Welcome to the April 14, 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."
|
|
|
Company Debuts 'World's First Smart Gun' with Fingerprint Unlocking System
CBS News
Annie Gimbel
April 13, 2023
U.S.-based Biofire Technologies says it has developed what it is calling the world's first biometric smart gun, a 9-mm firearm secured by fingerprint and three-dimensional infrared (IR) facial recognition. IR sensors in the grip keep the gun armed while an authorized user is holding it, making continuous biometrics authentication unnecessary. A rechargeable lithium-ion battery powers the gun; Biofire said the battery lasts for several months with average use and can maintain continuous fire for several hours. Biofire founder Kai Kloepfer said, "We've applied high-precision engineering principles to make a meaningful impact on preventable firearm deaths among children."
|
One-Size-Fits-All Content Moderation Fails the Global South
Cornell Chronicle
Patricia Waldron
April 13, 2023
Cornell University's Farhana Shahid and Aditya Vashistha found that content moderation systems frequently punish social media users in the Global South because they are based on Western social standards. The researchers interviewed Bangladeshis penalized for violating Facebook's community standards, who said the content moderation system often misread their posts, stripped content that was acceptable in their culture, and made unfair, vague, and arbitrary decisions. Shahid said current content moderation policies reinforce longstanding power imbalances dating from colonialism. Shahid and Vashistha suggested social media platforms embed local values, laws, and norms into their moderation systems through consultation with community representatives.
|
Sharp Image of Iconic Black Hole
Scientific American
Meghan Bartels
April 13, 2023
A multi-institutional team of researchers used machine learning to produce a sharper image of the supermassive black hole at the heart of the galaxy M87. The Event Horizon Telescope (EHT) Collaboration generated the original image of the black hole in 2019 with an artificial intelligence-enhanced method; the Institute for Advanced Study's Lia Medeiros and colleagues applied the principal-component interferometric modeling (PRIMO) algorithm to sharpen the image. The PRIMO machine learning approach was trained on simulated black holes with varying properties to formulate rules stemming from the expected appearance of black holes. The researchers processed initial EHT data using PRIMO, which yielded a sharper image of a narrower ring surrounding a truly black core.
|
RoboCop? No, RoboDog: Robotic Dog Rejoins New York Police
Associated Press
Karen Matthews
April 12, 2023
New high-tech policing devices rolled out in New York City this week include Digidog, a 70-pound robotic dog first used by the city in 2020, but pulled amid criticism it was "creepy" and "dystopian." New York City Mayor Eric Adams said the remote-controlled canine will be used beginning this summer in hostage standoffs and other situations in which sending police would be too risky. City law enforcement also will begin using the StarChase tracking system, which provides users a GPS tag that can be attached to cars so their locations can be tracked if they are stolen, and the Autonomous Security Robot, a cone-shaped security device that has been used in shopping centers and other locations.
|
Entering an Era of 3D Printing Even for DNA, Proteins
Pohang University of Science and Technology (South Korea)
April 5, 2023
A technique developed by researchers at South Korea's Pohang University of Science and Technology (POSTECH) preserves biopolymers' folding structure and molecular function through three-dimensional (3D) printing. The technique allows for the production of 3D biopolymeric architectures with precisely controlled sizes and geometries at submicron resolution with full mechanical stability and functional integrity. The printing process, performed at room temperature without additives, involves sequentially confining, evaporating, and solidifying a biopolymer-containing solution. POSTECH's Jung Ho Je said the research "holds the potential to expand to the printing of various materials with diverse optical and electrical properties, including complex materials such as quantum dots and carbon nanotubes."
|
AI-Descartes: A Scientific Renaissance
SciTechDaily
April 12, 2023
AI-Descartes, an "AI scientist" developed by researchers at IBM Research, Samsung AI, and the University of Maryland, Baltimore County, used logical reasoning and symbolic regression to reproduce Nobel Prize-winning work by U.S. chemist Irving Langmuir on the behavior of gas molecules adhering to a solid surface. It also "rediscovered" Kepler’s third law of planetary motion and recreated Einstein's relativistic time-dilation law. In addition to utilizing symbolic regression, AI-Descartes uses logical reasoning to determine which candidate equations fit the data best. Said Samsung AI's Cristina Cornelio, "In our work, we are merging a first-principles approach, which has been used by scientists for centuries to derive new formulas from existing background theories, with a data-driven approach that is more common in the machine learning era."
|
Computer Simulations Show How Drugs Get into the Blood
ETH Zurich (Switzerland)
Fabio Bergamin
April 13, 2023
Chemists and pharmaceutical scientists at ETH Zurich in Switzerland and Swiss drugmaker Novartis ran simulations on a supercomputer to gain insights into cyclic peptides' mechanism for penetrating cell membranes. Said ETH Zurich's Sereina Riniker, "Only modeling allows us such detailed, high-resolution insights, as there are no experiments that would let us observe an individual molecule crossing a membrane." The simulations revealed the peptide anchors itself to the membrane's surface, then reconfigures its three-dimensional shape before penetrating the membrane. "The more we know about this mechanism and the properties a molecule must have, the earlier and more effectively researchers can take this into account when developing new drugs," according to Riniker.
