Welcome to the September 9, 2015 edition of ACM TechNews, providing timely information for IT professionals three times a week.
Updated versions of the ACM TechNews mobile apps are available for Android phones and tablets (click here) and for iPhones (click here) and iPads (click here).
HEADLINES AT A GLANCE
Humanizing Technology: A History of Human-Computer Interaction
The New York Times (09/07/15) Steve Lohr
The history of computers also is the history of the ways humans interact with them, with more accessible and powerful interfaces accompanying more powerful and capable computers. University of Maryland, College Park professor Ben Schneiderman wants to make the history of human-computer interaction (HCI) more prominent with what he calls "the Human-Computer Interaction Pioneers Project," which so far has taken the form of a website on which Schneiderman profiles pioneers in the field, including himself. Schneiderman received the ACM SIGCHI Lifetime Achievement Award in 2001 for his work on "direct manipulation" software, which helped enable hyperlinks and touchscreen keyboards. Schneiderman's website currently has profiles on 45 HCI pioneers, including designers such as Douglas Engelbart, Alan Kay, and Ted Nelson, and researchers such as Sara Bly and Jennifer Preece. Beyond recognizing past achievements, Schneiderman also is looking to the future of HCI. He says the goal of future software will be less about specific features and more about encouraging specific outcomes for both individual users and the society at large. "The future lies in these social designs," he says. "Design to encourage trust, empathy, and responsibility, while protecting privacy. That's the next big thing."
MIT Hosts Discussion on Scaling STEM Education
MIT News (09/04/15) Abby Abazorius
The Massachusetts Institute of Technology's (MIT) Media Lab last week hosted an event examining the current and future state of science, technology, engineering, and mathematics (STEM) education. The "Scaling STEM" event was hosted by MIT president L. Rafael Reif and Rep. Joe Kennedy III (D-MA), and focused on how the use of technology can improve access to quality STEM education, including demonstrations of new learning tools and outreach efforts, as well as presentations by leading education experts and panel discussions. Sanjay Sarma, dean of digital learning at MIT, discussed how MIT had identified a need to extend its education efforts to students of all ages and levels, and how that need informs the desire for scalable research methods. Several presenters discussed the need to use education to help improve the diversity of the STEM fields by attracting more girls and minority students to the subjects at a young age. MIT professor John Gabrieli discussed how new technology can be used to give each student the best possible learning experience by first identifying their unique educational needs. The event concluded with a panel discussion centered around methods of retaining student interest in STEM fields, especially among students in underrepresented groups.
The Strange Link Between Global Climate Change and the Rise of the Robots
The Washington Post (09/08/15) Dominic Basulto
In a recent paper, computer scientists Joel Lehman and Risto Miikkulainen attempted to determine how humans might do in an evolutionary duel with super-intelligent robots following a hypothetical mass extinction event. The researchers concluded humans would not stand a chance against robots, because humans would evolve linearly, while the super-intelligent robots could evolve exponentially. Although the researchers' paper was purely speculative and was not specific about what might bring on a mass extinction event, futurist Dominic Basulto says one could be on its way due to the effects of climate change. Basulto notes the United Nations has predicted the major consequences of climate change, if left unaddressed, could begin to manifest by 2050, about the same time that many artificial-intelligence (AI) researchers expect to see the emergence of super-intelligence. Basulto says the confluence of these two events, combined with advances in robotic technology, could lead to just the scenario pondered by Lehman and Miikkulainen in their paper, which does not bode well for humanity in the long run. However, Basulto says it also might be possible that humans will use advances in AI to help mitigate climate change and avoid its worst effects.
Stanford, Toyota to Collaborate on AI Research Effort
Stanford Report (09/04/15) Bjorn Carey
Toyota is funding a new research center at Stanford University that will focus on artificial intelligence (AI) technologies. The collaboration builds on the AI research that has been pursued at the Stanford Artificial Intelligence Lab (SAIL) since the 1960s. An initial focus of the $25-million SAIL-Toyota Center for AI Research will be AI-assisted driving. Mimicking the brain's performance in an ever-changing environment makes autonomous driving one of the benchmark tasks for AI, says Stanford professor Fei-Fei Li, who heads SAIL and the new AI center. "AI-assisted driving is a perfect platform for advancing fundamental human-centric artificial intelligence research while also producing practical applications," Li says. Stanford computer scientists plan to train computers to recognize objects, speech, and data; use machine learning and statistical modeling to extract meaningful data points; and teach the AI platform to look at critical data sets and plot the safest driving action. "This support will enable us to expand our research in human-centered AI and innovate solutions to some of the world's most pressing challenges," says Persis Drell, dean of the Stanford School of Engineering. Toyota also is funding a parallel research center at the Massachusetts Institute of Technology, which will focus on autonomous systems.
