Welcome to the October 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."
|
|
|
Stanford Joins Group Offering Classes to Disadvantaged High-School Students
The Wall Street Journal
Lee Hawkins
October 24, 2021
Stanford University has joined some dozen colleges and universities that offer digitally taught college courses to economically disadvantaged high-schoolers. The school this fall launched a computer science (CS) class in which students watch lessons prerecorded by a Stanford professor, then meet with alumni who serve as teaching assistants. The course has enrolled 225 students across the U.S., and is part of the nonprofit National Education Equity Lab. All City Leadership's Michael Quinones said, "The Stanford alumni talk to the kids about their career and what they do. It's not just the teacher telling them about computer science."
*May Require Paid Registration
|
Self-Driving Race Cars Make History in Indianapolis
Yahoo! News
Thomas Urbain
October 24, 2021
The first Indy Autonomous Challenge in Indianapolis made history as the first racing competition between driverless vehicles, with a team from Germany's Technical University of Munich (TUM) claiming a $1-million prize. TUM’s car beat the EuroRacing team's Dallara IL-15 over two laps, although the Dallara achieved the fastest lap time ever recorded for an autonomous car, at 139 mph (223 kph). Each autonomous car used sensors, cameras, radar, and a global positioning system. The TUM team's Alexander Wischnewski said their car's 135 mph (218 kph) average speed "is not far away from what human drivers do" with the same vehicle.
|
Robot Artist Ai-Da Released by Egyptian Border Guards
BBC News
October 21, 2021
Egyptian border agents have released a British-built robot after detaining it at customs for 10 days due to concerns that it might have been concealing spy tools. The robot, named Ai-Da, creates abstract art using cameras and a robotic arm. The robot's creator, Aidan Meller, offered to remove the robot's modem, which had generated suspicion among the border agents. However, when they raised issues with its cameras, Meller said those could not be removed, as the cameras' recordings are turned into works of art using artificial intelligence algorithms. The robot's work was scheduled to be shown at the Great Pyramid of Giza in the first contemporary art exhibition at the Pyramids in 4,500 years.
|
Twitter Admits Bias in Algorithm for Rightwing Politicians, News Outlets
The Guardian
Dan Milmo
October 22, 2021
Researchers at Twitter found that the social media platform amplifies more tweets from rightwing politicians and news outlets than content from leftwing sources. The study looked at tweets from elected officials and political content from news outlets in the U.K., U.S., Canada, France, Germany, Spain, and Japan. The researchers also found that Twitter's algorithmic "Home" timeline amplified politicians' tweets in general more than the chronological timeline that ranks the most recent tweets first. Twitter's Rumman Chowdhury and Luca Belli wrote in a blog post, "Further root cause analysis is required in order to determine what, if any, changes are required to reduce adverse impacts by our Home timeline algorithm."
|
Russia Is Censoring the Internet, with Coercion and Black Boxes
The New York Times
Adam Satariano; Paul Mozur
October 23, 2021
The Russian government has forced the country's largest telecom and Internet service providers to install black boxes connected to a command center in Moscow, so that it can block, filter, and slow down Websites to control what the public sees online. The initiative was pushed to the forefront this spring when the filtering system was used for the first time to slow Twitter to a crawl. Twitter then agreed to remove dozens of posts the government determined were illegal. Russia's Internet regulator, Roskomnadzor, also has threatened to take down YouTube, Facebook, and Instagram if they do not block certain content on their own. The technology has been installed at 500 locations of telecom operators, and is expected to rise to more than 1,000 by next year.
*May Require Paid Registration
|
Quantum-Encrypted Information Transmitted Over Fiber More Than 600 Km Long
Optica
October 21, 2021
Researchers at the University of Leeds and Toshiba Europe in the U.K. established secure quantum communication over 605 kilometers (375 miles) of fiber through a new signal stabilization method. The researchers used the twin-field quantum key distribution protocol, which enables two geographically separated users to establish a common secret bit-string by sharing photons, which are usually transmitted over an optical fiber. The stabilization technique utilizes two optical reference signals at different wavelengths to minimize phase fluctuations over long distances. The researchers demonstrated that this method could support repeater-like performance while accommodating losses outside the traditional limit of 100 decibels over a 605-kilometer-long quantum channel. Toshiba Europe's Andrew Shields said, "This will allow us to build national- and continental-scale fiber networks connecting major metropolitan areas."
