Welcome to the August 31, 2020 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."
|
|
|
Technology Keeps Baltimore in Touch with Seniors Through Hurricanes, Pandemic
The Baltimore Sun
Meredith Cohn
August 27, 2020
Baltimore health officials are using a new computer system created by volunteer programmers to keep in touch with senior living facilities during crises like hurricanes and the current pandemic. The Code for Baltimore volunteer group worked with government technology development firm Bellese Technologies to develop the system, which automatically emails facilities to assess needs and supply information, with capability for texts soon to be added to its repertoire. In its initial weeks, the system has been used to survey senior facilities about specific needs related to the coronavirus pandemic, via email; health department personnel can respond directly to queries or refer them to outside resources. Facilities' responses also have alerted officials to residents' loneliness or boredom during quarantine, prompting them to link seniors with electronic devices virtually to outside groups.
|
Elon Musk Demonstrates Brain-Computer Tech Neuralink in Live Pigs
CNBC
Lori Ioannou; Christina Farr
August 28, 2020
Tesla and SpaceX CEO Elon Musk's brain-machine interface company, Neuralink, on Friday performed a live demonstration of a wireless interface in three pigs, allowing the animals' neural signals to be visualized in real time. Musk described the system as "like a Fitbit in your skull," and in July 2019 he expressed the hope of having such an implant in a human subject by the end of 2020. Neuralink said the system uses flexible "threads" a tenth the thickness of a human hair, and it has tested the technology on at least 19 different animals with robots, with a success rate of about 87%. Musk said the system will not be used on humans in the short term, as “There will be ethical and safety issues to work through, and for a long time, it’s likely that you’ll have to have a real medical need to access this technology.”
|
Data Scientists Salaries, Jobs Immune to Covid-19, Survey Finds
Forbes
Gil Press
August 27, 2020
A survey of data scientists by executive recruitment firm Burtch Works found the Covid-19 pandemic does not appear to be affecting their salaries or job availability. The researchers said, "The fact that 45% of organizations are keeping analytics front and center could be part of the reason why layoffs and furloughs are still the exception for analytics teams." In July, 7.6% of respondents said their team had increased hiring due to the pandemic. Median base salaries for data scientists as individual contributors range from $95,500 at level 1 to $165,000 at level 3; managers' median base salaries range from $150,750 at level 1 to $250,000 at level 3. Compared to 2019, median base salaries this year have either held steady or climbed moderately at most levels, driven by demand.
|
Robot Boats Leave Autonomous Cars in Their Wake
The Wall Street Journal
Christopher Mims
August 29, 2020
U.K. ocean-research nonprofit Promare will launch an autonomous boat next spring to retrace the trans-Atlantic journey of the Mayflower in a promotional demonstration of technology that can substitute for crew on voyages deemed too costly or risky. Marine AI is building the unmanned Mayflower, which is being programmed to manage storms. IBM's Andy Stanford Clark said another challenge is programming the ship's artificial intelligence not only to comply with maritime regulations in international waters, but also to explain its decisions. The ship's software is partly derived from repurposed financial-services technology, while Promare trained its computer-vision system on images of vessels, buoys, and floating debris.
*May Require Paid Registration
|
Researchers 3D-Print Lifelike Heart Valve Models
University of Minnesota Twin Cities
August 28, 2020
Researchers at the University of Minnesota (U of M) and medical device company Medtronic have developed a new process for three-dimensional (3D) printing of lifelike models of the heart's aortic valve and surrounding structures. The aortic root models are produced from computed tomography scans of each patient to match the exact shape of their hearts, then assembled with specialized silicone-based inks and a customized 3D-printing technique. The models are designed to help doctors prepare for a Transcatheter Aortic Valve Replacement, and the university's 3D printers can mimic both the soft tissue components of the model and the hard calcification on the valve flaps. U of M's Michael McAlpine said, "Someday maybe these ‘bionic’ organs can be as good as or better than their biological counterparts.”
|
Windows Computers Targets of 83% of Malware Attacks in Q1
PC Magazine
Jason Cohen
August 28, 2020
Microsoft Windows and Android security evaluator AV Test's 2019/2020 Security Report found malware targeting Windows computers constituted 83.45% of all malware attacks in the first quarter of this year. The report said 114 million new pieces of malware were developed last year, and anticipates 160 million new malicious programs by the end of this year, with many hackers exploiting Covid-19's disruption to spread malware and boost phishing attacks. The CVE database of known system vulnerabilities found that Microsoft has more than 660 dangerous flaws, with 357 of them attributed to Windows 10. Besides backdoor vulnerabilities, 64.31% of Windows-targeting malware were trojans, 15.52% were viruses, and 7.79% were worms. As a result of these various threats, Windows 10, which is loaded into about 51% of the world's computers, was rated the least secure operating system.
