Welcome to the June 25, 2021 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."
|
|
|
WWII Codebreaker Alan Turing Becomes 1st Gay Man on British Bank Note
NBC News
Shira Pinson
June 23, 2021
The Bank of England this week rolled out new £50 bank notes featuring World War II codebreaker Alan Turing, known as the "father of computer science and artificial intelligence," on what would have been his 109th birthday. Bank of England's Sarah John said, "He was a brilliant scientist whose thinking still shapes our lives today. However, his many contributions to society were still not enough to spare him the appalling treatment to which he was subjected simply because he was gay. By placing him on this new £50, we are celebrating his life and his achievements, of which we should all be very proud."
|
The Internet Eats Up Less Energy Than You Might Think
The New York Times
Steve Lohr
June 24, 2021
An analysis by two scientists found claims about the Internet's energy consumption are exaggerated by well-intentioned scientists who seldom account for the pace of energy efficiency advancements and how digital systems work. The researchers cite as examples international network operators Telefonica and Cogent's published data traffic and energy use statistics for 2020: Telefonica managed a 45% gain in network traffic with no boost in energy consumption, while Cogent's electricity use dropped 21% as traffic rose 38%. The researchers also criticized the practice of examining one high-growth sector of the technology industry and assuming proportional industry-wide growth of electricity use.
*May Require Paid Registration
|
GPS Cyberattack Falsely Placed U.K. Warship Near Russian Naval Base
New Scientist
David Hambling
June 24, 2021
A cyberattack may have been involved in a naval confrontation this week between Russia and a British warship in the Black Sea that never really happened. The global positioning system (GPS)-tracking Automatic Identification System (AIS) last week showed both a U.K. warship and a Dutch naval vessel coming within a few kilometers of a Russian naval base at Sevastopol, but a live Web camera feed confirmed that both ships were docked in Odessa, Ukraine, at the time. The spoofing in this case suggests a deliberate deception, as the ships' coordinates were changed gradually to imitate normal travel. Dana Goward at the Resilient Navigation and Timing Foundation said Russia could have executed the spoofing attack, and warned that such a hack "could easily lead to a shooting war by making things more confusing in a crisis."
|
ML Methods Could Improve Environmental Predictions
University of Minnesota News and Events
June 22, 2021
New process- or knowledge-guided machine learning (ML) techniques can predict flow and temperature in river networks more accurately even when data is scarce, according to researchers at the University of Minnesota, the University of Pittsburgh (Pitt), and the U.S. Geological Survey. The work involved an algorithm that was taught physical rules to generate more accurate forecasts and identify physically significant relationships between inputs and outputs. The method was designed to avoid common traps in ML-based prediction by informing the model through correlation across time, spatial links between streams, and energy budget equations. Pitt's Xiaowei Jia said, "Accurate prediction of water temperature and streamflow [can assist in] decision making for resource managers, for example helping them to determine when and how much water to release from reservoirs to downstream rivers."
|
'Huge Potential' in Virtual Clinical Trials
University of Leeds (U.K.)
June 23, 2021
An international research team led by the U.K.'s University of Leeds replicated the results of real-life clinical trials using digital simulations of patient groups. The proof-of-concept study assessed the use of in-silico trials to evaluate the effectiveness of a flow diverter in treating brain aneurysms. The researchers used clinical databases to build a virtual patient population that resembled actual patient populations in three previous flow diverter clinical trials. A computational model was used to analyze the implanted device's impact on blood flow in each virtual patient. The in-silico trial predicted 82.9% of virtual patients with normal blood pressure would be treated successfully with a flow diverter, compared with successful treatment rates of 86.8%, 74.8%, and 76.8% in the traditional clinical trials.
|
App Taps Unwitting Users Abroad to Gather Open Source Intelligence
The Wall Street Journal
Byron Tau
June 24, 2021
Gig workers, often in developing countries, are being recruited to gather open source intelligence for governments through a mobile phone application. San Francisco-based Premise Data pays workers to perform data collection and observational reporting tasks like capturing photos, and in recent years has tapped this workforce to conduct basic reconnaissance and gauge public opinion for the U.S. military and foreign governments. Premise's Maury Blackman said, "Data gained from our contributors helped inform government policymakers on how to best deal with vaccine hesitancy, susceptibility to foreign interference and misinformation in elections, as well as the location and nature of gang activity in Honduras."
