Welcome to the August 21, 2023, 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."
|
|
|
Adobe Co-founder John Warnock Dies at 82
KSL TV
Matt Brooks; Shelby Lofton
August 20, 2023
John Warnock, co-founder of computer software company Adobe, has passed away at age 82. While working on his doctorate in electrical engineering/computer science, Warnock and a colleague wrote the University of Utah's first computer-based registration program. At Xerox, Warnock helped create a system for interactive graphics, which he and colleague Charles Geschke further developed by co-founding Adobe in 1982. A 2020 statement by the University of Utah said, "The duo's innovations included scalable type, computer graphics, and printing processes" that transformed desktop publishing. Richard B. Brown, dean of the university’s College of Engineering, said Warnock's creation of the PDF "made desktop publishing practical" through the concept of software-generated fonts.
|
Trinity-IBM Team Simulates Super Diffusion on Quantum Computer
Trinity College Dublin (Ireland)
August 17, 2023
Quantum physicists at Ireland's Trinity College Dublin (TCD) and IBM Dublin used a quantum computer to model super diffusion in a system of interacting quantum particles. The Dublin-based researchers remotely programmed the 27-quantum-bit computer in New York. TCD's John Goold said the physicists simulated the Heisenberg spin chain, "and we were particularly interested in the long-time behavior of how spin excitations are transported across the system." The experiment marks the initial step in performing challenging quantum transport calculations on quantum computers.
|
Algorithm Reveals High Mutations Rates in Complex Genomic Regions
UC San Diego Today
Josh Baxt
August 16, 2023
The UniAligner algorithm developed by the University of California, San Diego's Pavel Pevzner and Andrey Bzikadze can detect mutations in biomedically critical genomic regions. UniAligner prioritizes rare substring matches to address classical sequence alignment approaches' tendency to generate millions of spurious matches between two sequences due to the repetitive nature of complex regions. The algorithm aligned human centromeres to reveal extremely high mutation rates, implying that centromeres may be some of the most rapidly developing regions in the human genome. UniAligner's speed also exceeds that of state-of-the-art sequence comparison algorithms.
|
Google Tests AI Assistant That Offers Life Advice
The New York Times
Nico Grant
August 16, 2023
Google is testing generative artificial intelligence (AI) technology programmed to serve as a life coach following the merger of its U.K.-based DeepMind research laboratory with its Brain AI development team in Silicon Valley. Materials reviewed by The New York Times indicate DeepMind has been developing generative AI to perform at least 21 personal and professional tasks, including providing life advice, ideas, planning instructions, and tutoring tips. Anonymous sources said worker teams organized by DeepMind contractor Scale AI are evaluating the AI assistant's capabilities, including its ability to answer sensitive questions about life challenges. According to Google, the program's creation feature could offer situation-based suggestions or recommendations, while its tutoring function can teach new skills or enhance current ones, and its planning capability can outline financial budgets and meal and workout plans.
*May Require Paid Registration
|
Government Regulation Can Curb Social Media Dangers Effectively
University of Technology Sydney (Australia)
August 14, 2023
Researchers from Australia's University of Technology Sydney (UTS) and Switzerland's École Polytechnique Fédérale de Lausanne (EPFL) have demonstrated that government policing of social media content can reduce its potential for harm. The researchers analyzed the dynamics of content dissemination by simulating the spread of information. They considered potential harm—the number of damaging ‘offspring’ generated by a single post—and content half-life—the time it takes to produce half of all offspring. UTS' Marian-Andrei Rizoiu and EPFL's Philipp J. Schneider found government-mandated external social media content moderation would likely limit harm, even with a 24-hour turnaround time.
|
Brazilian Researchers Propose Quantitative Diagnostic Method for Autism Spectrum Disorder
News-Medical Life Sciences
Danielle Ellis
August 16, 2023
Brazilian researchers have proposed a method for identifying autism spectrum disorder (ASD) through quantitative diagnostics. The researchers applied machine learning (ML) techniques to data on the brains of 500 people, including 242 diagnosed with ASD, acquired via functional magnetic resonance imaging (fMRI) and electroencephalogram. An ML algorithm was able to ascertain ASD-associated brain changes with more than 95% accuracy when fed patients' brain maps. Data analysis via fMRI underscored alterations in brain regions connected to cognitive, emotional, learning, and memory processes, while ASD patients' cortical networks exhibited more segregation, less information distribution, and less connectivity relative to controls. Said the University of São Paulo's Francisco Rodrigues, "Our study is an important step in the development of novel methodologies that can help us obtain a deeper understanding of this neurodivergence."
|
ChatGPT Leans Liberal
The Washington Post
Gerrit De Vynck
August 16, 2023
Research by scientists at the U.K.'s University of East Anglia suggests OpenAI's ChatGPT has a liberal slant. The researchers asked the chatbot to answer questions on political convictions as it assumed liberal supporters in the U.S., the U.K., and Brazil might answer them, then asked it to answer the same questions without prompting. The outcomes indicated "significant and systematic political bias toward the Democrats in the U.S., [leftist president] Lula in Brazil, and the Labor Party in the U.K.," the researchers wrote. These findings add to a growing body of evidence showing chatbots are rife with assumptions, beliefs, and stereotypes that were embedded in their training data.