Will Software that Writes Code Alter Tech's Script?
Financial Times (09/07/15) Richard Newton
Several companies are promising to turn the world of computer programming on its head by creating software that will be able to do most of the grunt work of writing software on its own. "I predict that the next computing language will not be computer language but natural language, human language," says Guillaume Bouchard, member of a team of researchers at University College London working on code-writing software. Bouchard also is co-founder of Bloomsbury.ai, which plans to release a public demo of a program that enables users without any background in programming to write software within minutes. Other companies seeking to develop such software include Queue Software and Bubble.is. Although they do not all share Bouchard's vision of software that could build any program based on a simple description, they do aspire to take the drudgery out of software development. Queue Software CEO Aidon Cunniffe says the inspiration for his company came from having to build the same code over and over for various clients and deciding to build a tool that could do that work while he focused on overall design. Bouchard says the goal of such software is to free the world's 11 million programmers to "use their smartness in other domains."
University of St Andrews Gets European Funding to Push the Limits of C++
ComputerWeekly.com (09/07/15) Cliff Saran
The European Union is funding a new initiative designed to develop effective software for emerging parallel computer platforms. Researchers at the University of St. Andrews will investigate how to make data-intensive applications run on highly parallel heterogeneous computing architectures. Improving the performance of data processing has the potential to significantly lower costs and energy consumption. The 3.5-million-euro initiative is part of the RePhrase project, which is tackling issues related to data management for parallel processing, including structuring data to make it efficient to access and to process; placement, migration, and replication of data to enable fast parallel access; and ensuring data consistency. The project will produce new software engineering tools, techniques, and methodologies for developing data-intensive applications in C++, targeting heterogeneous multicore/many core systems that combine central-processing and graphics-processing units into a coherent parallel platform. St. Andrews professor Kevin Hammond says the project will place the university at the forefront of software technology. "We will be able to study fundamental issues in software engineering for parallel programming and apply them to real-world problems," Hammond says. "The potential academic and commercial impact of this work is immense."
A Humanoid Robot to Liaise Between Space Station Crews
Researchers led by Peter Ford Dominey at the National Center for Scientific Research (CNRS) in France have developed "an autobiographical memory" for the robot Nao, enabling it to pass on knowledge between groups of people. The researchers' findings were presented at the 24th International Symposium on Robot and Human Interactive Communication on Sept. 3 in Kobe, Japan. The researchers developed a system in which a human agent can teach the Nao humanoid new actions through physical demonstrations, such as physically putting the robot's limbs in the correct position, imitation via the Kinect visual system, or voice command. The researchers developed a scenario in which an electronic card is damaged, and Nao plays the role of the scientist's assistant by following their directions, bringing or holding parts of the card during repair. If the same failure happens again, the memory of this event would enable the robot to use a video system to show the repair that was made to a new member of the crew. Researchers plan to test the Nao robot in the real conditions of space operations. They also would like to develop another area of application, assisting the elderly, with the robot playing the role of a personal memory aid.
Intelligent Cameras Can Put an End to Always-On Surveillance
New Scientist (09/02/15) Aviva Rutkin
During one of his visits to London, Microsoft Research's Victor Bahl noted that the many closed-circuit TV cameras around the city were mostly recording information that would never be of interest to anyone, wasting resources. Bahl's experience was the inspiration for Vigil, an intelligent camera system to be presented at the International Conference on Mobile Computing and Networking in Paris. Vigil works by examining every frame and counting the number of objects that might be noteworthy, such as people or car license plates. It then ranks snippets of video from most to least important and uploads them to the cloud. For the last two months, the Vigil team has had the system running at three sites, surveying labs or office hallways in London; Redmond, WA; and Madison, WI. In 2016, they hope to launch a bigger pilot that might include monitoring traffic on a city road. Bahl envisions a system such as Vigil could be used to detect the most exciting spots on the field during football games, switching TV coverage to those areas automatically. He says the system also could inform store managers when customers pick up certain products from their shelves.
Researcher Hacks Self-Driving Car Sensors
IEEE Spectrum (09/04/15) Mark Harris
Laser-ranging (lidar) systems that most self-driving cars use to detect obstacles can be hacked by a setup costing about $60, according to Security Innovation's Jonathan Petit. He says attackers can use such a system to trick a self-driving car into thinking something is directly ahead of it, thus forcing it to slow down, or overwhelm it with false signals to completely immobilize it. In a paper written as a research fellow at the University of Cork scheduled to be presented at the Black Hat Europe security conference in November, Petit described his simple setup using a low-power laser and a pulse generator. He found sensors are among a self-driving car's most susceptible technologies because lidar systems use easily-mimicked pulses of laser light to create three-dimensional pictures of the car's surroundings. Petit recorded pulses from a commercial IBEO Lux lidar unit, noting the pulses were not encoded or encrypted, allowing him to replay them at a later point. Petit argues it is never too early to start thinking about security, asserting a "strong system that does misbehavior detection could cross-check with other data and filter out those that aren't plausible."
