Welcome to the April 19, 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."

Firefighter with Colossus robot outside Notre Dame Firefighters Had Secret Weapon When Notre Dame Caught Fire: A Robot Named 'Colossus'
The Washington Post
Peter Holley
April 17, 2019

Paris firefighters used a 1,100-pound, tank-like robot to mitigate damage to Notre Dame cathedral and prevent the conflagration from spreading further, by entering environments too hazardous for humans. The "Colossus" robot employed a motorized water cannon to spray down the cathedral's stone walls, and extinguish the flames. Paris Fire Brigade Commander Jean-Claude Gallet credited Colossus with reducing temperatures within the building's glass-filled nave, and saving firefighters' lives. Colossus manufacturer Shark Robotics said the 2.5-feet-wide by 5.25-feet-long robot can carry 1,200 pounds, and be operated from nearly 1,000 feet away, via joystick. The battery-powered Colossus is waterproof and fireproof, as well as tolerant of thermal radiation; it also can crawl up stairs, and be outfitted with cameras, sensors, and a smoke-extracting fan.

Full Article
*May Require Paid Registration
University of Louisville Unveils 'Game-Changing' Partnership With IBM
Louisville Courier-Journal
Morgan Watkins
April 17, 2019

The University of Louisville (U of L) has partnered with IBM to establish the first skills academy in the U.S., to educate students in digital fields like artificial intelligence (AI). The collaborators said the goal is to help U of L, the city of Louisville, and the commonwealth prepare and train next-generation professional Kentuckians. The IBM Skills Academy is slated to open in the fall, offering students and faculty access to a curriculum, and educational tools to train them in rapidly expanding fields like AI, blockchain, cybersecurity, data science, and quantum computing. U of L participants will gain access to IBM's software and cloud technology, along with industry expertise. Faculty will be trained in data science, cloud, and cybersecurity, and assist others to attain IBM digital credentials or college credit, to help them pursue high-tech careers.

Full Article

rabbit in animal testing New Algorithm Could Save Thousands of Animals From Toxic Testing
Silicon Republic
Colm Gorey
April 17, 2019

Rutgers University researchers announced an algorithm designed to test for chemical toxicity, to keep people safe in various industries, and potentially sparing thousands of animals from being experimental subjects. The algorithm extracts data from the PubChem chemical database, then compares chemical fragments from tested compounds with those from untested compounds, mathematically grading similarities and differences to predict an untested chemical's toxicity. The researchers trained the algorithm using 7,385 compounds of known toxicity, and presented the program with 600 new compounds. The algorithm had a 62% to 100% success rate in predicting oral toxicity levels for several chemical groups, although this does not suggest animal testing will be completely phased out. Rutgers' Hao Zhu said, "This model takes an important step toward meeting the needs of industry, in which new chemicals are constantly under development, and for environmental and ecological safety."

Full Article
The Border Guards You Can't Win Over With a Smile
BBC News
Molly Kendrick
April 17, 2019

Governments are testing artificial intelligence (AI) to make passport control more efficient, identify potentially dangerous travelers, and spot smuggled goods at borders. U.S. technology company Unisys has been working with the U.S. Customs and Border Patrol to advance the LineSight threat assessment system, which quickly digests data about travelers from different government agencies and other sources, providing a mathematical risk rating. Experts worry that AI trained to recognize patterns or behavior with historic datasets can mirror biases within that data, leading to incorrect immigration and border admission decisions. Unisys' John Kendall said the company hopes to address this by letting the LineSight algorithm learn from its errors. Said the Brennan Center for Justice's Eric Posey, "We need transparency as to how the algorithm...is developed and implemented, how different types of data will be weighted in algorithmic calculations, how human decision-makers are trained to interpret AI conclusions, and how the system is audited."

Full Article
'Disastrous' Lack of Diversity in AI Industry Perpetuates Bias, Study Finds
The Guardian
Kari Paul
April 16, 2019

New York University's AI Now Institute recently conducted a survey of more than 150 studies and reports, and found that a lack of diversity in the artificial intelligence (AI) field has contributed to flawed systems that perpetuate gender and racial biases. The AI field, which is overwhelmingly white and male, is at risk of replicating or perpetuating historical biases and power imbalances. The report cites examples such as image recognition systems making offensive classifications of minorities, chatbots adopting hate speech, and Amazon technology failing to recognize users with darker skin colors. In addition, more than 80% of AI professors are men, and only 15% of AI researchers at Facebook and 10% of AI researchers at Google are women. The demographics of the AI field are reflective of a larger problem across computer science and other science, technology, engineering, and math (STEM) fields. Said Danaë Metaxa, a PhD candidate and researcher at Stanford focused on issues of the Internet and democracy, "Essentially, the lack of diversity in AI is concentrating an increasingly large amount of power and capital in the hands of a select subset of people."

