Welcome to the February 23, 2022, 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."

The Internal Revenue Service Building in Washington, D.C. IRS Allows Taxpayers to Forgo Facial Recognition Amid Blowback
The New York Times
Alan Rappeport
February 21, 2022


The Internal Revenue Service (IRS) has walked back a requirement for taxpayers to use facial recognition to access their online accounts amid negative feedback, and pledged to switch to a completely different identity verification system next year. This follows the agency's decision to "transition away" from using the ID.me third-party service to authenticate people creating accounts by using facial recognition to confirm their identities. Activists and lawmakers said using video selfies in this capacity was an invasion of privacy. The IRS will permit taxpayers to opt out of ID.me's biometric authentication in favor of live, virtual interviews with company representatives. The agency said it now plans to use Login.gov, which millions of Americans already use to confirm their identities for access to certain federal Websites, for account authentication.

Full Article
*May Require Paid Registration
Over 40% of West Coast Tech Job Listings Are Outside West Coast
Bloomberg
Reade Pickert
February 22, 2022


An analysis by the Conference Board indicates West Coast technology companies are listing 43% of their white-collar jobs in states outside that region. The top state destination for these jobs is Texas, while the leading metro areas are Washington, D.C., and New York City. The inland migration of tech hubs ramped up dramatically as companies went fully or partially remote during the pandemic, while expanding companies also are pursuing local talent. Texas has captured 7.3% of job postings from West Coast tech firms, and Austin has nearly 4% of the listings reviewed by the Conference Board. Virginia, Georgia, and North Carolina also score high in drawing tech companies.

Full Article

Mother Lizzie Fulton with baby Pippa and lead congenital cardiac surgeon Nicola Viola. 3D-Printed Heart Helps Save Life of Baby Girl with Rare Condition
University Hospital Southampton (U.K.)
February 22, 2022


U.K. surgeons saved the life of a baby girl with a rare heart condition using a three-dimensionally (3D)-printed model of her heart. The team at University Hospital Southampton (UHS) based the model of Pippa Fulton's heart on a computed tomography scan, which incorporated her multiple heart defects and allowed the surgeons to rehearse the surgical procedure before performing it. UHS cardiac surgeon Nicola Viola said, "In any major surgery like this, there is always the risk of serious complications, and making the 3D model of the heart gave us the chance to better understand and practice the procedure." The successful operation allowed the infant to go home after spending her first six months of life in the hospital.

Full Article

If every Ford Thunderbird in a machine learning model’s training dataset is shown from the front, an image of a Thunderbird shot from the side could be misclassified. Can ML Models Overcome Biased Datasets?
MIT News
Adam Zewe
February 21, 2022


Researchers at the Massachusetts Institute of Technology (MIT), Harvard University, and Japanese information and communications technology company Fujitsu investigated how machine learning (ML) models can surmount dataset bias. They applied neuroscience principles to determine whether an artificial neural network can learn to identify previously unseen objects. The results of the research indicate diverse training data can shape the network's ability to overcome bias, but can concurrently cause its performance to deteriorate. "When the neural network gets better at recognizing new things it hasn't seen, then it will become harder for it to recognize things it has already seen," explained MIT's Xavier Boix. The training method itself and the types of neurons that emerge during the training process also can influence the network's handling of biased datasets. Said Boix, "We need to be very careful about how we design datasets in the first place."

Full Article
AT&T First Top Network to Shut Down 3G
Independent (U.K.)
Io Dodds
February 22, 2022


AT&T on Feb. 22 became the first U.S. carrier to shutter its 3G mobile network, with T-Mobile to follow in its footsteps this summer, and Verizon in December. This means older mobile phones running on these carriers’ 3G networks will no longer work, in addition to millions of home security systems, car safety systems, GPS trackers, medical monitors, and even prisoner ankle tags that require 3G services. These Internet of Things devices are harder to replace than smartphones, and some customers will have to pay hefty prices to upgrade or continue using their old devices, but without key features. General Motors' OnStar is among the crash alert and roadside assistance systems expected to stop working unless vehicles receive software updates or physical replacements, depending on the model. Some companies are offering free software updates, but Tesla, for example, will charge customers $200 for new modems in any Model S vehicles built before June 2015.

Full Article
China's Driverless Car Dreams Troubled by U.S. Chip Dependency
Financial Times
Edward White
February 20, 2022


China's push for technologically independent self-driving cars is hindered by the vehicles' reliance on foreign-made semiconductors, mainly designed by U.S. companies like Nvidia. Chinese chip designers are years behind their American counterparts; Bain & Company China semiconductor expert Velu Sinha said Nvidia had the advantage of "a decade's head start" due to its successful graphical processing unit technology. Nvidia said its chips are a key component in at least 18 Chinese companies' autonomous driving plans. An anonymous China-based semiconductor consultant said indigenous automotive chip development faces "fundamental problems," including cost-inhibited scalability and incumbent chipmakers reworking technologies to fuel booms in cellphones, desktops, and datacenters in recent decades. Moreover, local manufacturing is unavailable for the most cutting-edge chips used by driverless platforms.

Full Article
*May Require Paid Registration

The Nos.e analyzed and identified different whiskey samples at CEBIT 2019. E-Nose Sniffs Out the Good Whiskey
IEEE Spectrum
Michelle Hampson
February 18, 2022


New research describes an electronic nose (e-nose) that can analyze whiskies and recognize a whiskey’s brand with over 95% accuracy after a single whiff. Researchers at Australia's University of Technology Sydney based NOS.E on an e-nose originally developed to detect illegal animal parts sold on the black market. NOS.E has a vial to insert the whiskey sample, and a gas sensor chamber to contain the whiskey's scent; the chamber reads the various odors and transmits the data to a computer, then key scent features are extracted and analyzed by machine learning algorithms to identify brand, region, and style of whiskey. The researchers compared NOS.E's analysis of a half-dozen whiskies to that of a state-of-the-art gas chromatography device, and found both methods yielded similar accuracy.

