Welcome to the October 2, 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."

CS student The State of Women in Computer Science
Alison DeNisco
September 29, 2017

Intersecting factors are preventing women from pursuing and persisting in computer science majors. Contributors include a dearth of exposure to computer science and engineering in middle school and high school, while well-meaning educators or parents are steering girls away from tech-focused classes; a general lack of awareness of potential tech careers also plays a key role. The Anita Borg Institute's Elizabeth Ames says many colleges realize the structure of their computer science programs--with entry-level courses assuming students are already well-versed in programming--are partly culpable for women's under-representation. Research also found when a male and a female student enter a computer science course at the same level, the male thinks he is more skilled than he is, and the female thinks she is less skilled than she is. Fewer women on computer science faculty also is discouraging female students from continuing with a major, while some students are not progressing to graduate school because of lucrative technology job offers.

Full Article
'Revolutionary' New Gesture Control Tech Turns Any Object Into a TV Remote
Lancaster University
October 2, 2017

Researchers at Lancaster University in the U.K. have demonstrated a new method of gesture control that can turn everyday objects into TV remotes. Their Matchpoint technology needs only a simple webcam, and it functions by displaying moving targets that spin around a small circular widget in the corner of the screen. The targets are aligned with different functions, such as volume, channel change, or menu views. The researchers say the user syncs up the target's trajectory with their hand, head, or an object to activate the desired command via "spontaneous spatial coupling." The Matchpoint software differs from current gesture-control techniques in that it does not seek a specific body part it has been trained to identify, offering much more flexibility and ease for users. "Our method allows for a much more user-friendly experience where you can change channels without having to put down your drink, or change your position," says Lancaster's Christopher Clarke.

Full Article

energy hog, illustration The Ridiculous Amount of Energy It Takes to Run Bitcoin
IEEE Spectrum
Peter Fairley
September 28, 2017

Keeping bitcoin running via the process of "mining" consumes a staggering amount of electricity, and experts concerned about an energy crisis stemming from slowing chip efficiency gains combined with the cryptocurrency's soaring value are investigating energy-saving solutions. Bitcoin mining's enormous energy expenditure is partly due to the computationally intense process of cryptographic hashing needed to encode each new block of transactions. Blockchain.info recently estimated that the world's bitcoin miners were producing about 5 quintillion 256-bit cryptographic hashes each second. Sebastiaan Deetman at Leiden University in the Netherlands warns that bitcoin power demand could potentially swell 20-fold to 14 gigawatts by 2020 if hash computations continue to accelerate. Remedies being explored by Intel's Michael Reed include a blockchain system that uses built-in security features in central-processing units. Reed notes one piece of Intel software randomly chooses which users will write each block, and idles all users' code for randomly determined intervals.

Full Article
UMIACS Partners with Fraunhofer, Signature Science on DNA Screening Technologies
University of Maryland
Tom Ventsias
September 28, 2017

Researchers at the University of Maryland Institute for Advanced Computer Studies (UMIACS) are developing next-generation DNA sequence-screening tools in collaboration with the Fraunhofer USA Center for Experimental Software Engineering and Signature Science (a subsidiary of the Southwest Research Institute). "The software underlying this project is extremely complex, involving an intricate chain of sophisticated software components," says UMIACS professor Mihai Pop. "This chain has to work seamlessly--not only to reliably identify biological threats, but to do so under strict time and resource constraints." UMIACS' Todd Treangen says his group "will develop software modules that can provide rapid sequence and protein structure comparisons to assess the threat potential of functional elements from short DNA sequences. The biggest challenge will be adapting current tools--and developing new tools--to perform accurate taxonomic assignment, function prediction, and threat assignment of these sequences." Fraunhofer executive director Adam Porter says the Fraunhofer team will incorporate these modules into an overarching infrastructure that includes visual performance monitoring dashboards.