Full Article

pilot approaching helicopter Aviation Automation Climbs New Heights With ALIAS
Federal Computer Week
Lauren C. Williams
April 17, 2019

The U.S. Defense Advanced Research Projects Agency (DARPA)'s Aircrew Labor In-Cockpit Automation System (ALIAS) project aims to develop autonomous artificial intelligences (AIs) to improve flight safety and performance in battlefield operations. ALIAS' goals include producing a customizable, drop-in, removable kit so fewer onboard crew members will be needed on military aircraft. With its initial fly-by-wire experiment led by Sikorsky scheduled for completion in May or June, ALIAS would enable advanced automation to be added to existing aircraft. DARPA in 2016 proved the effectiveness of the effort's sensory and avoidance capabilities with a Cessna 172G aircraft, approaching an unmanned aerial system from multiple angles. DARPA's Lt. Col. Philip Root said once a fly-by-wire AI has been successfully demonstrated, "we can begin adding the autonomy flight controls—operating in the background like a lane assist [feature in cars that helps] the human operator avoid a tree."

Full Article
Neuron, Synapse-Mimetic Spintronics Devices Developed
Tohoku University
April 16, 2019

Scientists at Tohoku University in Japan have developed devices that function like a human brain's neurons and synapses, which hold promise for future energy-efficient and adoptive computing systems. The researchers produced an artificial neuron and synapse, using spintronic technology designed to simultaneously employ an electron's electric and magnetic properties. The devices were microfabricated from an earlier functional antiferromagnetic/ferromagnetic material system, which exhibited basic neuronal and synaptic behavior, based on the same concept of spintronics. The resulting spiking neural network carries benefits beyond those of modern artificial intelligence for processing and prediction of temporal information, and the expansion of the developed technology to unit-circuit-, block-, and system-level should lead to computers that process time-varying data like voice and video, with a small amount of power or edge devices, that adjust to users and the environment.

Full Article

two-core biomaterial, illustration A Biosynthetic Dual-Core Cell Computer
ETH Zurich
Peter Rüegg
April 16, 2019

Researchers at ETH Zurich in Switzerland have integrated two CRISPR-Cas9-based core processors into human cells, marking a significant breakthrough toward creating powerful biocomputers. A special variant of the Cas9 protein forms the core of the processor. In response to input delivered by guide RNA sequences, the CPU regulates the expression of a specific gene, which then makes a particular protein. The method allows researchers to program scalable circuits in human cells, consisting of two inputs and two outputs that can add two single-digit binary numbers. The cell computer could be used to detect biological signals in the body, process them, and respond to them accordingly. The cell could use a properly programmed CPU to interpret two different biomarkers as input signals. Next, the researchers want to integrate a multicore computer structure into a cell, which would have even more computing power than the current dual core structure.

Full Article
Teaching Old Codes New Tricks
National Center for Atmospheric Research
Laura Snider
April 16, 2019

The National Center for Atmospheric Research (NCAR) and IBM's The Weather Company are collaborating to test the use of graphical-processing units (GPU) to produce accurate global weather forecasts. Said NCAR's Rich Loft, "The idea is that even though the processing elements in a GPU can do fewer mathematical calculations per second individually, you could have 100 times more of them than in a central-processing unit (CPU)]. By working together, they might be able to achieve a higher sustained performance." The partners have been working to enable NCAR's next-generation Model for Prediction Across Scales to operate on GPUs, retaining the original code to run on both CPUs and GPUs. IBM hopes this research could yield a nearly 200% upgrade in its forecasting resolution for much of the world, and possibly supply first-class forecasts for less technologically advanced countries.

