Welcome to the November 24, 2021 edition of ACM TechNews, providing timely information for IT professionals three times a week.
Please note: Please note: In observance of the U.S. Thanksgiving holiday, TechNews will not be published on Friday, November 26. Publication will resume Monday, November 29. 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."
|
|
|
For Rules in Technology, the Challenge Is to Balance Code, Law
The New York Times
Ephrat Livni
November 23, 2021
The idea that "code is law," observed in a 1999 book by Harvard University's Lawrence Lessig, has been embraced by the cryptocurrency industry, with some firms contending code can be a better arbitrator than traditional regulators. However, issues have emerged in the decentralized finance (DeFi) sector, with hackers this summer overriding smart contract instructions to steal $600 million from the Poly Network, which lets users transfer cryptocurrencies across blockchain networks. DeFi platforms typically are established as Decentralized Autonomous Organizations governed democratically by a community of users who vote with crypto tokens. Said Lessig, "We need a more sophisticated approach, with technologists and lawyers sitting next to behavioral psychologists and economists," to define parameters to code social values into programs.
*May Require Paid Registration
|
Kids, Teens Believe Girls Not Interested in CS: Study
University of Washington News
Kim Eckart
November 22, 2021
Researchers at the universities of Houston and Washington found that the majority of children surveyed believe girls are less interested than boys in computer science (CS) and engineering. The researchers gauged the beliefs of a racially diverse sample of children and teenagers in grades 1 through 12. Two studies surveying more than 2,200 children and teens found that an estimated 51% of respondents thought girls to be less interested than boys in CS, while 63% said girls are less interested in engineering. In experiments, girls were found to be less interested in a CS activity when told boys were more interested in it than girls, versus one in which they were told boys and girls were equally interested. Said the University of Washington's Andrew Meltzoff, "The mere presence of the stereotype influenced kids in dramatic ways. This brought home to us the pernicious effect of stereotypes on children and teens."
|
Deep Learning Method Adds 301 Planets to Kepler's Total Count
Jet Propulsion Laboratory
November 22, 2021
Scientists have added 301 validated exoplanets to the number of confirmed exoplanets through the use of the new ExoMiner deep learning neural network to analyze data from the U.S. National Aeronautics and Space Administration (NASA) Kepler and K2 missions. ExoMiner taps NASA's Pleiades supercomputer to identify new exoplanets within datasets, and can differentiate real exoplanets from false positives. Jon Jenkins at NASA's Ames Research Center said ExoMiner is not a "black box" that conceals its reasoning, so researchers know why it confirms or rejects exoplanets. Said Ames' Hamed Valizadegan, "ExoMiner is highly accurate and in some ways more reliable than both existing machine classifiers and the human experts it's meant to emulate because of the biases that come with human labeling."
|
How to Find Hidden Spy Cameras with a Smartphone
Help Net Security
Zeljka Zorz
November 23, 2021
Scientists at the National University of Singapore and South Korea's Yonsei University developed a smartphone application that can find tiny spy cameras concealed in everyday objects, using smartphones' time-of-flight (ToF) sensor. The researchers said the Laser-Assisted Photography Detection (LAPD) app spots hidden cameras better than commercial camera detectors, and much better than the human eye. The app, which works on any smartphone handset equipped with a time of flight (ToF) sensor, can only scan a single object at a time, and requires about a minutes to scan that object. The researchers said the app could be made more accurate by taking advantage of the handset’s flashlight and RGB cameras.
|
NC State to Add 4,000 Engineering, CS Students to Meet Worker Demand
WRAL Tech Wire
Mike Kulikowski
November 23, 2021
North Carolina State University (NC State) plans to add approximately 4,000 engineering and computer science (CS) students over the next five years, to meet the state's ever-growing demand for technology professionals. The Engineering North Carolina's Future initiative will provide the school $20 million in funding over the next two years to expand its faculty and support additional staff to accommodate the larger student body. Another $30 million allocated by North Carolina’s legislature will be used for facility upgrades to support the additional science, technology, engineering, and math (STEM) learners. Said NC State's Louis Martin-Vega, "This initiative will allow us to grow our faculty and teaching and research infrastructure to continue to provide access to an excellent engineering education for more students at the undergraduate and graduate level."
|
COVID Gets Airborne
UC San Diego News Center
Michelle Franklin
November 22, 2021
A team of U.S. and international researchers has developed the first model of the COVID-19 delta variant inside an aerosol, which the University of California, San Diego (UCSD)'s Rommie Amaro said "could change the way we view airborne diseases." Amaro and colleagues developed a framework to build, simulate, and analyze aerosol models, and used it to model the SARS-CoV-2 virus and its mechanism for infecting cells through the glycan membrane on its spike proteins. The researchers used the Summit supercomputer at the Department of Energy's Oak Ridge National Laboratory to simulate 1-billion-atom aerosol models. Said Amaro, "There is no experimental tool, no microscope that allows people to see the particles in this much detail, but this new computational microscope allows us to see what happens to the virus—how it moves, how it stays infectious during flight."
