Welcome to the November 18, 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."
|
|
|
'Extremely Aggressive' Internet Censorship Spreads in the World's Democracies
University of Michigan News
November 17, 2020
University of Michigan (U-M) researchers used an automated censorship tracking system to demonstrate that online censorship is proliferating in even the freest countries. The U-M team's Censored Planet tool collected more than 21 billion measurements over 20 months from 221 nations, and found Internet censorship growing in 10 countries, including unexpected places like Norway, Japan, Italy, India, Israel, and Poland. U-M's Ram Sundara Raman said in many cases, legislation compelling Internet service providers to block clearly objectionable material, like child pornography, sets the groundwork for more aggressive policies—and this means censorship measurement is essential. The U-M team said these findings highlight the effectiveness of Censored Planet's approach, which converts public Internet servers into automated sentries that track and report when access to websites is being inhibited.
|
Don't Fear the Robots, and Other Lessons From a Study of the Digital Economy
The New York Times
Steve Lohr
November 17, 2020
A task force established by the Massachusetts Institute of Technology (MIT) has published its findings on technology's effect on the U.S. workforce, and policies to boost opportunities for American workers in the digital economy. The team determined that automation via robots and artificial intelligence will not lead to a jobless future, but will create new jobs because the technologies cannot match all human abilities. Moreover, the task force believes U.S. training programs must align with market demands via public-private partnerships. Ginni Rometty on the MIT advisory board said, "Technology companies have an obligation to help responsibly prepare society for these higher-skill jobs."
*May Require Paid Registration
|
Israeli Researchers Create AI Capable of Writing Personalized Jazz Solos
Jerusalem Post
Idan Zonshine
November 10, 2020
Researchers at Technion – Israel Institute of Technology have created an artificial intelligence (AI) model that can generate personalized jazz solos. The researchers trained a music language model, BebopNet, to produce symbolic saxophone jazz improvisations to any chord progression, using hundreds of original jazz solos to build the initial data set. The AI trains a personal preference metric to predict notes that reflect the user's personal taste. The user is asked to rate a series of jazz improvisations according to their preference, and a regression model is used to predict their taste. The model then uses "beam search" to optimize the note-generation process in accordance with the user's taste. The researchers observed, “While our computer-generated solos are locally coherent and often interesting or pleasing, they lack the qualities of professional jazz solos related to general structure such as motif development and variations.”
|
World's Fastest Supercomputer Now Even More Powerful Than Before
TechRadar
Joel Khalili
November 16, 2020
Japan's Fugaku has extended its lead as the world's fastest supercomputer in the latest Top500 ranking, with a new peak performance of 442 petaFLOPS on the High Performance Linpack benchmark, up from 416 petaFLOPS in June. Fugaku also raised its score in the mixed-precision high-performance computing (HPC)-artificial intelligence (AI) workload, reaching 2.0 exaFLOPS. The improvements stem from 330,000 cores added to the system's Arm-based Fujitsu A64FX central processing unit capacity, for 7,630,848 cores total. The newest Top500 ranking indicates no major changes in terms of overall rankings, aggregate power, or new entries.
|
'Most Secure' U.S. Election Not Without Problems
Government Technology
Lucas Ropek
November 16, 2020
Although federal officials declared the 2020 presidential election the "most secure in American history," there were still technical problems. Alleged software glitches caused mistakes in vote tabulation for both presidential and local races in certain counties, while some communities suffered temporary miscounts due to clerical errors. Threats of foreign interference appear to have been countered by greater vigilance and stronger cyberdefenses by watchdogs like the Cybersecurity and Infrastructure Security Agency, and multi-stakeholder collaboration and information sharing. However, disinformation and misinformation have continued to fuel polarization of the electorate. Former ACM president Barbara Simons urges greater transparency and committed investment in auditable machinery as top priorities, along with curtailing the use of paperless voting machines.
|
Meat Giant Tyson Girds for Virus Surge, Tracking Covid-19 with Algorithms
The Wall Street Journal
Jacob Bunge
November 15, 2020
Tyson Foods CEO Dean Banks said the U.S. meatpacking company is using infection-tracking algorithms and employee testing to protect workers from a new Covid-19 surge. He said Tyson's Springdale, AR, headquarters uses a company-developed Web dashboard to monitor internal-testing results. In July, Tyson launched a "surveillance testing" program that tests a number of employees at each processing plant every week; to date, more than half of its roughly 141,000 employees have been tested. Banks said the algorithmic system compares positive tests with where employees work in the plants and with publicly reported infection rates in localities around its facilities. Said Banks, "We can dial up the algorithm when we sense there's something going on in the community, and we're much more prepared for a second wave."
