Welcome to the October 21, 2020 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."
|
|
WEF Says Machines Will Create More Jobs Than They Destroy, but Warns of Pandemic 'Double-Disruption'
CNBC Ryan Browne October 20, 2020
The World Economic Forum (WEF) forecasts that robotics and artificial intelligence advancements will result in a net increase in jobs over the next five years. Although the spread of machines and automation would cause the loss of 85 million jobs by 2025, 97 million new jobs would be created concurrently, for an overall gain of 12 million, according to a new WEF report. WEF predicts equal division of labor between humans and machines by 2025, which will require a significant amount of reskilling and upskilling from employers to ensure staff are sufficiently prepared for the future of work. WEF said the combination of technological adoption and the Covid-19 crisis creates a "double-disruption" for workers, with the former transforming the work environment and the latter leading to lob losses.
|
Microsoft, SpaceX Team Up to Bring Cloud Computing to Space
Nextgov Frank Konkel October 20, 2020
Microsoft has partnered with SpaceX and others to make its Azure cloud technology available and accessible to people anywhere on Earth, and potentially those in space. Microsoft will use SpaceX's forthcoming Starlink satellite constellation to bring customers in remote regions high-speed, low-latency broadband; the satellites will function as a channel for data between Microsoft's conventional datacenters and matched ground stations, and the company's modular datacenters. Microsoft also announced an expansion of its Azure Orbital partnership with satellite telecommunications company SES to broaden connectivity between its cloud data centers and edge devices.
|
GM to Run Driverless Cars in San Francisco Without Human Backups
Associated Press Tom Krisher October 16, 2020
General Motors' Cruise autonomous vehicle unit said it will remove human backup drivers from the driverless vehicles it is testing on San Francisco streets by year's end, as California's Department of Motor Vehicles has granted the company a permit to do so. This follows Google subsidiary Waymo's announcement last week that it would open its autonomous ride-hailing service in Phoenix, AZ, without human drivers. Said the University of California, Berkeley's Steven Shladover, "I don't see them as revolutionary steps, but they're part of this step-by-step progress toward getting the technology to be able to work under a wider range of conditions."
|
Drones to Drop Seeds to Boost Koala Gum Tree Numbers After Devastating Australia Bushfires
CNN October 19, 2020
Specialized drones are being tested in a program to increase koala populations on Australia's east coast by dropping gum tree seeds as part of a World Wildlife Fund (WWF) project to revive forest ravaged by bushfires. WWF's Australia unit aims to raise A$300 million ($210 million) over five years to fund the initiative to trial the seed drones and other techniques for restoring the forest, as a means of doubling koala numbers. Some drones can plant 40,000 seeds daily, and will help create corridors so koalas and other fauna can continue to exist in an environment fragmented by fire and land clearing.
|
Hackers Smell Blood as Schools Grapple with Virtual Instruction
The Wall Street Journal David Uberti October 19, 2020
Hackers are attacking U.S. schools struggling with virtual education amid the Covid-19 pandemic, launching ransomware and other attacks against school district networks. Doug Levin at consulting firm EdTech Strategies said at least 289 U.S. districts have suffered cyberattacks in 2020, with a spike in publicly reported attacks in August and September following a lull in the pandemic's early months. Although large school districts have strengthened their cybersecurity in recent years as digitalized instruction has increased, smaller districts lack personnel to police their networks. Reps. Doris Matsui (D-CA) and Jim Langevin (D-RI) have proposed legislation to fortify K-12 cybersecurity by tracking incidents at the federal level and establishing a $400-million grant program for schools.
*May Require Paid Registration
|
California May Replace Cash Bail With Algorithms—But Some Worry That Will Be Less Fair
NBC News Jon Schuppe October 17, 2020
California voters will vote on a ballot measure to replace bail with algorithms that aim to predict whether defendants should be released before trial. Supporters say these algorithms would make the system fairer, as poor people often cannot afford cash bail and either remain locked up or in debt to bail bondsmen. However, many researchers, computer scientists, and civil rights advocates are concerned that algorithmically assigning a person a risk score based on their background and criminal history could worsen discrimination, given that higher arrest rates for black people make them more likely to receive higher risk scores. Said the Pretrial Justice Institute's Meghan Guevara, "The problem with risk assessment tools is that everyone is ranked as having some kind of risk."
|
Experts: Florida Voting Machines Ripe for Foreign Hackers
Government Technology John Pacenti October 16, 2020
Computer scientists have expressed concerns about the security of voting machines used in 49 Florida counties. Although election officials claim the machines are not vulnerable to remote hacking because they are never connected to the Internet, the DS200 voting tabulator uses a wireless connection to transmit results. Finnish computer scientist Harri Hursti said the machine features software that operates like a cellphone and uses Internet Protocol when connecting to the wireless network. Princeton University's Andrew Appel said a hacker could penetrate a border router from the Internet or by walking near a polling place with a Stingray, a portable device that can capture data by mimicking a cellphone tower.
