Welcome to the November 2, 2016 edition of ACM TechNews, providing timely information for IT professionals three times a week.
Updated versions of the ACM TechNews mobile apps are available for Android phones and tablets (click here) and for iPhones (click here) and iPads (click here).
HEADLINES AT A GLANCE
A Billion Billion Calculations per Second: Where No Computer Has Gone Before
South China Morning Post (Hong Kong) (10/29/16) Viola Zhou
China has launched the development of its first exascale high-performance computer to maintain its lead position in the global supercomputing race. The system will run at 1,000 petaflops, topping the speed of China's Sunway TaihuLight computer by a factor of 10. China's Ministry of Science and Technology has allocated funding to three research institutions to devise prototypes to meet its five-year target of putting an exascale computer into operation. The participating institutions include Sugon, the National University of Defense Technology, and the National Research Center of Parallel Computer Engineering and Technology (developer of Sunway TaihuLight, currently the top supercomputer in the world). University of Science and Technology of China professor An Hong says once the prototypes are finished, the ministry will choose two teams with the best designs to construct the fully functioning exascale system. For the first time this year, China dethroned the U.S. as the country with the most supercomputers in the Top500 ranking. "The demand for computing speed has no limits," An says. "Now we have the money and technology, we can build better computers for scientists to use."
Lancaster University Hails Potential of AI Software to Cut Data Center Power Consumption
ComputerWeekly.com (11/01/16) Caroline Donnelly
Researchers from Lancaster University in the U.K. are developing REx, an artificial intelligence (AI)-based program designed to reduce the amount of energy consumed by data centers. REx consists of software components that can rapidly reassemble themselves into the most efficient form without the need for human intervention. The software uses a technique called microvariation, which takes software components like memory caches and various search and sort algorithms, and then automatically assembles the software components depending on the type of workloads the server is being used to run. "For the first time we have computer programs that are able to gain a deep understanding of their behavior and performance while they do their job," says Lancaster University lecturer Barry Porter. The researchers decided to focus on data centers because they are highly dynamic environments that experience major fluctuations in workloads over the course of the data. "By having the software that powers data centers built from this self-assembling software that learns its own behavior as it's running, we have the ability to deal with these dynamics much more efficiently and therefore save energy in the long term," Porter says. He notes REx also could help organizations cut software maintenance costs.
New Research Center to Explore Ethics of Artificial Intelligence
The New York Times (11/01/16) John Markoff
Carnegie Mellon University (CMU) on Wednesday will announce the launch of a new research center for investigating the ethics of artificial intelligence (AI). CMU president Subra Suresh says AI cannot advance without ethical discussions, noting "we are at a unique point in time where the technology is far ahead of society's ability to restrain it." Suresh also cites excessive optimism over certain people's claims about AI's sophistication, especially as it pertains to autonomous vehicles. The K&L Gates Endowment for Ethics and Computational Technologies will draw staff from multiple academic disciplines and initially add two faculty and three positions for graduate students. It also will set up a biennial conference on ethical issues facing the AI field. The K&L Gates law firm is the primary underwriter of the center with a $10-million endowment. K&L Gates chairman Peter J. Kalis says the potential economic and cultural impact of AI makes it vital that society makes ethical choices about its use. "Carnegie Mellon resides at the intersection of many disciplines," Kalis says. "It will take a synthesis of the best thinking of all of these disciplines for society to define the ethical constraints on the emerging AI technologies."
China's Policing Robot: Cattle Prod Meets Supercomputer
Computerworld (10/31/16) Patrick Thibodeau
Chinese researchers have developed AnBot, an "intelligent security robot" deployed in a Shenzhen airport. The backend of AnBot is linked to China's Tianhe-2 supercomputer, where it has access to cloud services. AnBot uses these technologies to conduct patrols, recognize threats, and identify people with multiple cameras and facial recognition. The cloud services give the robots petascale processing power, well beyond the processing capabilities in the robot itself. The supercomputer connection enhances the intelligent learning capabilities and human-machine interface of the devices, according to a U.S.-China Economic and Security Review report that focuses on China's autonomous systems development efforts. The report found the ability of robotics to improve depends on the linking of artificial intelligence (AI), data science, and computing technologies. In addition, the report notes simultaneous development of high-performance computing systems and robotic mechanical manipulation give AI the potential to unleash smarter robotic devices that are capable of learning as well as integrating inputs from large databases. The report says the U.S. government should increase its own efforts in developing manufacturing technology in critical areas, as well as monitoring China's growing investments in robotics and AI companies in the U.S.