*May Require Paid Registration
|
Algorithm Helps Autonomous Vehicles Find Themselves, Summer or Winter
California Institute of Technology
Robert Perkins
June 23, 2021
Visual terrain-relative navigation (VTRN) now can operate effectively regardless of seasonal changes, thanks to a new algorithm. California Institute of Technology researchers applied deep learning and artificial intelligence to eliminate seasonal content that can trip up VTRN systems, which rely on close similarity between the terrain they are looking at and database images. The algorithm utilizes self-supervised learning to educate itself, seeking patterns in images by parsing out details and properties that humans likely would overlook. VTRN systems equipped with the algorithm can localize more accurately: one upgraded system could match 92% of images of summer foliage against winter leaf-off imagery, with the remaining 8% easily addressed through other methods.
|
Rembrandt's 'Night Watch' on Display with Missing Figures Restored by AI
Reuters
Toby Sterling
June 23, 2021
Researchers at the Rijksmuseum in the Netherlands used artificial intelligence to restore missing parts of Rembrandt’s "The Night Watch" for a new exhibit. This marks the first time in 300 years that the 1642 painting is on display in its original size. Strips that were cut from all four sides of the painting during a 1715 move and later lost were recreated by restorers and computer scientists with the help of a copy made by another artist of the time. Images of the original painting and the smaller 1655 copy attributed to Gerrit Lundens were scaled to the same size, with the Lundens work warped to fit with the Rembrandt where the placement of figures and objects slightly differed.
|
Average Time to Fix Critical Cybersecurity Vulnerabilities is 205 Days: Report
ZDNet
Jonathan Greig
June 22, 2021
Software security adviser WhiteHat Security has estimated that the average time to correct critical cybersecurity vulnerabilities increased from 197 days to 205 days between April and May 2021. WhiteHat researchers determined that 66% of all apps used by the utility sector had at least one exploitable bug exposed throughout the year. The top five vulnerability classes WhiteHat researchers observed over the last three months were information leakage, insufficient session expiration, cross-site scripting, insufficient transport layer protection, and content spoofing; many such bugs also can be found and leveraged with little skill or effort. WhiteHat's Setu Kulkarni said the situation highlights a dearth of cybersecurity talent available to most organizations, and an overall scarcity of resources for many sectors wrestling with updates and patches for numerous apps.
|
Implantable Brain Device Relieves Pain in Early Study
NYU Langone NewsHub
June 21, 2021
Experiments by New York University (NYU) Langone scientists demonstrated a computerized brain implant's ability to relieve short-term and chronic pain in rodents. The closed-loop brain-machine interface identifies activity in the anterior cingulate cortex, an area of the brain essential for pain processing; when a computer connected to the implant automatically detects electrical patterns closely associated with pain, it directs the implant to relieve it by stimulating the prefrontal cortex. The NYU Langone researchers implanted electrodes in the brains of dozens of rats, then exposed them to carefully measured amounts of pain, monitoring for how fast they withdrew from the pain source. The researchers said the device accurately detected pain up to 80% of the time.
|
Major Step Forward for Quantum Error Algorithms
NCI Australia
June 21, 2021
Researchers at the University of Sydney have raised the threshold for correcting quantum calculation errors with the help of the Gadi supercomputer of Australia’s National Computational Infrastructure (NCI) organization. The researchers used Gadi to run about 87 million simulations for all possible qubit arrangements and aligned the threshold with the actual error rates of physical quantum computing systems. Said Sydney's David Tuckett, "This step brings us closer to making practical quantum computing possible. Quickly being able to run these simulations on NCI is central to understanding the effectiveness of our qubit arrangements."
|
Scientists Obtain Real-Time Look at How Cancers Evolve
Memorial Sloan Kettering Cancer Center
June 23, 2021
Scientists at New York's Memorial Sloan Kettering Cancer Center (MSK) and Canada's University of British Columbia have demonstrated a machine learning (ML) technique that uses population genetics to accurately forecast the evolution of breast cancer. The researchers analyzed tumor models called patient xenografts, human cancers transplanted into mice, which they subjected to platinum-based chemotherapy treatment and treatment withdrawal. They concurrently employed single-cell sequencing to automatically record the genetic makeup of thousands of individual cancer cells in the tumor, then used the ML fitClone tool to apply population genetics mathematics to cancer cells; the result was a model of individual cells and their clones' behavior. MSK's Sohrab Shah said the goal is to use this method on blood samples to identify specific clones in a tumor, predict their evolution, and personalize treatment.
|
|