*May Require Paid Registration
|
Stretchable Sensor Gives Robots, VR a Human Touch
Cornell Chronicle
David Nutt
November 12, 2020
Cornell University researchers have developed a fiber-optic sensor that, when integrated with low-cost light-emitting diodes (LEDs) and dyes, results in a stretchable skin that senses deformations like pressure, bending, and strain. The sensor employs a stretchable lightguide for multimodal sensing (SLIMS) featuring transparent and dye-impregnated polyurethane elastomeric cores linked to an LED; each core is mated to a red-green-blue sensor chip to detect geometric changes in the optical path of light. The dyes function as spatial encoders. The sensor is coupled with a mathematical model that can separate distinct deformations and pinpoint their precise locations and magnitudes. The researchers engineered a glove with a SLIMS sensor on each finger, as well as a battery and Bluetooth, that can transmit basic data in order to reconstruct its movements and deformations in real time.
|
In Step Toward Quantum Tech, Scientists Synthesize 'Bright' Quantum Bits
UChicago News
Emily Ayshford
November 12, 2020
Physicists and chemists at the University of Chicago and Northwestern University have developed a new technique for generating customized quantum bits (qubits) by chemically synthesizing molecules that encode quantum data into their spin states. The team utilized organometallic chromium molecules to create a spin state controlled via light and microwaves. Exciting the molecules with laser pulses and measuring the light emissions enabled the reading of spin states after placement in a superposition. Varying a few different atoms on the molecules through synthetic chemistry also tweaked their optical and magnetic properties, highlighting the potential for tailor-made qubits. Northwestern's Danna Freedman said, "Our bottom-up approach enables both functionalization of individual units as 'designer qubits' for target applications and the creation of arrays of readily controllable quantum states, offering the possibility of scalable quantum systems."
|
Employee Surveillance Software Demand Increased as Workers Transitioned to Home Working
ZDNet
Eileen Brown
November 16, 2020
A study by virtual private network (VPN) review website Top10VPN found that demand for employee surveillance software rose 55% in June 2020, compared to the average before the Covid-19 pandemic. Worldwide demand climbed 108% in April and 70% in May 2020, compared to searches conducted in 2019. Moreover, queries on "how to monitor employees working from home" increased by 1,705% in April and 652% in May 2020, versus queries in the preceding year. The growth in such a phrase's popularity reflects a lack of readiness among many companies for the sudden transition to working at home. The most popular surveillance tools—Time Doctor, Hubstaff, and FlexiSPY—constitute nearly 60% of global demand for surveillance software, given the range of features offered; 81% of the most in-demand tools offer keystroke logging, and 61% offer Instant Messaging monitoring.
|
Boston Dynamics Dog Robot Learns New Tricks on BP Oil Rig
Reuters
Catherine Koppel
November 13, 2020
Boston Dynamics is programming its "Spot" dog robot to read gauges, look for corrosion, construct maps, and detect methane leaks, now that it is working on a British Petroleum (BP) oil rig in the Gulf of Mexico. BP's Adam Ballard said Spot's observational functions will improve the safety of rig operations by reducing the need for personnel. He said, "We believe a lot of that up-front, remote work preparation can be done with a remotely controlled robot ... being able to pan, tilt, zoom, and really understand the entire area in real conditions, real time." BP hopes to expand Spot's data-gathering capability in the future to enhance areas where humans are limited.
|
New Fiber-Optic Sensors Transmit Data Up to 100 Times Faster
EPFL (Switzerland)
November 13, 2020
Engineers at the Swiss Federal Institute of Technology, Lausanne (EPFL) and China's Beijing University of Posts and Telecommunications have developed an encoding/decoding system that enables fiber-optic sensors to transmit data up to 100 times faster and over a wider area than previous systems. EPFL's Zhisheng Yang and Simon Zaslawski said the system functions like an echo, with light pulses emitted along a fiber, and the reflected signals decoded by a device. EPFL's Luc Thévenaz said, "Other systems are either limited in scope or expensive. But with ours, you just have to add a software program to your existing equipment. No need to adapt your sensors or use complex devices."
|
Engineers Develop Mobile App to Monitor Glucose Levels in Diabetics
News-Medical
Emily Henderson
November 16, 2020
Engineers at the U.K.'s University of Cambridge have developed a computer-vision technology into a free mobile phone application for monitoring glucose levels in diabetics. Created in collaboration with U.K. glucose testing company GlucoRx, the GlucoRx Vision app captures glucose concentrations, time, and date displayed on a conventional glucose test when users take a photo of the test results with a mobile phone camera. Cambridge's Roberto Cipolla said, "The app will vibrate when it's read the information, so you get a clear signal when you've done it correctly. The system is accurate across a range of different types of meters, with read accuracies close to 100%."
|
PNNL Researchers Speed Power Grid Simulations Using AI
Pacific Northwest National Laboratory
Lynne Roeder
November 9, 2020
The Smart Power Grid Simulator (Smart-PGSim) developed by researchers at the Pacific Northwest National Laboratory (PNNL) and the University of California, Merced, efficiently solves power grid simulations using multi-task learning modeling, a novel neural network approach. The researchers integrated grid-specific domain knowledge into the layers of a multi-task learning model. They found that the Smart-PGSim solved power flow calculations about three times faster than a traditional numerical model, with nearly the same accuracy. Said PNNL's Gokcen Kestor, "These techniques can be broadly applied to many scientific high-performance computing applications, not just the optimization problem in power grid simulations. Other potential applications could include fluid dynamic simulation, molecular dynamics simulation, and cosmology modeling."
|
|