This Preschool Is for Robots
Bloomberg (09/02/15) Jack Clark
University of California, Berkeley researchers have developed a robot named Brett, short for Berkeley Robot for the Elimination of Tedious Tasks. The robot is part of an initiative to develop artificial intelligence (AI) that lets machines learn the way humans do. Pieter Abbeel, who runs the robotics group at Berkeley, says his research has been partially inspired by watching child psychology tapes. Abbeel says Brett's mind currently is somewhere between the infant and toddler stage, but is picking things up at a rapid pace. The Brett project teaches robots to learn by repeatedly trying to solve problems and adjust their behavior each time to get closer to the goal. "Every five steps or so, it thinks about what it has experienced and updates its behavior," says Sergey Levine, a postdoctoral researcher at Berkeley's technology research lab. Brett's brain relies on a combination of deep learning and reinforcement learning. Combinations of the two have been tried in software before, but the two areas have never been integrated so well in a single robot, according to AI researchers familiar with the Berkeley project. "That's been the holy grail of robotics," says the University of Washington's Carlos Guestrin.
New Technique Lowers Cost of Energy-Efficient Embedded Computer Systems
NCSU News (09/01/15) Matt Shipman
North Carolina State University (NCSU) researchers have developed a technique for creating less-expensive, low-power embedded systems. "Using our techniques, we've been able to create prototype systems with power converters that have a combination of energy efficiency and low cost that...is unmatched by anything currently on the market," says NCSU professor Alex Dean. The researchers used design principles from real-time systems and incorporated the power converter software into the embedded system processor. They note this technique guarantees the other software on the embedded system's processor will not disturb the power converter's correct operation. "This eliminates the need for a separate processor or controller circuit on the power converter itself, which in turn makes the overall system less expensive," Dean says. The new technique also makes the embedded systems smaller, lighter, and more flexible. The embedded system and power converter software are using a shared processor on a single chip, giving developers more coordinated control over both the system's functions and related demands those functions make on the power converter. The researchers made two prototype converters using the new technique, and found both performed better, in terms of cost and efficiency, than dozens of other compatible power converters on the market.
Researchers Turn Clothes Into Electronic Displays
The Engineer (United Kingdom) (09/02/15)
Researchers from the Holst Center in Holland, imec in Belgium, and the CMST lab at Ghent University say they have achieved an advance in wearable devices that could lead to wearable displays in clothing that provide users with feedback. The team has demonstrated the world's first stretchable and comfortable thin-film transistor-driven light-emitting diode display laminated into textiles. Wearable devices such as healthcare monitors and activity trackers already are commonplace, and the team's work could help make wearable devices less obtrusive and more comfortable. The researchers say the technology should encourage more people to use wearable devices regularly, which would increase the quality of their data. "By combining imec's patented stretch technology with our expertise in active-matrix backplanes and integrating electronics into fabrics, we've taken a giant step towards that possibility," says the Holst Center's Edsger Smits. The team reports the display is very thin and mechanically stretchable. Moreover, the researchers say they used fabrication steps that are known to the manufacturing industry.
How a 1,200-Year-Old Hacking Technique Can Already Crack Tomorrow's Encrypted Vaults
Forbes (09/03/15) Thomas Fox-Brewster
In a new paper, Microsoft security researcher Seny Kamara, along with co-researchers Muhammad Naveed from the University of Illinois and Charles Wright from Portland State University, discuss how they were able to extract medical information from hospital databases even when protected by advanced encryption. The researchers downloaded real patient data from 200 U.S. hospitals from the National Inpatient Sample database of the Healthcare Cost and Utilization Project. They targeted two types of the "property preserving elements" of encryption scheme CryptDB: the order preserving encryption (OPE) component and the deterministic encryption (DTE) component. Using a technique developed 1,200 years ago by Arab scholar al-Kindi on the DTE-protected columns in the database, it was possible to look at the scrambled versions of the medical information to see what groups of encrypted data occurred most often. Hackers hoping to exploit these techniques would first have to gain access to the server on which the database was held. They would then have to wait for queries to be made on the vault to get to the right layer and successfully expose the information. The researchers found if they sought names, hackers could use de-anonymization attacks by correlating auxiliary datasets to expose whose details they had uncovered.
Abstract News © Copyright 2015 INFORMATION, INC.
To submit feedback about ACM TechNews, contact: [email protected]
Current ACM Members: Unsubscribe/Change your email subscription by logging in at myACM.