Welcome to the April 10, 2019 edition of ACM TechNews, providing timely information for IT professionals three times a week.

ACM TechNews mobile apps are available for Android phones and tablets (click here) and for iPhones (click here) and iPads (click here).

To view "Headlines At A Glance," hit the link labeled "Click here to view this online" found at the top of the page in the html version. The online version now has a button at the top labeled "Show Headlines."

A photo of Mendel Rosenblum Inaugural ACM Chuck Thacker Breakthrough Award Recognizes Fundamental Contributions That Enabled Cloud Computing
Association for Computing Machinery
Jim Ormond
April 10, 2019


ACM has named Stanford University's Mendel Rosenblum to receive the inaugural ACM Charles P. "Chuck" Thacker Breakthrough in Computing Award. Rosenblum is recognized for reinventing the virtual machine for the current era, revolutionizing datacenters and facilitating modern cloud computing. In the late 1990s, Rosenblum and students at Stanford revived virtual machines, using them to solve technical problems in building system software for scalable multiprocessors. In 1998, Rosenblum and colleagues founded VMware, which mainstreamed the use of virtual machines to support different software environments to share processor resources within a data center, enabling modern cloud computing. Said ACM president Cherri M. Pancake, "Cloud computing, as we know it today, would not be possible without Rosenblum's reinvention of virtual machines." Rosenblum will receive the award in June at ACM's annual Awards Banquet in San Francisco.

Full Article
New Technique Cuts AI Training Time by More Than 60%
NC State News
Matt Shipman
April 8, 2019


North Carolina State University (NC State) researchers have developed a method that substantially reduces training time for deep learning networks without losing accuracy, expediting development of new artificial intelligence (AI) applications. According to NC State's Xipeng Shen, the Adaptive Deep Reuse technique has demonstrated a maximum AI training time reduction of 69%. The method is founded on the realization that many chunks of data in a dataset are similar to each other, which enables a deep learning network to save power by applying filters to one data chunk, then applying the results to all similar chunks in the same set. Said Shen, "We think Adaptive Deep Reuse is a valuable tool, and look forward to working with industry and research partners to demonstrate how it can be used to advance AI."

Full Article
Fake Cancerous Nodes in CT Scans, Created by Malware, Trick Radiologists
The Washington Post
Kim Zetter
April 3, 2019


Researchers in Israel say they have developed malware that tricks medical scanners into detecting bogus cancerous nodes, pointing out serious security vulnerabilities in critical medical imaging equipment and the networks that transmit those images. The researchers at Ben-Gurion University Cyber Security Research Center used machine learning to train their code to rapidly assess scans passing through a picture archiving and communication system (PACS) network, and adjust and scale false tumors to conform to a patient's anatomy and dimensions. The malware would let hackers automatically add malignant-seeming masses to scans before clinicians examine them, or eliminate real cancerous growths without detection. Malefactors could opt to modify random scans to sow chaos and mistrust in hospital equipment, or could target specific patients, looking for scans with a specific individual's name or identification number. Aiding attackers is the fact that PACS networks are usually not encrypted.

Full Article
*May Require Paid Registration

Smart windows that have darkened as the sun became brighter As Office Tenants Expect More Tech, Even the Windows Get Smart
The New York Times
Jane Margolies
April 9, 2019


Demand for new technologies from office developers and tenants is fueling competition between suppliers, with upgrades like smart windows, or fiber-optic cable to improve redundancy, attracting attention from investors. For example, developer DivcoWest is installing an estimated 10 miles of fiber-optic cable beneath and within six office buildings in a mixed-use development in Cambridge, MA, promising Wi-Fi with uninterrupted connectivity. According to New York consultant WiredScore, more than 1,800 office buildings worldwide have so far registered for or been accorded certification for in-building Internet connectivity. Some office providers are enhancing other kinds of technology in buildings, like installing antennae for picking up and amplifying cellphone signals and lighting sensors that track the brightness of the sun. Even more sophisticated capabilities are expected, with Rao Mulpuri, CEO of smart window maker View, predicting windows would eventually be used like computer screens.