Full Article

Plasma inside the TCV fusion reactor in Switzerland. DeepMind Uses AI to Control Plasma Inside Tokamak Fusion Reactor
New Scientist
Matthew Sparkes
February 16, 2022


Scientists at U.K. artificial intelligence (AI) company DeepMind and the Swiss Federal Institute of Technology in Lausanne (EPFL) have developed a neural network that can control the magnetic fields within EPFL's Variable Configuration Tokamak (TCV) fusion reactor. This eliminated the need to monitor the TCV's interior with 19 magnetic coils, each controlled by an individual algorithm; the network automatically learns which voltages must be supplied to best contain plasma within the reactor. The researchers trained the AI on a digital model of the reactor and it successfully contained the plasma for about two seconds, which is approaching the reactor's threshold. The algorithm also was able to configure and move the plasma within the reactor, which may help to improve efficiency or stability in new reactors.

Full Article
Social Media Echo Chambers Spread Vaccine Misinformation
Technical University of Denmark
Hanne Kokkegard
February 16, 2022


Technical University of Denmark (DTU) researchers analyzed 60 billion tweets posted before the pandemic to find that social media is spreading misinformation about vaccines. The researchers developed deep learning and natural language processing techniques to train a computer to identify the vaccine view expressed in a given tweet, and were able to identify users who consistently were for or against vaccines, and from which sources the profiles shared their vaccine information. DTU's Bjarke Mønsted said, "Where vaccine supporters often refer to news media and science sites when sharing knowledge about vaccines on Twitter, we can see that profiles belonging to anti-vaccine profiles far more often share links to YouTube videos and to sites that are known to spread fake news and conspiracy theories." The results confirmed that social media is an echo chamber that inhibits vaccine advocates and opponents from encountering each other's views online.

Full Article
Research Could Make Quantum Tech Viable in Outer Space
Tom's Hardware
Francisco Pires
February 14, 2022


Researchers at the U.K.'s University of Sussex have developed a remote monitoring and control system template for quantum devices and experiments that could be practical for application in space or other extreme environments. The project specifies control and key experimental parameters, like vacuum chamber pressure, laser beam power, or conducted resistances, needed in quantum computing environments. It also highlights how to mitigate negative environmental factors by implementing automated control sequences coupled with evolutionary algorithms and machine learning protocols. Sussex's Peter Krüger suggests this "has far-reaching implications that could pave the way to new smart technologies utilizing AI [artificial intelligence]/human collaboration."

Full Article
Ancient Dwarf Galaxy Reconstructed with MilkyWay@Home Volunteer Computer
Rensselaer Polytechnic Institute News
Mary L. Martialay
February 17, 2022


A multi-institutional team of researchers from Rensselaer Polytechnic Institute (RPI), the California Institute of Technology, and Canada's Queen's University reconstructed an ancient dwarf galaxy using the MilkyWay@Home distributed supercomputer. The supercomputer harnesses 1.5 petaflops of crowdsourced computer power donated by volunteers across the globe. The researchers estimated the total mass of the original galaxy whose stars now make up the Orphan-Chenab Stream as 2x107 times the mass of our sun. They determined that ordinary matter accounts for only slightly more than 1% of that mass, with dark matter making up the remainder. RPI's Heidi Newberg said, "By looking at the current speeds of stars along a tidal stream, and knowing they all used to be in about the same place and moving at the same speed, we can figure out how much the gravity changes along that stream. And that will tell us where the dark matter is in the Milky Way."

Full Article

“Understanding these processes can provide solutions in many areas,” said URV researcher Àlex Arenas. Researchers Discover How to Predict Degradation of Neural Network
URV Activ@
February 16, 2022


Researchers at Spain's Universitat Rovira i Virgili (URV) have identified the theoretical underpinnings for predicting how neural networks will function and degrade over time. These findings indicate how much damage a system can endure before it will completely degrade and lose its functionality, known as the phase transition of percolation degradation. URV's Alex Arenas said, "We have been able to find this transition and we have also been able to calculate the homeostatic response [i.e., the ability to find alternatives and continue functioning] of the network." He added that a set of mathematical tools "that can be very useful not only in neuroscience but in any type of network" is now available to the scientific community.

Full Article
'Underground Maps' Segment Cities Using Fashion, AI
Cornell Chronicle
Tom Fleischman
February 15, 2022


Artificial intelligence (AI) developed by scientists at Cornell University, the University of Texas at Austin, and Facebook can automatically segment cities into "underground maps" -based commonalities of interests. The work builds on the Cornell team's development of the GeoStyle tool, which can discover geospatial events and predict fashion trends. The researchers applied a fashion recognition algorithm to photos geolocated from 37 large cities, in order to detect clothing styles, then mixes of those styles within a given radius. They then used AI to identify spatially and stylistically coherent segments within a city. Two human-centered benchmark programs, HoodMaps and OpenStreetMaps, were used with resident surveys to calculate the technique's accuracy; in every instance, it described the “sense” of a neighborhood better than existing methods.

Full Article
Peered Review Resources for Engaging Students
 
ACM Special Interest Groups
 

Association for Computing Machinery

1601 Broadway, 10th Floor
New York, NY 10019-7434
1-800-342-6626
(U.S./Canada)



ACM Media Sales

If you are interested in advertising in ACM TechNews or other ACM publications, please contact ACM Media Sales or (212) 626-0686, or visit ACM Media for more information.

To submit feedback about ACM TechNews, contact: [email protected]