Full Article

medical testing, illustrative photo AI Test Pinpoints More Cancers Targeted by Astra, Glaxo Drugs
John Lauerman
April 15, 2019

Researchers at Harvard Medical School and the University of Cambridge in the U.K. suggest using artificial intelligence (AI) to locate certain tumor cells could help identify patients that drugs from AstraZeneca and GlaxoSmithKline would benefit. The researchers' machine learning algorithm can analyze genetic tests, which detect patterns in tumors demonstrating whether they have a deficiency in homologous DNA repair. This deficiency makes those growths susceptible to treatment by medications called PARP inhibitors, like Astra's Lynparza or Glaxo's Zejus. Oncologists usually identify this susceptibility by testing for mutations that suppress cellular DNA repair, using a specific mechanism; however, many tumors lacking such flaws also may possess the same weakness, but are harder to detect in the absence of a specific gene test. The researchers said the new algorithm is applicable to genetic tests that are already performed on many tumors.

Full Article

Geofense, illustration Tracking Phones, Google Is a Dragnet for the Police
The New York Times
Jennifer Valentino-DeVries
April 13, 2019

A new kind of search warrant, called a "geofence" warrant, that draws on an enormous Google database employees call Sensorvault, turns the tactic of tracking cellphone users' locations into a digital dragnet for law enforcement, marking the latest example of how personal information is being used for purposes many users never expected. However, as privacy concerns have mounted among consumers, policymakers, and regulators, technology companies have come under intensifying scrutiny over their data collection practices. It is unclear how often these search requests have led to arrests of convictions, because many of the investigations are still open and judges frequently seal the warrants. The technique illustrates a troubling phenomenon for privacy advocates: anytime a technology is developed that could be used in surveillance, law enforcement will eventually try to utilize it. For example, Sensorvault includes detailed location records involving at least hundreds of millions of devices worldwide and dating back nearly a decade.

Full Article
Programming Languages: Don't Bother Learning These Ones in 2019
Liam Tung
April 12, 2019

Codementor, a startup that connects developers with questions to developers with answers, has created a list of the worst programming languages to learn, based on community engagement, growth, and the job market. Last year, the company decided that Dart, Objective-C, CoffeeScript, Lua, and Erlang were the top five programming languages not worth learning. This year, the company focused on which languages aspiring developers should not learn as a first programming language. For this reason, it excluded the top three most popular languages: JavaScript, Python, and Java. The company's data suggests that languages to not bother learning this year are Elm, CoffeeScript, Erlang, and Perl. Kotlin, a popular language for building Android apps, rose from 18th place to 11th place on Codementor's worst-to-best list, while Dart was named the "most improved" language.

Full Article
Homeland Security Warns of Security Flaws in Enterprise VPN Apps
Zack Whittaker
April 12, 2019

Several enterprise virtual private networking (VPN) apps are vulnerable to a security bug that can allow an attacker to remotely break into an organization's internal network, according to the U.S. Department of Homeland Security's Cybersecurity and Infrastructure Security Agency, which published an alert about the vulnerability following the recent public disclosure by Carnegie Mellon University's CERT/CC. The government found that VPN apps built by four vendors—Cisco, Palo Alto Networks, Pulse Secure, and F5 Networks—improperly store authentication tokens and session cookies on a user's computer. The apps generate tokens from a user's password, which are stored on the computer to keep the user logged in without having to reenter the password every time. However, if these tokens are stolen, they can allow access to the user's account without needing the password. To date, only Palo Alto Networks has confirmed its GlobalProtect app was vulnerable, and the company issued a patch for both its Windows and Mac clients.

Full Article
How Nintendo Made VR Work—Without Those Dorky Helmets
The Wall Street Journal
Matthew Kitchen
April 11, 2019

Nintendo's LABO virtual reality (VR) system does not come with a cumbersome, uncomfortable helmet that many other VR systems have. Instead, the user attaches an interchangeable series of toys onto a set of cardboard goggles, and holds them up to the eyes using organic movements. In addition, LABO's images are exceedingly cartoonish compared with other, more realistic virtual worlds. The untethered system also allows the user to remain seated while swiveling 360 degrees to navigate the VR environment. The kit includes the goggles, plus five intuitive toys: the Blaster; the Camera; a Bird with flappable wings; an Elephant that lets the user paints with its trunk; and a Wind Pedal the user pumps to keep virtual balls in the air.

Full Article
*May Require Paid Registration
2019 ACM Europe Summer School
ACM WiSec 2019 Conference

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: [email protected]