Welcome to the December 10, 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."
|
|
|
Tech Helping Poor People Get Government Aid
The New York Times
Jason DeParle
December 8, 2021
Various organizations are developing technologies to help lower-income Americans overcome bureaucratic obstacles to access government relief. The Propel startup offers the free Providers app, used by 5 million households to manage their food stamp benefits. Meanwhile, the Civilla nonprofit helped Michigan reduce the size of its application for food stamps by 60%. The Code for America nonprofit developed a phone-optimized Web portal that shortens the application process for Californians by three-quarters or more, and designed portals to help poor households claim stimulus checks and the expanded child tax credit during the pandemic.
*May Require Paid Registration
|
Wildfire Dataset Helps Firefighters Save Lives, Property
UC Riverside News
Holly Ober
December 8, 2021
Scientists at the University of California, Riverside (UCR), Stanford University, and Vanderbilt University have built the WildfireDB dataset for simulating the spread of wildfires to help study wildfires and guide emergency response and evacuations. UCR's Ahmed Eldawy called WildfireDB "the first comprehensive and open-source dataset that relates historical fire data with relevant covariates such as weather, vegetation, and topography." The dataset is a compilation of information on the spread of fires in the contiguous U.S. over the last decade. The researchers employed the Raptor satellite data processing system to combine data on historical wildfires with other geospatial features; researchers or firefighters can select relevant data from the dataset to train machine learning models to forecast wildfire dynamics.
|
Generating a Realistic 3D World
MIT News
Lauren Hinkel
December 6, 2021
Researchers at the Massachusetts Institute of Technology (MIT), the MIT-IBM Watson Artificial Intelligence (AI) Laboratory, and Harvard and Stanford universities developed the ThreeDWorld (TDW) platform to simulate high-fidelity audio and visual environments. Said the MIT-IBM Watson AI Lab's Chuang Gan, "We are trying to build a general-purpose simulation platform that mimics the interactive richness of the real world for a variety of AI applications." TDW can produce photorealistic scenes and render audio in real time for compilation into audio-visual datasets, modified through interaction, and adapted for human and neural network learning and prediction tests.
|
AI Tool Detects Most Common Climate Falsehoods
The Washington Post
Kasha Patel
December 9, 2021
A machine learning algorithm developed by researchers at Australia's Monash University can detect climate misinformation online. The researchers programmed the artificial intelligence tool to flag Websites that present false or misleading information about climate change science and remedies in multiple categories. The team analyzed over 250,000 documents from 1998 to 2020 from 20 popular, mostly U.S.-based conservative think tanks, and 33 central contrarian blogs that mainly refuted climate information about science/scientists and solutions. Monash's John Cook said, “There is a dearth of research into understanding attacks on climate science and scientists themselves, let alone developing solutions.”
*May Require Paid Registration
|
Physical Features Boost Efficiency of Quantum Simulations
Los Alamos National Laboratory News
December 6, 2021
Researchers at the U.S. Department of Energy's Los Alamos National Laboratory have published two papers explaining how physical properties can speed quantum simulation. One paper illustrates that a quantum simulation algorithm's complexity depends on the relevant energy scale, and not the system's full range of energies. A quantum system that solely explores low-energy states could be modeled with low complexity on a quantum computer without crashing from errors. A second paper, co-authored with the California Institute of Technology's Shouzhen Gu, details a quantum simulation algorithm that can surpass speed limits suggested by the time-energy uncertainty principle. Such quantum systems also can be modeled very efficiently on quantum computers, the researchers said.
|
Researchers Explain Why They Believe Facebook Mishandles Political Ads
NPR
Bill Chappell
December 9, 2021
Facebook allowed more than 70,000 political ads to run during last year's U.S. elections, according to researchers at New York University (NYU) and Belgium's KU Leuven university, despite a ban on such advertising. In analyzing more than 4.2 million political and 29.6 million non-political ads on Facebook from more than 215,000 advertisers, the researchers found that 61% more ads are missed than are detected worldwide, and 55% of U.S. detected ads actually were non-political. NYU's Laura Edelson put some of the blame for the high false positive rate on the "rudimentary" way Facebook seems to use keyword models to classify advertising and content. Said Edelson, "Facebook does involve humans in some portion of the ad and content moderation policy, but it's definitely automated first. That approach just has accuracy problems."
|
Computational Comparison of SARS-CoV-2 Delta, Omicron Variant Binding Affinity with ACE2
News-Medical Life Sciences
Sam Hancock
December 8, 2021
Researchers at Malaysia's Management and Science University used computational tools and modeling to probe the omicron variant of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). The researchers used various online tools to calculate omicron's molecular weight, secondary structure, and other properties, and identified 30 mutations in the spike protein, mainly within the receptor-binding domain. The researchers compared omicron's binding affinity with angiotensin-converting enzyme 2 (ACE2) with that of the delta variant, and found omicron exhibits a less disordered viral protein area than both delta and wild-type SARS-CoV-2. This data could be used to help scientists further investigate omicron, and to predict its spread and the harm it could cause.
