Welcome to the October 18, 2017 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."
|
|
How Google's Quantum Computer Could Change the World
The Wall Street Journal Jack Nicas October 16, 2017
A team of Google scientists is using the physics of quantum mechanics to build a reliable, large-scale computer of potentially staggering power, with the ability to radically change industries from artificial intelligence to chemistry, accelerating machine learning and engineering new materials, chemicals, and drugs. "It's a fundamentally new way of harnessing nature to do computations," says the University of Texas at Austin's Scott Aaronson. Google's quantum system in 2018 will be put to the test by attempting to solve a computational problem that a classical computer would take billions of years to crack, thus marking the "quantum supremacy" milestone. Google researcher Hartmut Neven's team at the company's Santa Barbara lab is in a race to reach this milestone by completing a transistor-less 49-quantum bit (qubit) chip. The quantum system is contained in a cryostat vat that keeps magnetic fields from interfering, and the qubits are extremely chilled to sustain their conductivity for running calculations.
*May Require Paid Subscription
|
Liquid Metal Brings Soft Robotics a Step Closer
University of Sussex (United Kingdom) James Hakner October 17, 2017
Researchers at the University of Sussex and Swansea University in the U.K. say they have applied electrical charges to manipulate liquid metal into two-dimensional shapes such as letters and a heart. The team notes the results mark an "extremely promising" new class of materials that can be programmed to seamlessly change shape, a breakthrough they say could lead to new possibilities in soft robotics and shape-changing displays. "This is a new class of programmable materials in a liquid state which can dynamically transform from a simple droplet shape to many other complex geometry in a controllable manner," says University of Sussex researcher Yutaka Tokuda. The team says the electric fields used to shape the liquid are created by a computer, meaning the position and shape of the liquid metal can be programmed and controlled dynamically. The research was presented this week at the ACM Interactive Surfaces and Spaces (ISS 2017) conference in Brighton, U.K.
|
Serious Flaw in WPA2 Protocol Lets Attackers Intercept Passwords and Much More
Ars Technica Dan Goodin October 16, 2017
Researchers at the University of Leuven (KU Leuven) in Belgium have discovered a severe flaw in the WPA2 protocol that enables hackers within range of a vulnerable device or access point to intercept passwords and other sensitive data presumed to be shielded by the Wi-Fi encryption protocol. The researchers say the Key Reinstallation Attack (KRACK) exploit targets the core WPA2 protocol itself and can be waged against devices running Android, Linux, and OpenBSD. KRACK attackers can deceive such devices to reinstall an all-zero encryption key instead of the actual key, forcing the client to reset packet numbers containing a cryptographic nonce and other parameters to their initial values. This causes the nonce to be reused in a manner that permits circumvention of encryption. KU Leuven's Mathy Vanhoef warns the flaw also can be used to infect websites with ransomware or other malware, and the vulnerability will likely be most threatening to large corporate and government Wi-Fi networks.
|
Berkeley Experts on How to Build More Secure, Faster AI Systems
Berkeley News Brett Israel October 16, 2017
Experts at the University of California, Berkeley's Real-Time Intelligent Secure Execution Lab (RISELab) have released a report outlining new research pathways to address issues that could hinder the progress of artificial intelligence (AI). RISELab director Ion Stoica says many of the issues stem from systems, security, and computer architecture. The experts cite security challenges inherent in many companies deploying AI applications in the public cloud on servers they do not control. RISELab's Raluca Ada Popa says securing AI systems could involve leveraging secure multiparty computation to manage sensitive data via encryption, or keeping critical AI system components in a secure execution environment. Adding robustness against adversarial learning attacks is another security measure AI needs, and new machine-learning models and network architectures will likely have to be designed, says RISELab's Dawn Song. Meanwhile, RISELab member David Patterson, a past president of ACM, notes accommodating the exponential growth of device-generated data could be achieved by designing secure, AI-specific computers that are optimized for a given task.