Researchers Uncover Hidden Censorship on Chinese Live Streaming Apps
U of T News (11/01/16) Dena Allen
University of Toronto (U of T) researchers have found hidden keyword blacklists that are used to censor chats on three popular Chinese live-streaming applications. The researchers also found uneven implementation of censorship on the live-streaming apps they studied, a result that goes against prior research and assumptions that Internet censorship in China operates under a uniform set of guidelines. "Social media companies in China are held responsible and liable for content on their platforms, and are expected to control content, or face punishment from the government," says U of T researcher Masashi Crete-Nishihata. "Our research shows how this system works in practice." The researchers studied how censorship works on these applications by reverse-engineering them, which led them to find censorship is executed on the client side, meaning all the rules to perform censorship are inside of the application running on the users' phones and computers. In addition, the researchers were able to collect the keyword lists used to trigger censorship of chat messages. "These keyword lists give a behind the scenes look into how social media is censored in China," says U of T researcher Jeffery Knockel. The researchers tracked updates to the keyword lists over a year and found new terms were often added in reaction to sensitive events.
U.S. Convenes Chip Study Group
EE Times (10/31/16) Rick Merritt
President Barack Obama has convened a new working group under the President's Council of Advisers on Science and Technology to examine ways to fortify the U.S. semiconductor industry as competition from China intensifies and the cost and complexity of upholding Moore's Law increases. The Semiconductor Industry Association (SIA) recommended establishing such a task force on policy for the chip sector, notes SIA chief John Neuffer. "It's very important for the U.S. government to signal the semiconductor industry is important to the U.S economy and national security," he says. The SIA has long pressured the federal government to boost spending on semiconductor research and reduce U.S. statutory tax rates that top those of competing countries. It also has pushed for more visas for foreign students receiving advanced science, technology, engineering, and math degrees in the U.S. and for broadening free trade. Independently, the association has begun a study outlining the chip industry's vision statement. The White House indicates it will urge "additional public and private investments in [research and development]" as part of a "set of recommendations on initial actions the federal government, industry, and academia could pursue to maintain U.S. leadership in this crucial domain."
Researchers Bring Eyewear-Free 3D Capabilities to Small Screen
Researchers from Seoul National University (SNU) in South Korea have developed a new method of making a convertible video display that offers both two- and three-dimensional imaging. In addition, the researchers say the new method has near-viewing capabilities, and simplifies and shrinks the architecture of the technology. They say the method is based on a monolithic structure that combines the active parallax barrier, a polarizing sheet, and an image layer into a single panel. Instead of two separate image and barrier panels, as is the case with conventional convertible video displays, the researchers used a polarizing interlayer with the image layer in direct contact with one side of the interlayer, while the active parallax barrier of a liquid-crystal layer is formed on the other side as an array of periodically patterned indium-tin-oxide electrodes. This interlayer enables the minimum separation of the image and barrier layers, which provides the short viewing distance required for the smaller screens of mobile devices. "The polarizing interlayer approach here will allow high resolution together with design flexibility of the displays, and will be applicable for fabricating other types of displays such as viewing-angle switchable devices," says SNU professor Sin-Doo Lee.
The Kids Who Might Save the Internet
Christian Science Monitor (10/31/16) Sara Sorcher
The future of the Internet may lie in the hands of a new generation of computer-savvy children forming a hacker community disposed toward fortifying Web security. "The next generation...are more willing to challenge assumptions in technology than older people, who may feel things are established or difficult to change," says ESET researcher Stephen Cobb. "It's the idealism of youth which may inspire alternative approaches to design and deployment of digital technology." One teenage hacker, "CyFi," hacked a game app at age 10 and has co-founded a cybersecurity conference for children that hosts ethical hacking workshops. She says her generation "has a responsibility to make the Internet safer and better," especially as its increasing connectedness to homes and institutions creates new vulnerabilities. Among the incentives luring children into hacking are financial rewards for the discovery of digital weaknesses. Corporate funding for training also is encouraging young people to fill a void in the cybersecurity workforce, while the Air Force Association's CyberPatriot contest seeks to test the skills of middle- and high-school youth and inspire them to pursue cybersecurity careers. Some children want to pass on their technical knowledge to other children, with one 10-year-old heading a nonprofit that produces videos and games to educate children in cybersecurity.
Wearable Health: Exploring Human-Centered Solutions of On-Body Technologies to Improve Healthcare
CCC Blog (10/31/16) Helen Wright
George Mason University professor Vivian Motti proposes the concept of noninvasive, wearable technologies to improve healthcare as well as education by designing improved interfaces at the university's Human Centric Design Lab. "A more comprehensive understanding about the users' perspectives on wearable technologies facilitates the identification, classification, and analysis of the benefits and drawbacks of current solutions, helping in the definition of new design guidelines and patterns to improve the development of next-generation technologies," Motti says. "We expect to improve the interfaces and interactive solutions of wearables by iteratively involving human users in the design process and by understanding how different contexts of use impact the overall user experience with emerging technologies." Among the applications Motti's team is investigating is the use of a smartwatch app to support learning-disabled students so they can be more autonomous. "We also investigate privacy aspects in the context of mobile and wearable applications, seeking to understand better the threats, risks, and implications involved in these interaction scenarios, to devise solutions that are easier for users to understand and use," Motti notes. "Our research on usable privacy focuses on small devices, such as smartphones and watches, and graphic user interfaces."