*May Require Paid Registration
|
Researchers Demo Fake Airplane Mode Exploit That Tricks iPhone Users
Computer Weekly
Alex Scroxton
August 17, 2023
Jamf Threat Labs researchers demonstrated an exploit chain that allows attackers to use an artificial ‘airplane mode’ to remain connected to exposed devices that users believe are offline. The researchers created a fake airplane mode by identifying a specific string in the device's console log, "#N User airplane mode preference changing from kFalse to KTrue," accessing the device's code, and replacing the function with an empty or ‘do nothing’ function. They also accessed the user interface to add a small piece of code to dim the mobile connectivity icon and highlight the airplane mode icon, then exploited the CommCentre to block mobile data access for certain apps so the user received a "turn off airplane mode" notification. The researchers believe the technique is most likely to be used in a targeted attack.
|
Humanoid Robot Can Pilot an Airplane Better Than a Human
EuroNews
Roselyne Min
August 15, 2023
Engineers and researchers at South Korea's Korea Advanced Institute of Science & Technology are developing a humanoid robot that can fly airplanes without requiring cockpit modifications. "Pibot" can operate the same flight instruments that humans use, even amid severe vibration through high-precision control technology. The robot can monitor the current state of the aircraft using external cameras and manipulate switches on the control panel using its internal cameras. Pibot also can memorize complex manuals in natural language, including all of the Jeppesen aeronautical navigation charts. The researchers expect Pibot, when completed in 2026, to pilot airplanes without error and overtake humans' reaction time in emergency situations thanks to advances in large language models.
|
Self-Driving Cars Can Slow Traffic
North Carolina State University
Matt Shipman
August 14, 2023
A study by North Carolina State University (NC State) researchers found that unconnected automated vehicles (AVs) can slow travel time through intersections. The researchers simulated traffic conditions using a computational model that took into account AVs, human-driven vehicles, connected vehicles (CVs) driven by humans but connected to traffic light control systems and which share data with other CVs, and connected automated vehicles (CAVs). They ran 57 traffic simulations, looking at the impact of different combinations of these vehicles and found the presence of more CVs and CAVs improved traffic flow at intersections, but higher percentages of unconnected AVs reduced travel time at intersections. Explained NC State's Ali Hajbabaie, "This is because those AVs are programmed to drive conservatively in order to reduce the risk of collisions."
|
Over 100K Hacking Forums Accounts Exposed by Info-Stealing Malware
BleepingComputer
Ionut Ilascu
August 14, 2023
Researchers at threat intelligence firm Hudson Rock identified 120,000 infected systems containing more than 140,000 credentials for cybercrime forums. Using data from info-stealer logs, the researchers found that around 100,000 of those compromised computers belonged to hackers, most likely less-skilled ones. The info-stealer logs also detailed additional credentials (emails and usernames), auto-filled personal data (names, addresses, phone numbers), and system information (computer names and IP addresses). Of the compromised users, over 57,000 had accounts with the cybercrime community Nulled. The researchers found BreachForums users had the strongest passwords, and credentials for cybercrime forums were slightly stronger than government website logins. Three info-stealers, RedLine, Raccoon, and Azorult, accounted for a majority of infections.
|
Israeli Co. Uses AI to Save Bees
The Jerusalem Post (Israel)
Zachy Hennessey
August 10, 2023
Israeli agricultural technology company BeeHero has introduced the Pollination Insight Platform to monitor pollinating bees, enhance pollination efficiency, and improve crop production. The platform uses in-field sensors to track pollinator activity for various crops in real time, for inclusion in the world's largest dataset on bee behavior. Artificial intelligence-driven analytics convert the data into insights that inform decision-making by growers that can augment crop yields. BeeHero developed the platform in collaboration with Israel-based global vegetable seeds company Hazera, and has deployed it in the U.S., Europe, and Israel. Hazera's Avi Gabai said, "The introduction of this in-field sensing solution marks a significant milestone in the agricultural industry's ongoing efforts to address the challenges posed by declining bee populations."
|
|