|
Researchers Create Digital Objects From Incomplete 3D Data
Saarland University October 13, 2017
Researchers at Saarland University and the Max Planck Institute for Informatics in Germany say they have developed a computational method for reconstructing a digital object from incomplete images. Their method uses a neural network to reconstruct digital objects from incomplete datasets. "Our method requires no supervision during the learning phase, which is novel for this type," notes Max Planck researcher Mario Fritz. For example, the researchers could use the new model to reconstruct a flat monitor, whose digital representation after a three-dimensional scan looked more like a paneled wall, so viewers could once again recognize a monitor in the digital object. The team intends to further develop the method so it also will work on deformable objects and larger scenes. "In the future, it will have to be possible to capture real-world objects simply and quickly, and project them in a realistic way into the digital world," says Saarland University professor Philipp Slusallek.
|
University Robots to Improve the Care of Older People, Patients and Autistic Children
Lancashire Post Iain Lynn October 16, 2017
Researchers at Edge Hill University in the U.K. are developing robots designed to help monitor and care for the elderly, patients, and autistic children. The researchers have spent the past several months training a toddler-sized robot, called Robbie, to recognize 90 common objects as well as human actions and emotions. Robbie uses large-scale annotated image databases to learn about the world and then recognize objects and actions, guess the gender and estimate the age of a person, and determine what emotions that person is feeling. The researchers say this type of robotic system could become a companion to both children and older adults. "Initially we see Robbie being most useful in residential care homes where he can be a companion to residents and can keep an eye on them, watching and recording what they eat, drink, if they take their medication, their emotions, and more," says Edge Hill University researcher Ardhendu Behera.
|
Study Finds Auto-Fix Tool Gets More Programmers to Upgrade Code
NC State News Matt Shipman October 16, 2017
Researchers at North Carolina State University (NCSU) have found that auto-fix tools are effective ways to get programmers to make relevant upgrades. Most programmers rely on code in external libraries to perform some of their functions, and these libraries are periodically updated to address flaws, but many coders put off these updates. "Our goal with this project was to assess tools designed to get more programmers to upgrade their out-of-date dependencies," says NCSU professor Chris Parnin. The researchers examined thousands of open source projects on GitHub, focusing on what methods were used to incentivize or facilitate upgrades and whether those incentives made any difference. The researchers found projects with automated pull requests made 60 percent more of the necessary upgrades than projects that did not use incentives. The team will present its research at the IEEE/ACM International Conference on Automated Software Engineering (ASE 2017), which takes place Oct. 30-Nov.3 at the University of Illinois at Urbana-Champaign, IL.
|
Now There's an IQ Test for Siri and Friends
Technology Review October 13, 2017
Researchers at the Chinese Academy of Sciences in Beijing have developed an intelligence test that both machines and humans can take, and used it to rank intelligent assistants such as Google Assistant and Siri on the same scale used for humans. The test is based on the "standard intelligence model," in which systems must have a way of obtaining data from the outside world. They must be able to transform the data into a form they can process, they must be able to use this knowledge in an innovative way, and they must feed the resulting knowledge back into the outside world. The researchers found that even a six-year-old human outperforms the most advanced digital assistant, which according to this test is Google Assistant. However, machine intelligence is rapidly improving. In 2014, Google Assistant scored 26.4 on this test, compared with a score of 47.28 in the most recent test.
|
Using Facebook Data as a Real-Time Census
UW News (WA) Kim Eckart October 12, 2017
Researchers at the University of Washington (UW) recently conducted a study demonstrating how Facebook can serve as a more current source of information, especially about migrants, when it comes to conducting a census. The researchers say the study marks the first demonstration of how present-day migration statistics can be obtained by compiling the same data advertisers use to target their audience on Facebook, and combining it with information from the U.S. Census Bureau. In addition, UW professor Emilio Zagheni says as researchers further explore the increasing number of databases produced for advertisers, social scientists could leverage social networks to collect information on geography, mobility, behavior, and employment. The study focused on Facebook's Ads Manager service, which enables users to enter information on a target audience and then receive data on that population. The UW team developed a program for extracting data from Ads Manager about expatriates from more than 50 countries to every U.S. state.