Google Teaches 'AIs' to Invent Their Own Crypto and Avoid Eavesdropping
Ars Technica (10/28/16) Sebastian Anthony
Google Brain researchers say they have developed artificial intelligence (AI) technology that can form encryptions and communicate securely. The researchers found by teaching an AI to value secrecy, neural networks can learn to protect their communications without being prescribed a set of cryptographic algorithms. Two neural networks, Alice and Bob, were given a shared secret key, and a separate network, Eve, was tasked with decrypting the communication between Alice and Bob. All three robots had the same neural network architecture but had no connection other than Alice and Bob's shared key. The key and plaintext were entered into the first layer of Alice's neural network, while the key and Alice's ciphertext output were input for Bob; Eve received just the ciphertext. Each network was provided a loss function, and Eve and Bob's plaintext had to be as close to the input plaintext as possible, while Alice's loss depended on Eve's accuracy, creating an adversarial generative network among the robots. Alice and Bob created a system in which they could communicate with few errors, but although Eve showed some improved degree of accuracy, Alice and Bob responded by improving their cryptography technique. Researchers say future work might focus on steganography, as well as asymmetric (public-key) encryption.
Study: People Can Tell If They Are Voting on a Secure System
Rice University (10/24/16) Amy McCaig
Despite election rhetoric aiming to cast doubt on the security of the voting system, U.S. voters have a good sense of whether or not a voting system is secure, according to a Rice University study. Researchers assessed 90 participants' perceptions of three voting systems in a mock election. Each system was armed with a different level of security: a standard paper ballot, a paper ballot with enhanced security features, and a paper ballot with fake security features that did nothing to make the system safer. After voting, participants were asked questions about the security of the voting system they used. Voters found the system with enhanced security features to be the most secure and were not fooled by the system with fake security mechanisms. Participants also were asked about their level of confidence in the voting systems. Half of the voters who used the enhanced security system felt more confident their votes would be counted accurately, whereas just 16.7 percent of voters using the other two systems were confident. "If voters suspect any type of security flaws, then they might not see the point in participating in an election," says Rice researcher Claudia Ziegler Acemyan. "This results in disenfranchisement, potentially impacted election results, and possibly an overall lack of confidence in the resulting government and policies."
Making Computers Explain Themselves
MIT News (10/27/16) Larry Hardesty
Researchers from the Massachusetts Institute of Technology's (MIT) Computer Science and Artificial Intelligence Laboratory (CSAIL) have developed a method for training neural networks so they provide not only predictions and classifications, but also rationales for their decisions. "In real-world applications, sometimes people really want to know why the model makes the predictions it does," says MIT graduate student Tao Lei. The neural nets are trained on textual data, with the CSAIL researchers splitting the net into two modules. The first module filters out segments of text from the training data, which are scored based on length and coherence; higher-scoring segments are shorter and are extracted more from strings of consecutive words. The selected segments are then handed to the second module, which conducts prediction or classification. The models undergo concurrent training to maximize both the score of the extracted segments and the accuracy of prediction/classification. One dataset on which the researchers tested their method was a set of reviews from a beer-assessing website. Tests showed the system's concurrence with human annotations was 96 percent and 95 percent, respectively, for ratings of appearance and aroma, and 80 percent for palate.
Paralyzed People Inhabit Distant Robot Bodies With Thought Alone
New Scientist (10/26/16) Helen Thomson
The European Union's VERE project aims to dissolve the boundary between the human body and a surrogate. The system was tested with three volunteers, each of whom wore an electroencephalogram (EEG) cap and a head-mounted display that showed what a robot in Japan was seeing. The volunteers made the robot move by concentrating on arrows superimposed across the display, each flashing at different frequencies. A computer detected which arrow a participant was looking at using the EEG readings that each frequency provoked, and it sent the corresponding movement to the robot. The researchers found this system enabled the participants to control the robot in near-real time, and they were able to make the robot pick up a drink, move across the room, and put the drink on a table. The researchers tried to improve the feeling of embodiment using auditory feedback. As they controlled the robot, both able-bodied volunteers and those with spinal cord injuries were able to place the drink closer to a target location when they heard footsteps as they walked, instead of a beep or no noise at all. The improved control suggests users feel more in tune with the robot itself when there is auditory feedback, says University of Rome researcher Emmanuele Tidoni.
Abstract News © Copyright 2016 INFORMATION, INC.
To submit feedback about ACM TechNews, contact: [email protected]