|
World's Largest Lab-Grown Steak Unveiled by Israeli Firm
The Guardian
Damian Carrington
December 8, 2021
Researchers at the Israeli firm MeaTech 3D said they have grown a 4-ounce steak in a lab, the largest produced thus far. The steak was produced by incorporating live bovine stem cells into "bio-inks" used in a three-dimensional (3D) printer. After it was “printed,” the steak was placed in an incubator to mature, allowing the stem cells to differentiate into fat and muscle cells. MeaTech initially plans to sell the cultured fat as an ingredient for other products, with the goal of producing cultured meat at the same cost as conventional meat. Good Food Institute Europe's Seren Kell said 3D printing “enables companies to create more sophisticated 'whole cut' products which can authentically recreate the taste, texture, and mouthfeel of conventional meat."
|
HPC Helps Scientists Extract Data From Satellite, Drone Imagery
Government Computer News
Stephanie Kanowitz
December 7, 2021
Philipe Dias and Lexie Yang at the U.S. Department of Energy's Oak Ridge National Laboratory (ORNL) are using machine learning and high-performance computing to optimize the extraction of data from satellite and drone imagery. The researchers pulled data on building footprints and roads from satellite imagery by labeling pixels of buildings from remote sensing images. Dias said they map Earth's surface at a resolution of 5 meters (16 feet) per pixel, requiring 100 trillion pixels to be processed; generalization becomes challenging because satellite and drone sensors capture images and domain distribution shifts at varying resolutions.
|
The Digital Car Is Coming. It Might Not Be More Profitable.
The Wall Street Journal
Stephen Wilmot
December 7, 2021
Although investors hope to reap dividends from software-based in-vehicle services, the upgrade cost for automakers so far suggests little profit potential. Multinational automaker Stellantis hopes its "STLA Brain" technology platform will lead to more powerful vehicle software updates, and projects $22.5 billion in annual revenue from software-enabled products and subscriptions by 2030. Meanwhile, General Motors has a plan to make $80 billion from software and other new businesses in 2030, including $50 billion from its Cruise driverless taxi service. Implementing a software-centric business model at a meaningful scale is proving elusive, with revenue only expected to see significant growth in the second half of this decade.
*May Require Paid Registration
|
Google Announces Lawsuit, Technical Action Against Blockchain Botnet Glupteba
ZDNet
Jonathan Greig
December 7, 2021
Google has filed suit against Glupteba, a Russia-based blockchain-backed botnet, to "create legal liability for the botnet operators, and help deter future activity." Google's Threat Analysis Group found the Glupteba botnet has compromised about 1 million Windows devices globally. Google's Royal Hansen and Halimah DeLaine Prado wrote in a blog post that the botnet "is notorious for stealing users' credentials and data, mining cryptocurrencies on infected hosts, and setting up proxies to funnel other people's Internet traffic through infected machines and routers." Google said it also has disrupted Glupteba's command and control infrastructure, although that may only be temporary due to its "sophisticated architecture and the recent actions that its organizers have taken to maintain the botnet, scale its operations, and conduct widespread criminal activity."
|
Australia's AI Cameras Catch Over 270,000 Drivers Using Phones
New Scientist
Alice Klein
December 8, 2021
Artificial intelligence (AI)-equipped cameras have spotted more than 270,000 drivers using phones while driving in New South Wales (NSW), Australia, since the state began issuing fines in March 2020. The cameras capture high-definition images of the front of each passing vehicle, and AI software analyzes them to identify drivers using a handheld cellphone; officers vet images flagged as potentially showing violations before fining those drivers. Transport for NSW's Tara McCarthy said, "We know that mobile phone detection cameras are working and people are getting the message not to use their phone illegally, as we have seen a significant drop in offenses."
|
Scientists Turn to Drones to Learn About Climate Quality
Associated Press
Jennifer McDermott
December 9, 2021
Drone developer Saildrone this week launched three wind- and solar-powered ocean drones from Rhode Island to collect climate quality data along the Gulf Stream over the next six months. Scientists at the University of Rhode Island and the European Center for Medium-Range Weather Forecasts hope to use the information gathered by the drones to enhance medium- and long-range weather forecasting, and to help account for how much man-made carbon dioxide the Gulf Stream can absorb. Saildrone's Susan Ryan said collecting data along the Gulf Stream in harsh winter months is challenging due to strong currents and intense storms, adding that she thinks the data to be collected could help improve the models used to hold countries accountable for their emission-reduction goals.
|
|