|
Technology Streamlines Development of Gene Therapy for Genetic Blinding Disorders
News-Medical Life Sciences
Emily Henderson
October 19, 2021
Researchers at the University of Pittsburgh School of Medicine (UPSM) have developed the novel scAAVengr computational platform to identify adeno-associated virus (AAV) vectors that could deliver ultra-precise and efficient retinal gene therapies for blinding disorders. The platform employs single-cell RNA sequencing to quickly screen numerous AAVs for those best-suited for gene therapy, shortening the process from years to months. The researchers found scAAVengr also performs equally well in identifying AAVs that target the brain, heart, and liver. UPSM's Leah Byrne said the platform "will allow us to translate emergent therapies that are already working for some patients into the clinic much more rapidly."
|
Supercomputer Simulations Reveal How Protein Crowding in Cells Impacts Interactions
RIKEN
October 21, 2021
Researchers from Japan's RIKEN research organization and Michigan State University used supercomputer simulations to determine how crowding in cells impacts drug binding to a protein target, potentially helping explain why some drugs work in theory but not in practice. The researchers factored in cells' normally crowded environments to improve the accuracy of these simulations. They developed the GENESIS software program and performed microsecond-scale simulations of the interaction of the c-Src kinase enzyme with the PP1 inhibitor amid different concentrations of bovine serum albumin (BSA). The researchers determined that BSA crowding limited the amount of PP1 that could reach the enzyme, and covered some binding sites on the enzyme and changed its shape. Laboratory tests validated the results of the simulations, finding that increased BSA crowding reduced PP1's efficacy in inhibiting c-Src kinase.
|
Meet Snitch: The Small, Agile RISC-V Processor
IEEE Spectrum
Michelle Hampson
October 21, 2021
Tests by researchers at the Swiss Federal Institute of Technology, Zurich (ETH Zurich) suggest the new RISC-V-based Snitch processor used in parallel can be 3.5 times more energy-efficient than other benchmark designs, and as much as six times faster. Snitch can perform most basic instructions within a single clock cycle, and execute longer latency instructions without stalling and awaiting their completion. ETH Zurich's Florian Zaruba said, "This leads to a very compact and high-performance design compared to conventional processors that achieve high performance." Zaruba said while programming Snitch is more complicated than standard processors, he believes its energy efficiency will make it appealing for next-generation computers.
|
Researchers Use Computer Modeling to Predict Patient Tumor Responses to Radiation Therapy
Moffitt Cancer Center
October 18, 2021
Researchers at the Moffitt Cancer Center in Tampa, FL, used computers to model the effects of interactions between cancer cells and immune cells in tumors, along with their responses to radiation. The team incorporated factors including cell proliferation, migration, cell death, immune cell motility, cell-to-cell interactions, and the cytotoxic impact of radiation within the model. They determined tumor cells will either evade immune predation or be destroyed by the immune system, based on the numbers of immune effector and suppressor cells present. The model predicted that about half of non-small cell lung cancer patients treated with radiation would have durable tumor control with a lower radiation dose, while 40% would require a higher dose.
|
Artificial Networks Learn to Smell Like the Brain
MIT News
Jennifer Michalowski
October 18, 2021
A team of researchers from the Massachusetts Institute of Technology (MIT) and Columbia University found a machine learning model can train itself to smell by building an artificial neural network that mimics the brain's odor-processing olfactory circuits. The researchers used the fruit fly's olfactory system as a template, building an artificial network comprised of an input layer, a compression layer, and an expansion layer; links between neurons would be rewired as the model learned to classify smells. The network self-organized in minutes into a structure closely resembling the fly brain's olfactory network. MIT's Guangyu Robert Yang said, "By showing that we can match the architecture [of the biological system] very precisely, I think that gives more confidence that these neural networks can continue to be useful tools for modeling the brain."
|
AI Can Help Diagnose Mental Health Disorders Where Access to Care Is Lacking
UPI
Brian P. Dunleavy
October 15, 2021
At the French Institute for Research in Computer Science and Automation, researchers have developed a machine learning approach that accurately diagnosed mental health disorders based on "proxy measures" for mental health about 90% of the time. The researchers based their models on sociodemographic and behavioral data on more than 500,000 adults from the U.K. Biobank database, along with magnetic resonance imaging (MRI) brain scans of more than 10,000 participants. In tests on data from 7,000 U.K. Biobank participants, about half of whom had a diagnosed mental health disorder, the researchers found a combination of sociodemographic information and brain imaging to assess brain age was the most accurate proxy measure for mental health. Said lead researcher Denis Engemann, "We showed that useful proxy measures can be derived from other inputs than brain images."
|
|