|
Research Team Advances Brain-Inspired Computing
USC Viterbi News
Amy Blumenthal
November 21, 2021
Researchers at the University of Southern California, the U.S. Army Research Laboratory, the University of Florida, and the Georgia Institute of Technology have unveiled advancements in brain-inspired computing that tap randomness to augment performing optimization. The researchers, aiming to mimic neurons as much as possible, developed a circuit to address combinatorial optimization problems, a hetero-memristor that dynamically tunes randomness features through a memory-computation combination by adding a third electrical terminal and modulating its voltage. "The randomness introduced in the new device demonstrated in this work can prevent it from getting stuck at a not-so-viable solution, and instead continue to search until it finds a close-to-optimal result," according to the researchers.
|
Sundance 'Biodigital' Film Festival Will Try to Bridge Gap Between VR, Reality
Engadget
Devindra Hardawar
November 22, 2021
The 2022 Sundance Film Festival in Park City, UT, plans to expand its virtual interactive venue, The Spaceship, as part of its New Frontier exhibition. The event will also host a "Biodigital Bridge" which Sundance programmer Shari Frilot called a human-scale screen-enabling interaction between physical and digital attendees. Created in partnership with immersive studio Active Theory, the bridge will support basic chat, but will mainly allow festival-goers to see how other attendees experience Sundance. Frilot said the virtual platform "points to the vital landscape of how cinema and storytelling is going to manifest [going forward]. We're here to not only contribute to it, but to meet it and support it."
|
Deep Learning Reveals How Proteins Interact
UW Medicine
Ian Haydon
November 19, 2021
A multi-institutional group of researchers has constructed three-dimensional (3D) models of protein interactions in eukaryotic cells using evolutionary analysis and deep learning. The researchers analyzed known gene sequences in yeast to map the interactions that create protein complexes, and statistically identified gene pairs that naturally acquire mutations in a connected manner. They also employed UW Medicine's RoseTTAFold and DeepMind's AlphaFold deep learning software to simulate the interacting proteins' 3D shapes. UW Medicine's David Baker said the research yielded insights into protein interactions for nearly all core eukaryotic cellular processes, including "over 100 interactions that have never been seen before."
|
Imaging Technology May Reduce Need for Skin Biopsies
UCLA Samueli School of Engineering
November 18, 2021
University of California, Los Angeles (UCLA) researchers used virtual histology technology to assess images of skin tissues, which could reduce the need for skin biopsies to detect skin cancers. The researchers invented a deep learning framework to convert images of intact skin obtained by reflectance confocal microscopy (RCM) into a format that dermatologists and pathologists can use. The convolutional neural network was trained to rapidly render RCM images of unstained skin as virtually stained three-dimensional images. Said UCLA's Aydogan Ozcan, "This approach may allow diagnosticians to see the overall histological features of intact skin without invasive skin biopsies or the time-consuming work of chemical processing and labeling of tissue."
|
Wikipedia Tests AI for Spotting Contradictory Claims in Articles
New Scientist
Matthew Sparkes
November 19, 2021
Researchers at Taiwan's National Cheng Kung University, in conjunction with the Wikimedia Foundation, have developed artificial intelligence technology which they say can identify contradictory claims in Wikipedia articles and flag them for human review. The researchers found 2,321 contradiction warnings in all English Wikipedia articles posted by March 2020. They used 80% of 1,105 examples of contradictions and solutions by human editors to train the neural network to detect contradictions on its own. The remaining 20% of the data then was used to test the neural network, which was found to have an accuracy rate of up to 65%.
*May Require Paid Registration
|
Croatia's Seed-Scattering Drones Replant Forests Hurt by Fire
Reuters
Antonio Bronic; Igor Ilic
November 17, 2021
Flying drones are dispersing seeds in Croatia to replant hard-to-reach fire-damaged forests. Goran Ladisic of the Magic Forest company, which is running the project, said tests have showed more than 40% of the seeds dispersed by the drones have taken root. Drone flight controller Ivan Vidakovic said four or five drones can cover approximately 10 hectares (24 acres) in eight hours. The goal is to replant the area with native trees. Magic Forest said it has been in touch with potential partners in Ukraine, Bosnia, Montenegro, Austria, and California about deploying the service in those areas next year.
|
|