|
Foiling Illicit Cryptocurrency Mining with AI
Los Alamos National Laboratory News
August 20, 2020
Computer scientists at the U.S. Department of Energy's Los Alamos National Laboratory have developed an artificial intelligence (AI) system that could potentially detect malware that hijacks supercomputers for cryptocurrency mining. The system compares the contours in an algorithm's flow-control graph to a catalog of graphs for programs permitted to run on a given computer, but also checks for the presence of a graph that identifies programs that should be running. The researchers compared a known benign code to a malicious bitcoin-mining code with their system, which identified the illicit mining operation faster and more reliably than conventional, non-AI analyses. The system's reliance on graph comparisons renders it immune to common code-masking deceptions that illicit cryptocurrency miners employ.
|
MIT Wireless System Can Monitor What Care Facility Residents Are Doing While Preserving Privacy
TechCrunch
Darrell Etherington
August 25, 2020
Researchers at the Massachusetts Institute of Technology's Computer Science and Artificial Intelligence Laboratory developed a wireless system that can track the movements, vital signs, and activities of long-term care and assisted-living facility residents without breaching their privacy. The system, dubbed RF-Diary, combines a map of a person's living space with the kinds of activities that happen in different scenarios. The team trained RF-Diary on wireless signals generated by people performing known activities in these spaces, and taught it to categorize activities from new people in completely new locations. The system was better than video monitoring at identifying the activities of individuals even in low-light settings, or when occluded by other objects.
|
QUT Algorithm Could Quash Twitter Abuse of Women
Queensland University of Technology
August 28, 2020
An algorithm developed at the Queensland University of Technology (QUT) in Australia can identify misogynistic content on Twitter with 75% accuracy. The researchers mined a dataset of 1 million tweets, which they refined to 5,000 by searching for those containing certain abusive keywords. Those tweets were categorized as misogynistic or not based on context and intent and input to the machine learning classifier to build its classification model. Said QUT's Richi Nayak, "Teaching a machine to differentiate context, without the help of tone and through text alone, was key to this project's success, and we were very happy when our algorithm identified 'go back to the kitchen' as misogynistic; it demonstrated that the context learning works."
|
Parking Technology Aims to Manage Curb Space Virtually
The Washington Post
Lori Aratani
August 29, 2020
Washington, D.C., is the first U.S. city to test a system that transmits real-time information about curbside parking availability and space size to delivery drivers. The system was developed by curbFlow, and company founder Ali Vahabzadeh said it relies on a batch of "computer vision devices" installed in businesses' windows. The camera- and microprocessor-equipped devices record when curb space is occupied near the establishment, and send that information to drivers. Vahabzadeh said the virtual curbFlow system's development was driven by Covid-19 forcing many people to work from home, vastly increasing demand for takeout and delivery services, with an accompanying surge in congestion as drivers vie for curb space. Said Vahabzadeh, "We're effectively greasing the wheels of the delivery economy."
*May Require Paid Registration
|
Google Performed the First Quantum Simulation of a Chemical Reaction
New Scientist
Leah Crane
August 27, 2020
Researchers at Google have simulated a chemical reaction using a quantum computer for the first time. Because it is difficult for quantum computing to achieve the precision necessary to simulate large atoms or chemical reactions, the research marks a step toward finding a practical use for quantum computers. The researchers used Google's Sycamore device to simulate a diazene molecule undergoing a reaction in which the two hydrogen atoms move into different configurations around the two nitrogen atoms. The researchers said more qubits will be needed, along with adjustments to the calculation, to scale up the algorithm to simulate more complex reactions. Said Google's Ryan Babbush, "We're doing quantum computations of chemistry at a fundamentally different scale now. The prior work consisted of calculations you could basically do with pencil and paper by hand, but the demonstrations we're looking at now, you'd certainly need a computer to do it."
|
Algorithm Aims to Alert Consumers Before They Use Illicit Online Pharmacies
Penn State Institute for Computational and Data Sciences
August 27, 2020
Pennsylvania State University (Penn State) researchers said a new algorithm may be able to identify illicit online pharmacies that provide customers with substandard medications. The computer model uses a comparative method that starts from a baseline of reputable online pharmacies, in order to spot any apparent abnormalities. The team examined several traits of online pharmacies, and identified relationships between them and other websites as vital for ascertaining their legitimacy. Former Penn State researcher Sowmyasri Muthupandi said, "Among all the attributes we found that it's these referral websites that paint a clearer picture when it comes to classifying online pharmacies."
|
Machine Learning Peeks Into Nano-Aquariums
University of Illinois News Bureau
Lois Yoksoulian
August 24, 2020
Researchers at the University of Illinois, Urbana-Champaign (U of I) have developed a machine learning workflow that streamlines the process of interpreting data from watching nanoparticles interact via liquid-phase electron microscopy. The machine learning workflow builds off an existing neural network, known as U-Net, that successfully identified irregular cellular features using other types of microscopy. Measurements from about 300,000 pairs of interacting nanoparticles were collected for the study. Said U of I's Qian Chen, "Developing a method even to see these particles was a huge challenge. Figuring out how to efficiently get the useful data pieces from a sea of outliers and noise has become the new challenge.” The researchers have made the source code for the machine learning program publicly available.
|
|