Full Article

flexible circuitry Team Builds Flexible New Platform for High-Performance Electronics
UW-Madison News
Renee Meiller
September 28, 2017

Engineers at the University of Wisconsin-Madison (UW-Madison) say they have constructed the world's most functional flexible transistor, along with a fast, simple, and affordable manufacturing process that can be scaled to the commercial level. The team has tweaked the design of a bipolar complementary metal-oxide semiconductor thin-film transistor, which combines speed, high current, and low-power dissipation in the form of heat and wasted energy. "The industry standard is very good," says UW-Madison professor Zhenqiang Ma. "Now we can do the same things with our transistor--but it can bend." Ma's team manufactured the flexible electronics on a single-crystal silicon nanomembrane on one flexible plastic substrate, and their process removes many steps and trims time and cost off of transistor fabrication. Ma says his group completed fabrication in a week, versus the industry standard of three months. "One high-temperature step fixes everything--like glue," Ma notes. "Now, we have more powerful mixed-signal tools."

Full Article
The Benefits and Risks of Using Talking Software to Address Mental Health
Stanford News
Milenko Martinovich
September 25, 2017

In an interview, Stanford University scholars Adam Miner, Arnold Milstein, and Jeff Hancock discuss the growing use of conversational software programs as a tool in mental health therapy, along with their potential advantages and drawbacks. "Conversational agents may allow people to share experiences they don't want to talk about with another person," and thus encourage more patient honesty, Miner says. The availability of such programs to patients amid a limited number of human mental health professionals is another potential benefit that Hancock notes. However, Miner says negative experiences with conversational agents could make patients less willing to seek aid in the future. "A balance must be struck between high-tech and high-touch treatment," he says. Meanwhile, Milstein says mainstream healthcare organizations will likely not adopt the technology until it demonstrates clear therapeutic benefits. "Clinical trials can address safety and efficacy, but clarity around user expectations and rules governing medical devices are needed," Miner says.

Full Article
Are Computers Already Smarter Than Humans?
Lance Whitney
September 29, 2017

In some respects, computers can be more intelligent than humans in terms of learning and problem solving, according to experts such as Technion-Israel Institute of Technology professor Shlomo Maital. "Today, computers can learn faster than humans, e.g., [IBM's] Watson can read and remember all the research on cancer, no human could," Maital notes. Meanwhile, LearnOpenCV.com's Satya Mallick says artificial intelligence can be trained to outperform humans in certain tasks such as visual recognition, in which "there is a vast amount of data we can gather to solve these tasks and/or they are repetitive tasks." However, experts concur computers still cannot match humans for their general intelligence, creativity, or common-sense knowledge or understanding of the world. PsychCentral.com CEO John Grohol sees the attainment of creativity and intuition by computers as perhaps an unreachable goal, while Nara Logics CEO Jana Eggers says computers lack the human concept of meaning, such as emotional experience.

Full Article
ASU Engineer Works to Develop a Portable Brain-like Computer
Monique Clement
September 27, 2017

Arizona State University professor Jae-sun Seo has a goal to fit state-of-the-art learning algorithms on small footprint devices such as smartphones with the help of custom-designed hardware, using a U.S. National Science Foundation grant. "The overarching goal of this project is to build brain-inspired intelligent computing systems using custom hardware designs that are energy efficient and programmable for various cognitive tasks, including autonomous driving, speech, and biomedical applications," Seo says. To reduce the resources required by learning algorithms, Seo plans to mimic the human brain's ability to selectively and adaptively learn and recognize real-world data instead of using the redundant computations and exhaustive searches now used by algorithms. Seo also plans to investigate low-power, real-time on-chip learning methods, along with new memory compression schemes for software and hardware design, efficient on-chip power management capable of adapting to abrupt changes in cognitive workloads, and cross-layer optimization of circuits, architectures, and algorithms.

