Welcome to the March 6, 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).
Defining AI Beyond a Narrow Cognitive Computing View
The Huffington Post
Mark Skilton
March 4, 2017

Mark Skilton, a professor at the Warwick Business School in the U.K., proposes an artificial intelligence (AI) test that goes beyond the imitation test pioneered by mathematician Alan Turing. He stresses achievements such as IBM Watson beating human Jeopardy players are examples of specialized intelligence, but are very narrow elements of general intelligence. "Intelligence can be broadened into a deeper set of dimensions that reflect not specialized specific mechanistic tasks but a more real interpretation of physical environment generalization and human existence," Skilton says. He envisions a future integration of perception, action, language, and cognition capabilities into a holistic system that can process general unstructured data, manipulate free-form objects into art or structures, and interpret and infer to higher forms of contextual and information awareness. "A public debate would be helpful in exploring these issues in separating the difference of intelligence, artificial intelligence, and reasoning," Skilton says.

Full Article
Social Media Algorithm Says Trump 'Tweets Like a Bachelor'
ITMO University
Anna Huddleston
March 6, 2017

Researchers at ITMO University in Russia have developed an algorithm that predicts a user's marital status with 86-percent accuracy using data from Twitter, Foursquare, and Instagram. However, the algorithm identified U.S. President Donald Trump, who is married, as single, according to his social media activity. The researchers concluded the inconsistency was due to Trump's abnormal activity in the media, indicating he uses Twitter like a bachelor. The researchers say the personality-characterizing algorithm could be used to identify illegal groups, or to find people prone to depression and suicide. ITMO's Kseniya Buraya is studying approaches for describing human personality through social networks. Buraya processes user data with the new algorithm and then adapts the information to the Myers-Briggs Type Indicator, a scale of psychological types based on Jung theses.

Full Article

robot hand at keyboard AI Scientists Gather to Plot Doomsday Scenarios (and Solutions)
Dina Bass
March 2, 2017

Arizona State University (ASU) artificial intelligence (AI) researcher Eric Horvitz and Doomsday Clock expert Lawrence Krauss recently convened about 40 specialists at ASU to brainstorm worst-case AI outcomes ranging from stock market rigging to global war. "To maximally gain from the upside [to AI] we also have to think through possible outcomes in more detail than we have before and think about how we'd deal with them," Horvitz says. Experts whose doomsday scenarios were considered most threatening led panels that discussed attack and prevention strategies. The U.S. Defense Advanced Research Projects Agency's John Launchbury and Tufts University's Kathleen Fisher led a session on automated cyberattacks. Organizers hope the workshop raises awareness of the threat, encourages dialogue of what needs to happen to avoid such outcomes, and merges ideas from different fields. Launchbury believes the participation of policy figures at the workshop will nurture significant progress.

Full Article
Software Engineers Have Figured Out a Way to Turn Charts Into Music for the Blind
Corinne Purtill
March 3, 2017

Legally blind SAS software engineer Ed Summers in February guided the company's release of the Graphics Accelerator, a free browser plug-in that scans Web pages for graphs or charts and translates them into piano notes. The Graphics Accelerator follows the principles of sonification, the art of transforming data into sound. The notes the Graphics Accelerator generates are the result of feedback from students at a school for the blind, who tested the tool's prototypes and found their tinny, robotic sounds to be annoying. Pennsylvania State University professor Mark Ballora notes the human ear is especially capable of recognizing patterns, reading nuanced changes, and identifying multiple information streams simultaneously. He envisions sonification as having broad utility for both sighted and vision-impaired people, and says the SAS tool "has a lot of implications for what science education could look like in 20, 30 years."

Full Article
Artificial Data Give the Same Results as Real Data--Without Compromising Privacy
MIT News
Stefanie Koperniak
March 3, 2017

Researchers at the Massachusetts Institute of Technology (MIT) have developed the Synthetic Data Vault (SDV), a machine-learning system that automatically creates synthetic data. Such artificial data can be used in data science efforts that otherwise would be thwarted due to limited access to authentic data. The use of authentic data raises significant privacy concerns, and the synthetic data can still be used to develop and test data science algorithms and models. The SDV algorithm, known as a recursive conditional parameter aggregation, exploits the hierarchical organization of data common to all databases. The researchers found the synthetic data can successfully replace real data in software writing and testing. They also note the SDV can be scaled to create very small or very large synthetic datasets, facilitating rapid development cycles or stress tests for big data systems.

Full Article

tech image of world map Computer Scientists and Law Scholars Untangle Complexities of Cyberconflict
William Weir
February 28, 2017

Computer science and law professors at Yale University are working together to promote research into the legal and technological aspects of cyberconflict, or politically motivated cyberattacks that do not always escalate to the level of cyberwarfare. Funded with a two-year $406,000 grant from the William and Flora Hewlett Foundation, the cross-disciplinary project will forge a network of experts in the field and include speakers, conferences, and a Yale course co-taught by the three professors. Although there are extensive laws governing physical attacks by state actors, national and international law is largely unclear about state-sponsored cyberattacks. "It's not just that the law's unclear, it's that there are multiple bodies of law--criminal, international, security, and privacy, and civil rights," says Yale professor Scott Shapiro. "One important goal of the project is figuring out the minimal amount that's needed to know to just to have a conversation."

