A study by MIT researchers examining adversarial images finds that AI systems pick up on tiny details in images that are imperceptible to the human eye, which can lead to misidentification of objects, reports Louise Matsakis for Wired. “It’s not something that the model is doing weird, it’s just that you don’t see these things that are really predictive,” says graduate student Shibani Santurkar.
WCVB-TV’s Jennifer Eagan reports that researchers from MIT and MGH have developed a deep learning model that can predict a patient’s risk of developing breast cancer in the future from a mammogram image. Prof. Regina Barzilay explains that the model “can look at lots of pixels and variations of the pixels and capture very subtle patterns.”
Scientific American reporter Jeremy Hsu highlights how CSAIL researchers have developed a robot that can automatically sort recycling. The robot “uses soft Teflon ‘fingers,’ which have fingertip sensors to detect object size and stiffness,” Hsu explains.
MIT researchers have identified a new method to engineer neural networks in a way that allows them to be a tenth of the size of current networks without losing any computational ability, reports Avery Thompson for Popular Mechanics. “The breakthrough could allow other researchers to build AI that are smaller, faster, and just as smart as those that exist today,” Thompson explains.
HealthDay News reporter Amy Norton writes that MIT researchers have developed an AI system that can help predict a woman’s risk of developing breast cancer and provide more personalized care. “If you know a woman is at high risk, maybe she can be screened more frequently, or be screened using MRI,” explains graduate student Adam Yala.
Technology Review reporter Will Knight spotlights how MIT researchers have developed a new chip that is many times more efficient than silicon chips and could help bring AI to a multitude of devices where power is limited. “We need new hardware because Moore’s law has slowed down,” explains Prof. Vivienne Sze.
Writing for Forbes, research engineer Bryan Reimer examines Elon Musk’s recent comments about the future of driverless vehicles. Reimer explains that while there likely won’t be a fully self-driving vehicle system available in the next few years, there will be “an evolution of features that utilize drivers as a backup to the automation in situations requiring intervention.”
Writing for Wired, Prof. Joi Ito, director of the Media Lab, argues against the notion of singularity, the concept that AI will supersede humans. “Instead of thinking about machine intelligence in terms of humans vs machines, we should consider the system that integrates humans and machines – not artificial intelligence but extended intelligence,” writes Ito.
WBUR reporter Pamela Reynolds highlights graduate student Joy Buolamwini’s piece, “The Coded Gaze,” which is currently on display as part of the “Avatars//Futures” exhibit at the Nave Gallery. Reynolds writes that Buolamwini’s piece “questions the inherent bias of coding in artificial intelligence, which has resulted in facial recognition technology unable to recognize black faces.”
Quartz reporter Anne Quito spotlights how graduate student Arnav Kapur has developed a wearable device that allows users to access the internet without speech or text and could help people who have lost the ability to speak vocalize their thoughts. Kapur explains that the device is aimed at augmenting ability.
Inside Science reporter Yuen Yiu writes that MIT researchers have developed a new AI system that can summarize scientific research papers filled with technical terms. Yiu writes that the system “is a dramatic improvement from current programs, and could help scientists or science writers sift through large numbers of papers for the ones that catch their interest.”
Axios reporter Ina Fried spotlights how graduate student Arnav Kapur has developed a system that can detect speech signals. “The technology could allow those who have lost the ability to speak to regain a voice while also opening up possibilities of new interfaces for general purpose computing,” Fried explains.
BBC Click spotlights how CSAIL researchers have developed a robot that can automatically sort recycling. “Many paper and plastic cups look the same, but by introducing the ability to squeeze the object and to know whether it is flexible or not we are able to go one step beyond what today’s methods can do, explains Prof. Daniela Rus, director of CSAIL.
TechCrunch reporter Brian Heater writes that MIT researchers have developed a robot that can recycle materials using sensors that allow it to differentiate between objects. Heater explains that “the system utilizes a Teflon gripper with built in sensors that are capable of determining an object’s makeup based on size and stiffness.”
Writing for The Wall Street Journal, Prof. Thomas Malone examines how AI could transform business hierarchies. “AI may create some more centralized hierarchies, and even more situations that call for flexible structures,” writes Malone. The overall goal, though, will remain “figuring out how to combine the different capabilities of people and computers into ‘superminds’ that are smarter than anything we’ve ever had before.”