Full Article
Programming Language Popularity: C++ Bounces Back at Python's Expense
ZDNet
Liam Tung
April 8, 2019


C++ had a share of 8.83% in Tiobe's April 2019 programming community index of the popularity of programming languages, up 1.62 percentage points from April 2018 and moving the 35-year-old programming language to third place on the list. The ratings are based on results in Google, Bing, Yahoo, Wikipedia, YouTube, and Baidu. The index is meant to indicate a language's popularity among developers, but does not take into account which is the best programming language or the language that has the most lines of code written in it. This month, C++ overtook the increasingly popular Python, which has seen its usage grow massively among developers in the past few years. Python this month had a 8.166% share, up 2.36 percentage points over the same month of last year. The top 10 languages in this month’s Tiobe index are, in descending order, Java, C, C++, Python, Visual Basic .NET, C#, JavaScript, SQL, PHP, and Assembly Language.

Full Article

A picture of Blue, a new robot arm Blue Is a New Low-Cost Force-Controlled Robot Arm From UC Berkeley
IEEE Spectrum
Evan Ackerman
April 9, 2019


University of California, Berkeley (UC Berkeley) roboticists have designed a low-cost robot arm for manipulation tasks in unconstrained human environments. Blue is a seven-degrees-of-freedom appendage expected to cost less than $5,000 when mass-produced, built around quasi-direct drive actuators, which support robust force control and selectable impedance, and are highly backdrivable. Blue also features brushless motors to drive a 7.125:1 timing belt, paired with a differential, as well as custom motor driver boards, each outfitted with 14-bit absolute magnetic encoder for motor commutation and robot position sensing. The encoder also has 12-bit current sensing for closed-loop current/torque control of each servomotor; a three-axis accelerometer for state estimation, collision detection, and calibration, and temperature sensors for thermal monitoring.

Full Article
Companies Take a Piecemeal Approach to Automation Tech
The Wall Street Journal
Angus Loten
April 4, 2019


KPMG surveyed about 600 C-suite executives to determine how companies are approaching the deployment of automation technologies. About 30% of respondents said their companies have apportioned $50 million or more into smart automation projects, and more than 50% have already spent at least $10 million. Such projects include diverse combinations of robotic process automation, artificial intelligence, machine learning, cognitive computing, and analytics; thus far, funding is going into corporate finance and accounting functions, followed by group benefits strategies, compliance, and industry-specific core operations. More than 50% of respondents listed improving or streamlining customer services and front-office effectiveness as their primary goal. Many companies have adopted a piecemeal strategy to automation, due to uncertainty about how much investment will be needed to make deployments worthwhile, as well as a dearth of "organizational clarity and accountability."

Full Article
*May Require Paid Registration

A parking lot of predominantly new Tesla Model 3 electric vehicles New Feature Enables Car to Drive Itself Through a Parking Lot to Pick You Up
CNBC
Tom Huddleston Jr.
April 9, 2019


Tesla has unveiled a new feature allowing drivers to remotely call their Tesla car to drive itself through a parking lot to pick them up, provided the vehicle is within 150 feet. Tesla CEO Elon Musk said the Enhanced Summon feature will be available within the week for any U.S. Tesla owners who opted for the Enhanced Autopilot, or "Full Self-Driving," package. With Enhanced Summon, drivers can tap a button in the Tesla mobile app to summon their car. They can either choose the "Find Me" option to bring the vehicle directly to them using global-positioning systems, or drop a pin at a location within 150 feet of where the Tesla is parked. Enhanced Summon builds off of the Summon feature in Tesla's Autopilot software, which lets users remotely park their auto using the mobile app.

Full Article

Dancing in Class. Through Hip-Hop and Dance, This Nonprofit Teaches Girls of Color How to Code
Philadelphia Inquirer
Heather Khalifa
April 8, 2019


The danceLogic program is offered by Philadelphia’s West Park Cultural Center, a nonprofit that aims to provide affordable activities for children over the summer and after school. The weekly classes include one hour of dance instruction and practice, followed by an hour of coding in an effort to promote science, technology, engineering, the arts, and math (STEAM). The idea for danceLogic came to founder Betty Lindley in 2001 after she read an article about a teacher in New York who brought her background of technology and dance together. Lindley knew Franklyn Athias of Comcast Cable spent his Saturday mornings teaching free coding classes, and she approached him about partnering for danceLogic. Said Athias, "My view is if you can train them younger, we can get them into this program and one day they either become programmers or something in the technology field."

