Power of the Cloud for Film Projects

Award-winning engineer Yves Faroudja has come out of retirement to explore how new approaches to compression and processing can help with the transition to Ultra HD. Los Altos-based startup Faroudja Enterprises is currently testing a prototype technology to move 4K in real time, while preserving image quality at reduced bit rates with existing compression systems. Faroudja has filed for eight related patents, and the company has plans to either license the tech or introduce it as software. Continue reading Ultra HD: New Processing Tech Improves Video Compression

Solid-state batteries, which have been used for wireless sensors but are typically considered too expensive for most devices, can now be manufactured much cheaper, according to Applied Materials. The company, which supplies equipment for semiconductor and display industries, says that these longer lasting batteries can be used in anything from smartwatches to electric cars. The company plans its first commercial use of the batteries in wearable devices, where size is a limitation.  Continue reading Longer-Lasting Solid-State Batteries May Power New Wearables

Amazon Local Register, the company’s new card reader, is a black rectangular device that plugs into a merchant’s smartphone or tablet. For $10, it works with an app that tracks business transactions for merchants. Amazon offers lower processing fees with the device if a merchant signs up before the end of October, while the standard deal will remain a 2.5 percent charge, which is less than other card reader offers. The lower charge means that Amazon is likely making little profit. Continue reading Mobile Payments: Amazon Debuts New Credit Card Reader

Researchers at MIT have developed a method for teaching computers to understand what is happening in video content. The method uses a similar approach to textual analysis, such as natural language processing, by looking at each part of a video to figure out what the whole thing means. The researchers created an algorithm to identify what occurs in individual frames of the video, and then determines what those mean when combined in a certain order. Continue reading Researchers Use Algorithm for Computers to Understand Video

Google recently announced a new initiative dubbed “Project Tango” that plans to integrate 3D sensor technology in a series of prototype Android smartphones. The venture involves a 5-inch phone built by Google that features sensors for tracking the device’s movement while modeling the surrounding space and landscape. “This could impact future game design, virtual set construction and location scouting,” notes ETC staffer Phil Lelyveld. “It could enable crowd-sourced 3D stitching of location modeling data.” Continue reading Project Tango: Google Aims to Bring 3D Sensor Tech to Phones

Intel, Broadcom and other chip makers are working to develop the next major advance in cellular network technology, a topic of discussion this week at the Mobile World Congress in Barcelona. Spurred by market leader Qualcomm and rivals that hope to reduce its dominance, companies are demonstrating extensions of the tech presently used in the fastest networks. One new service, LTE Advanced, is expected to offer download rates of 300 megabits per second, four to six times faster than current LTE. Continue reading Chip Makers Race to Deliver Tech for Faster Mobile Networks

In another deal involving “deep learning,” Google is purchasing London-based DeepMind Technologies, a somewhat secretive artifical intelligence startup. The move is viewed as a talent acquisition to bring CEO Dennis Hassabis to Google. The games prodigy and neuroscientist was named “probably the best games player in history” by the Mind Sports Olympiad. While it is unclear what DeepMind does exactly, its website describes building algorithms for games, e-commerce and simulations. Continue reading Deep Learning: Google Plans to Acquire AI Startup DeepMind

Apple is reportedly looking into ways to expand into a mobile-payments service that would go beyond the iTunes store. Down the line, it might compete directly with the likes of Google, eBay’s PayPal and Square, which have become ubiquitous in mobile payment processing for physical goods and services. For Apple, that would likely mean consumers would be able to use their iPhones or iPads to make in-store and online purchases with greater ease. Continue reading Apple Could Make Shopping Easier with Mobile Payments Service

Video processing company Folded Space announced its deep color content encoding (DCE) and decoding algorithms, which are now available for new HD and Ultra HD displays. The algorithms allow these newest displays to show off more vibrant colors captured by HDR cameras. DCE is a much more efficient process than other similar methods that use a lot of bandwidth, and Folded Space intends to license the encoding algorithm to software partners. Continue reading Folded Space Unveils Deep Color Content Encoding for UHD

Yahoo announced it has acquired SkyPhrase, a four-person startup that develops natural language processing technology. SkyPhrase will join the Yahoo Labs team in New York and will help with Yahoo’s efforts in content delivery for mobile users, who are becoming more accustomed to devices that respond to conversational queries. SkyPhrase’s NLP technology is expected to be used for fantasy sports and other areas that rely on queries involving complex data sets. Continue reading Yahoo Picks Up Natural Language Processing Startup SkyPhrase

A team from Disney Research Zürich led by Wojciech Jarosz has developed a new rendering technique that is expected to save animators significant time in the editing process of feature films. The new computational algorithm, which is being presented at this week’s ACM SIGGRAPH Asia conference in Hong Kong, can reduce rendering times for scenes that are affected by light — like those with water and smoke — by a factor of up to 1,000. Continue reading New Animation Technology to Save Production Time and Cost