Full Article

Teen looking at poster Singing Posters and Talking Shirts: UW Engineers Turn Everyday Objects Into FM Radio Stations
UW Today
Jennifer Langston
March 1, 2017

Researchers at the University of Washington (UW) are enabling everyday objects to communicate directly with smartphones and car radios and transmit information to them. "The challenge is that radio technologies like Wi-Fi, Bluetooth, and conventional FM radios would last less than half a day with a coin cell battery when transmitting," says UW's Vikram Iyer. "So we developed a new way of communication where we send information by reflecting ambient FM radio signals that are already in the air, which consumes close to zero power." The UW team's "backscattering" system sends messages by reflecting and encoding audio and data within outdoor FM radio transmissions without disrupting the original signals. A demonstration of a "singing poster" is one example of the backscattering system the team implemented. They also are exploring "smart fabric" applications such as sensor-outfitted apparel to monitor vital signs that are sent to a user's phone.

Full Article
Machine-Learning Algorithm Predicts Laboratory Earthquakes
Technology Review
March 3, 2017

Los Alamos National Laboratory researchers have trained a machine-learning algorithm to identify the signs that a laboratory earthquake is about to take place using only the sounds it emits under strain. Although the researchers are hesitant to make any predictions about the algorithm's ability to predict real earthquakes, they say their study opens up new avenues of research in this field. The researchers first created artificial earthquakes by pulling on one block sandwiched between two others. They also packed a mixture of rocky material in between the blocks to simulate the properties of real faults. The researchers then recorded the acoustic emissions and fed them into a machine-learning algorithm to see if it could find a pattern. The researchers found the acoustic emissions were good predictors of a future artificial earthquake, and are looking to apply it to real earthquakes in the near future.

Full Article

Mom and daughter look at laptop Kids Want Parental Help With Online Risk, but Fear Parental Freak Outs
Penn State News
Matt Swayne
February 27, 2017

Researchers at Pennsylvania State University (PSU) conducted a study examining how teenagers communicate with their parents about their online behavior. They found teens rarely talked to their parents about potentially risky online experiences, and parents and children often have greatly different perceptions of and reactions to the same online situations, such as cyberbullying, sexual exchanges, and viewing inappropriate content. "Teens tended to be more nonchalant and say that the incident made them embarrassed, while parents, even though they were reporting more low-risk events, emoted much stronger feelings, becoming angry and scared," says former PSU researcher Pamela Wisniewski, who is now a professor at the University of Central Florida. She says this disconnect could lead teens to refrain from talking about situations that may upset their parents. The researchers presented their findings last week at the ACM Conference on Computer-Supported Cooperative Work and Social Computing (CSCW 2017) in Portland, OR.

Full Article
Virginia Tech Leads DARPA Project to Develop Algorithms and Software Tools for Next Generation Social Science
Virginia Tech News
Barbara Micale
February 27, 2017

The U.S. Defense Advanced Research Projects Agency's (DARPA) Next Generation Social Science (NGS2) program aims to build and evaluate new modeling processes, experimental designs, and methodology to advance rigorous and reproducible large-scale social science studies. DARPA has awarded Virginia Polytechnic Institute and State University, Stanford University, Cornell University, Johns Hopkins University, and Northwestern University a three-year cooperative agreement for up to $3 million for the experimental effort that will leverage sociology, network science, statistical physics, and machine learning. NGS2 will challenge researchers to develop new tools and methods to identify mechanisms of collective identity formation, or how a group of individuals becomes a unified whole. The project will use a novel approach that combines large-scale agent-based casual modeling, data-driven model development, and statistical methods. Researchers also will develop algorithms and software to simulate situations in which individuals must make decisions for personal or group betterment.

Full Article
Using Game Theory to Predict Cyberattacks on Elections and Voting Machines
Research News @ Vanderbilt
Heidi Hall
February 27, 2017

An algorithm based on game theory can be used to protect against efforts to tamper with election results, according to Vanderbilt University professor Yevgeniy Vorobeychik. He says people committing election fraud will want to evade detection, so their interference will likely target voting machines in such a way that the outcomes are close to a tie, eliminating their district's influence on the total vote count. The algorithm Vorobeychik developed generates a random list of districts likely to be targeted, and enables auditors to check for discrepancies in vote totals. Although doing that process by hand would be labor-intensive and could permit fraudsters to determine the selection system, the algorithm makes the process completely unpredictable. Vorobeychik plans to release a version of the software to public agencies for use in polling places during elections or for use in audits after voting ends.

Full Article
Will Democracy Survive Big Data and Artificial Intelligence?
Scientific American
Dirk Helbing; Bruno S. Frey; Gerd Gigerenzer
February 25, 2017; et al.

Artificial intelligence and big data are expected to fundamentally transform both economic structures and human society. Humanity has reached a crossroads in which new technologies, if managed without regard for ethics or regulation, could spell disaster. One example is erosion of systems from within due to a lack of transparency and democratic control, where the manipulation of people by the technology's controllers for political profit can go undetected. This could destroy social cohesion and cause societal fragmentation or even disintegration. Irresponsible manipulation of people and society demands imposition of high standards and a Hippocratic Oath-like code of conduct. Similarly, placing trust in a super-intelligent machine would reduce the quality and diversity of solutions to societal problems. However, this future could be avoided by decentralizing the function of information systems, supporting informational self-determination and participation, improving transparency, reducing information distortion, and supporting collective intelligence and responsible digital behavior by citizens.

Full Article
Reactive Internet Programming
ACM Conferences

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