Full Article

A challenge to find toothbrushes within a bathroom scene. In Plain Sight
The UC Santa Barbara Current
Sonia Fernandez
September 26, 2017

Researchers at the University of California, Santa Barbara (UCSB) have found deep neural networks can be superior to humans in conducting visual searches, because the networks do not overlook mis-scaled targets. The networks search across whole scenes and use the visual characteristics of the object itself, while humans also employ the relationships between objects and their context within the scene to direct their eyes. On the other hand, humans were found to be better than computer vision in verifying the presence of different target objects in real-world scenes that may or may not feature them, as networks returned a higher number of false positives. The researchers note this mechanism is a useful approach of the human brain for processing scenes quickly, removing distractors, and lowering false positives. "The findings might suggest ways to improve computer vision by implementing some of the tricks the brain utilizes to reduce false positives," says Middle East Technical University in Turkey professor Emre Akbas, who worked on the project while at UCSB.

Full Article

Berkeley Lab members sitting around a table Quantum Computation to Tackle Fundamental Science Problems
Berkeley Lab News Center
Linda Vu
September 26, 2017

Two research teams at Lawrence Berkeley National Laboratory (Berkeley Lab) will use U.S. Department of Energy (DoE) funding to develop and test algorithms, compiling methods, and scheduling tools to enable near-term quantum computing platforms for scientific discovery in the chemical sciences. "The question we are currently asking is whether there are specific problems that we can solve with more specialized quantum computers," says Berkeley Lab researcher Irfan Siddiqi, who will co-direct DoE's Advanced Quantum-Enabled Simulation (AQuES) Testbed project. AQuES co-director Jonathan Carter says Berkeley Lab is co-designing quantum processors in close collaboration with the scientists developing the quantum algorithms, compiling techniques, and scheduling tools, which should be very useful for addressing the challenge of sustaining quantum coherence. Carter compares quantum computing to a chess game, noting, "[Quantum bits] lose coherence in a really short amount of time, so it's up to us to figure out the most useful set of moves we can make."

Full Article
Computer Science Professors Awarded $1 Million From National Science Foundation Funding
Pipe Dream-Binghamton University
Rachel Lacklow; Ari Bateman
September 28, 2017

Binghamton University professors Ping Yang and Guanhua Yan, along with colleagues at Wayne State University, will use a $999,999 U.S. National Science Foundation (NSF) grant to develop software to elevate the credibility of scientific research by preventing the falsification of research findings. "If research results are questionable because the underlying data has been corrupted or altered, faith in all scientific results can be lost, something we definitely don't want to happen," says NSF's Anita Nikolich. Yang cites current scientific workflow systems as imperfect and lacking required security infrastructure, and he and Yan will develop software that maintains a secure log of any workflow activity, including alterations to the software and parameters, data input, and the results of data analysis. The software also will monitor who is responsible for any changes, and will enable scientists to find honest errors hindering a study's reproducibility by supplying a comprehensive log of all recorded data.

Full Article

Sebastian Thrun Sebastian Thrun Talks Flying Cars, Automated Teaching, and an AI Arms Race With China
Dave Gershgorn
September 25, 2017

In an interview, Google moonshot lab X co-founder Sebastian Thrun says flying cars are ready to be built, given advances in unmanned drone technology. Current-generation drones' use of autonomous software to perform the most difficult flight tasks "gives the user the easy part," Thrun notes. For flying cars to move forward, Thrun says knowing how to interface with the ground and different structures will be essential, since power lines, foliage, airstreams, noise regulations, failure modes, and many other factors are in play. Thrun also envisions artificial intelligence becoming more deeply incorporated into education, including grading. "My general hypothesis is that almost any generally repetitive task, if you have enough data, an AI that looks over your shoulder can take over more and more of it," he says. With China making gains in AI, Thrun recommends the U.S. hold to its present course of sustaining its global leadership and doubling down on AI investment.

Full Article
Here's to Adele, Dan and Alan for giving us Smalltalk
ACM Ambassador Program

Association for Computing Machinery

2 Penn Plaza, Suite 701
New York, NY 10121-0701

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: technews@hq.acm.org