|
Raspberry Pi Code Editor Wants to Help Next-Generation Programmers
TechRadar
Craig Hale
April 12, 2023
The Raspberry Pi Foundation has launched a free coding tool to help next-generation programmers globally hone their skills. The Code Editor is accessible on any platform through a Web browser and is designed for children over seven who attend the foundation's Code Clubs or CodeDojos. The beta-format editor can handle Python scripts, which the foundation said is often preferred because of its English-language resemblance and popularity among professional coders. Users are able to broaden the editor's use by creating an account to save and revisit work, enabling homework or general experiments outside the classroom.
|
The Complex Math of Counterfactuals Could Help Spotify Pick Your Next Favorite Song
MIT Technology Review
Will Douglas Heaven
April 4, 2023
Researchers at music-streaming company Spotify have developed a machine learning model that aims to improve automated decision-making. The model is based on counterfactual analysis, complex math used to determine the causes of past events and predict the effects of future events. The researchers used the theoretical framework of twin networks, which views counterfactuals as a pair of probabilistic models (one representing the real world, the other representing a fictional world), as a blueprint for a neural network. They trained the neural network to predict how events would occur in the fictional world, resulting in a computer program that can perform counterfactual reasoning.
|
Soft Robot Hand Learns to Avoid Butter Fingers
New Scientist
Karmela Padavic-Callaghan
April 12, 2023
Scientists at the U.K.'s University of Cambridge have developed a soft robotic hand that learns to prevent objects from slipping out of its grasp. The researchers built the appendage using a three-dimensionally-printed plastic skeleton, molded soft silicone material, and 32 flexible sensors to mimic human skin's pressure sensitivity. They linked the hand to a movable arm with a wrist motor and connected the sensors to a computer. Former Cambridge researcher Thomas Thuruthel said the hand was able to reliably grasp 11 out of 14 randomly selected household objects in more than 1,000 tries and managed to prevent slippage with wrist adjustments about 79% of the time.
|
U.S. Cracked $3.4-Billion Crypto Heist, Bitcoin's Anonymity
The Wall Street Journal
Robert McMillan
April 12, 2023
U.S. authorities cracked bitcoin's anonymity and unraveled a $3.4-billion cryptocurrency heist by exploiting permanent and transactional evidence in the blockchain's online ledger. The authorities used this strategy to unmask James Zhong's crypto theft, stolen via a software bug that enabled fraudulent bitcoin withdrawals on the Silk Road online marketplace. Since then, government officials and private companies have essentially aggregated a blockchain address book from earlier investigations to help federal, state, and local authorities probing cybercrimes. Blockchain analytics firm Chainalysis said it has surveyed more than 1 billion wallet addresses, filtering out legitimate and dubious assets and identifying crypto exchanges. Blockchain analytics charts the flow of cryptocurrency owned by individuals and groups; the Internal Revenue Service says the U.S. has recovered more than $10 billion in stolen digital currency through prosecutions in the past two years.
|
Quantum Approach to Solving Electronic Structures of Complex Materials
Argonne National Laboratory
Sarah C.P. Williams
April 7, 2023
Scientists at the U.S. Department of Energy's Argonne National Laboratory and the University of Chicago have developed a quantum computational solution for electronic structures in materials engineering. The researchers used quantum computers to formulate a hybrid simulation in which four to six quantum bits (qubits) partly perform calculations, while a conventional system further processes the results. The approach yielded the correct electronic structures of several spin defects in solid-state materials after several iterations. The researchers also created a new error-mitigation method to help account for innate quantum-computer noise and to ensure the accuracy of the results. Argonne's Benchen Huang said, "When we scale this up to 100 qubits instead of four or six, we think we might have an advantage over conventional computers. But only time will tell for sure."
|
Detection Tool to Fight Flood of Fake Academic Papers
Financial Times
Clive Cookson
April 13, 2023
Academic publishers have introduced an online tool developed by academic and professional publishing association STM that can identify bogus papers fabricated by "paper mills." Publishers registered with the STM Integrity Hub can upload manuscripts to the cloud-based tool, which scans them for signs of potential fraud like doctored images and similarities to papers previously associated with paper mills. Said Joris van Rossum, who led development of the STM Integrity Hub, “The academic world is traditionally built on trust and in the past few years we have seen that process challenged by systemic fraud. The paper mill industry is a growing problem, made bigger recently by its use of more sophisticated technologies and AI-generated content.”
*May Require Paid Registration
|
|