Full Article
A Powerful Spyware App Now Targets iPhone Owners
TechCrunch
Zack Whittaker
April 8, 2019


Researchers at mobile security firm Lookout have discovered a powerful spy app originally designed for Android devices can now target iPhone owners. The investigators said the app's developer, Italy’s Connexxa, abused Apple-issued enterprise certificates to circumvent the Android app store to infect unsuspecting victims. Once installed, the app can silently steal a target device’s information, including real-time location data, and can be remotely activated to eavesdrop on people's conversations. Connexxa also created Exodus, an Android spy app used by Italian authorities. Lookout's Adam Bauer said both apps employ the same backend infrastructure, while the iOS app used multiple methods, including certificate pinning, to thwart analysis of network traffic. Disclosure of these findings prompted Apple to revoke the app maker's enterprise certificate.

Full Article

Two people with virtual reality gears on China's Virtual Reality Arcades Aim for Real-World Success
South China Morning Post
April 7, 2019


China was home to an estimated 3,000 virtual reality (VR) arcades in 2016, when the market was forecast to grow 13-fold between then and 2021 to $782 million, according to a joint report by iResearch Consulting Group and Greenlight Insights. This market potential is why the VR and augmented reality (AR) industries are excited by China. A 2018 report from Digi-Capital said that Chinese growth in the next five years could see it dominate AR/VR long-term, and not by a small margin. A key factor in China's lead in the industry is its government. Beijing is pushing hard for China to become a world leader in next-generation technologies including artificial intelligence and autonomous vehicles. VR has been added into that group, benefiting from a range of preferential government policies. Several towns and cities in China have already declared themselves as incubator zones that are integrating VR into research, manufacturing, education, and other sectors.

Full Article
Europe Making AI Rules Now to Avoid New Tech Crisis
CNN
Ivana Kottasova
April 8, 2019


The European Commission (EC) this week introduced ethics guidelines to inform the development of artificial intelligence (AI). IBM's Liam Benham said, "It's like putting the foundations in before you build a house...now is the time to do it." The EC announcement included seven nonbinding principles for guiding AI development and building trust. European digital economy official Mariya Gabriel emphasized the need for companies using AI to be transparent, as well as non-discriminatory. The EC said future AI systems must be safe and reliable across their entire life cycle, while data protection must be prioritized, with users controlling their own information. The nonprofit AlgorithmWatch organization raises issues with Europe's approach because the term "trustworthy AI" is not well defined, while Thomas Metzinger at the University of Mainz in Germany criticized the guidelines for not banning AI weaponization.

Full Article
When Computers Collude
NPR Online
Greg Rosalsky
April 2, 2019


A team of Italian economists built algorithms and set them up to compete in selling imaginary goods in a virtual marketplace, and found the programs appeared to collude to elevate prices. This has raised issues about the regulation of pricing algorithms, which are currently used in various industries. The University of Bologna's Emilio Calvano cited several views of this issue, including one that regulation should be avoided because the market is ultimately self-correcting. Another school of thought urges the creation of a new regulator that tests algorithms in government computers before they are allowed in the marketplace. Calvino said he supports a third option, in which companies would be permitted to innovate and deploy such algorithms in marketplaces, "because they can do a lot of good." The government's antitrust authorities should then monitor them, Calvano said, and intercede if they are being used anti-competitively, which might require updating antitrust statutes.

Full Article
Sao Paulo School of Advanced Science on Learning from Data
 
ACM Learning Center
 

Association for Computing Machinery

2 Penn Plaza, Suite 701
New York, NY 10121-0701
1-800-342-6626
(U.S./Canada)



ACM Media Sales

If you are interested in advertising in ACM TechNews or other ACM publications, please contact ACM Media Sales or (212) 626-0686, or visit ACM Media for more information.

To submit feedback about ACM TechNews, contact: [email protected]