|
Use AI to Convert Ancient Maps Into Satellite-Like Images
IEEE Spectrum Michelle Hampson October 16, 2020
Researchers at the Polytechnic University of Pernambuco in Brazil developed a machine learning algorithm that can transform ancient maps into Google-style satellite images, which could help determine how land use has changed over time, as well as displaying the social and economic impacts of urbanization. The researchers used an existing artificial intelligence (AI) tool, Pix2pix, which uses two neural networks—one that creates images based on the input set, and another that determines whether the generated image is fake or not—that ultimately create realistic-looking images based on the historical data. The researchers said they are working to improve the resolution of the resulting images.
|
The Real Promise of Synthetic Data
MIT News October 16, 2020
Researchers at the Massachusetts Institute of Technology (MIT) have created the Synthetic Data Vault, a set of open-source data-generation tools which students, software developers, and others can use to obtain as much synthetic data as is needed for their projects. Synthetic data can be used as a stand-in for real data, but to be effective, a synthetic dataset must have the same mathematical and statistical properties as the real-world dataset, and cannot contain or hint at the information from the real dataset on which it is based. MIT's Kalyan Veeramachaneni said the interface allows users to tell the synthetic data generator the constraints under which the data is to be generated.
|
The Demon Is in the Detail—Trinity Team Discovers Uniquely Quantum Effect in Erasing Information
Trinity College Dublin (Ireland) October 16, 2020
Researchers at Ireland's Trinity College Dublin and the U.K.'s University of Manchester discovered a uniquely quantum effect in deleting information that may have ramifications for the design of quantum computing processors. They found the quantum nature of information to be erased can give rise to significant deviations in heat dissipation not present in conventional bit erasure. The Trinity and Manchester scientists analyzed an experimentally realistic erasure process that permits quantum superposition, which results in rare events exhibiting heat dissipation far greater than the Landauer limit. Trinity's John Goold said, "This is a highly unusual finding that could be really important for heat management on future quantum chips—although there is much more work to be done, in particular in analyzing faster operations and the thermodynamics of other gate implementations."
|
Researchers Detail LaND, AI That Learns From Autonomous Vehicle Disengagements
VentureBeat Khari Johnson October 15, 2020
Artificial intelligence (AI) researchers at the University of California, Berkeley (UC Berkeley) said they have created AI for autonomous vehicles driving in unseen, real-world conditions that outperforms top techniques in use by delivery robots driving on sidewalks. The Learning to Navigate from Disengagements (LaND) navigation system examines disengagement events (in which the vehicle gives control to a human operator) and anticipates their future recurrence. LaND treats disengagements as part of training data, which enables the robot to learn from datasets gathered during testing. The UC Berkeley researchers said their key insight was that “if the robot can successfully learn to execute actions that avoid disengagement, then the robot will successfully perform the desired task.”
|
Google Warns of 'BleedingTooth' Bluetooth Flaw in Linux Kernel
ZDNet Liam Tung October 14, 2020
Google has shared details of a high-severity flaw affecting the Bluetooth stack in Linux kernel versions below Linux 5.9 that support BlueZ on the Google Security Research Repository on GitHub. An Intel advisory recommends updating the Linux kernel to version 5.9 or later, because "Improper input validation in BlueZ may allow an unauthenticated user to potentially enable escalation of privilege via adjacent access." BlueZ, the official Linux Bluetooth stack, is rolling out Linux kernel fixes for the high-severity flaw, CVE-2020-12351, and for two medium-severity flaws, CVE-2020-12352 and CVE-2020-24490. Google has published a proof-of-concept exploit code for the BleedingTooth vulnerability.
|
Opening the Black Box of Neural Networks
Pacific Northwest National Laboratory Allan Brettman October 13, 2020
Pacific Northwest National Laboratory (PNNL) researchers used deep learning neural networks to model water molecule interactions, unearthing data about hydrogen bonds and structural patterns. The PNNL team employed 500,000 water clusters from a database of more than 5 million water cluster minima to train a neural network, relying on graph theory to extract structural patterns of the molecules' aggregation. The method provides additional analysis after the network has been trained, allowing comparison between measurements of the water cluster networks' structural traits and the predicted neural network, enhancing the network's understanding in subsequent analyses. PNNL's Jenna Pope said, "If you were able to train a neural network, that neural network would be able to do computational chemistry on larger systems. And then you could make similar insights in computational chemistry about chemical structure or hydrogen bonding or the molecules’ response to temperature changes.”
|
|