|
Scientists Develop Machine-Learning Method to Predict the Behavior of Molecules
New York University James Devitt October 12, 2017
Researchers at New York University (NYU) are leading an international team that has developed a machine-learning method that predicts molecular behavior. "By identifying patterns in molecular behavior, the learning algorithm or 'machine' we created builds a knowledge base about atomic interactions within a molecule and then draws on that information to predict new phenomena," says NYU professor Mark Tuckerman. He notes the system can learn complex interatomic interactions, which are normally prescribed by complex quantum mechanical calculations, without performing such intricate calculations. The team created the machine by developing a small sample set of the molecule to be studied to train the algorithm. They then used the system to simulate complex chemical behavior within the molecule. "We have reached the ability to not only use (artificial intelligence) to learn from data, but we can probe the AI model to further our scientific understanding and gain new insights," says Technical University of Berlin professor Klaus-Robert Muller in Germany.
|
New Software Speeds Origami Structure Designs
Georgia Tech Research Horizons Josh Brown October 11, 2017
Researchers at the Georgia Institute of Technology (Georgia Tech) have developed MERLIN, a computer-aided approach for streamlining the design process for origami-based structures. The researchers say MERLIN is a breakthrough that makes it easier for engineers and scientists to conceptualize new ideas graphically while generating the underlying mathematical data for building the structure. "With the new software, we can easily visualize and, most importantly, engineer the behavior of deployable, self-assembling, and adaptable origami systems," says Georgia Tech professor Glaucio Paulino. The research involved building a computer model to simulate the interaction between the two facets of a folded sheet--how easily and how far the folds would bend, and how much the flat planes would deform during movement. MERLIN lets users simulate how origami structures will respond to compression forces at different angles. "The software also allows us to see where the energy is stored in the structure and better understand and predict how the objects will bend, twist, and snap," Paulino says.
|
USC ISI Leads IARPA Contract for Developing Hybrid Forecasting Systems
USC Viterbi School of Engineering Caitlin Dawson October 11, 2017
Researchers at the University of Southern California's (USC) Viterbi School of Engineering Information Sciences Institute have received a U.S. Intelligence Advanced Research Projects Activity grant to develop a human-machine hybrid forecasting system for worldwide socioeconomic and geopolitical issues. The Synergistic Anticipation of Geopolitical Events team will create a scalable platform that incorporates human insights and machine-learning models. This human-machine system will enable data-driven platforms to incorporate human feedback in real time for emerging issues that lack historical precedent. The researchers say they will use an interactive platform to efficiently search and organize information, including traditional news reporting, social media, and financial indicators. The platform also will enable forecasters to interact with machine models by changing the model-generated forecasts. "By the end of the program, our hybrid forecasting system needs to be capable of handling at least 500 individual forecasting problems a year, which is significantly more than was processed in the previous program," says USC Viterbi's Fred Morstatter.
|
There's a Huge Opportunity in Robotics for Early-Career Computer Scientists and Serious Software Engineers
ZDNet Greg Nichols October 17, 2017
In an interview, University of Washington professor Maya Cakmak discusses the role of programming by demonstration (PbD) in her work on human-machine interaction, and how it could help enterprise robotics make market gains by appealing to early-career computer scientists and software engineers. Cakmak notes with PbD, "you demonstrate a task and the robot figures out what the program should be to recreate what you demonstrated." Using an autonomous mobile robot called Fetch, Cakmak says her students are developing unique ways for non-experts to program the robot to perform specific jobs. "What we're trying to do is let a person program manipulator actions by demonstration, and then define those," she notes. Cakmak envisions PbD enabling previously unconsidered use cases for robots. "People who see problems can figure out how to use robots to solve them," she says. "It will empower those people to program robots for themselves."
|
|