It’s a small comparison percentage, however, it matters.
The big idea.
Simonkr/iStock.
Continue reading “Women get to speak less in TV debates than men, AI analyzes 625,409 dialogues” »
Traumatic brain injuries might have faded from the headlines since the NFL reached a $765 million settlement for concussion-related brain injuries, but professional football players aren’t the only ones impacted by these injuries. Each year, between 2 million and 3 million Americans suffer from traumatic brain injuries—from elderly people who fall and hit their head, to adolescents playing sports or falling out of trees, to people in motor vehicle accidents.
There are currently no treatments to stop the long-term effects of a traumatic brain injury (TBI), and accurate diagnosis requires a visit to a medical center for a CT scan or MRI, both of which involve large, expensive equipment.
UC San Diego bioengineering Professor Ester Kwon, who leads the Nanoscale Bioengineering research lab at the Jacobs School of Engineering, aims to change that. Kwon’s team is developing nanomaterials—materials with dimensions on the nanometer scale—that could be used to diagnose traumatic brain injury on the spot, be it a sports field, the scene of a car accident, or a clinical setting. They’re also engineering nanoparticles that could target the portion of the patient’s brain that was injured, delivering specific therapeutics to treat the injury and improve the patient’s long-term quality of life.
Los Alamos National Laboratory researchers have developed a novel method for comparing neural networks that looks into the “black box” of artificial intelligence to help researchers comprehend neural network behavior. Neural networks identify patterns in datasets and are utilized in applications as diverse as virtual assistants, facial recognition systems, and self-driving vehicles.
“The artificial intelligence research community doesn’t necessarily have a complete understanding of what neural networks are doing; they give us good results, but we don’t know how or why,” said Haydn Jones, a researcher in the Advanced Research in Cyber Systems group at Los Alamos. “Our new method does a better job of comparing neural networks, which is a crucial step toward better understanding the mathematics behind AI.”
To foster empathy in conversation, scientists at Kyoto University developed a shared-laughter AI system that reacts properly to human laughter.
What makes something hilarious has baffled philosophers and scientists since at least the time of inquiring minds like Plato. The Greeks believed that feeling superior at others’ expense was the source of humor. Sigmund Freud, a German psychologist, thought humor was a means to let off pent-up energy. In order to make people laugh, US comedian Robin Williams tapped his anger at the absurd.
No one appears to be able to agree on the answer to the question, “What’s so funny?” So picture attempting to train a robot to laugh. But by creating an AI that gets its signals from a shared laughing system, a team of researchers at Kyoto University in Japan is trying to do that. The researchers describe their novel technique for creating a funny bone for the Japanese robot ‘Erica’ in the journal Frontiers in Robotics and AI.
Bias in AI systems is proving to be a major stumbling block in efforts to more broadly integrate the technology into our society.
A new initiative that will reward researchers for finding any prejudices in AI systems could help solve the problem.
The effort is modeled on the bug bounties that software companies pay to cybersecurity experts who alert them of any potential security flaws in their products.
Topological materials are a special kind of material that have different functional properties on their surfaces than on their interiors. One of these properties is electrical. These materials have the potential to make electronic and optical devices much more efficient or serve as key components of quantum computers. But recent theories and calculations have shown that there can be thousands of compounds that have topological properties, and testing all of them to determine their topological properties through experiments will take years of work and analysis. Hence, there is a dire need for faster methods to test and study topological materials.
A team of researchers from MIT, Harvard University, Princeton University, and Argonne National Laboratory proposed a new approach that is faster at screening the candidate materials and can predict with more than 90 percent accuracy whether a material is topological or not. The traditional way of solving this problem is quite complicated and can be explained as follows: Firstly, a method called density functional theory is used to perform initial calculations, which are then followed by complex experiments that involve cutting a piece of material to atomic-level flatness and probing it with instruments under high vacuum.
The new proposed method is based on how the material absorbs X-rays, which is different from the old methods, which were based on photoemissions or tunneling electrons. There are certain significant advantages to using X-ray absorption data, which can be listed as follows: Firstly, there is no requirement for expensive lab apparatus. X-ray absorption spectrometers are used, which are readily available and can work in a typical environment, hence the low cost of setting up an experiment. Secondly, such measurements have already been done in chemistry and biology for other applications, so the data is already available for numerous materials.
More and more companies and scientists are working to equip contact lenses with applications that not long ago still seemed like science fiction, such as the ability to record videos or diagnose and even treat diseases. Mojo Vision, an American startup, is one company that has been improving its prototypes since 2015. It is currently developing an ambitious project involving augmented reality lenses that, in addition to correcting your vision, will let you consult all kinds of information, from the trails on a ski slope to your pace when you run, all through microLED displays the size of a grain of sand.
“In the short term, it sounds like a futuristic idea, but 20 years ago we couldn’t even imagine many of the technological advances that we have today,” says Ana Belén Cisneros del Río, deputy dean of the College of Opticians-Optometrists in the Spanish region of Castilla y León, of the Mojo Vision project. However, Daniel Elies, a specialist in cornea, cataract and refractive surgery and medical director of the Institute of Ocular Microsurgery (IMO) Miranza Group in Madrid, does not believe that this type of contact lens will become part of everyday life anytime soon, “especially due to cost issues.”
One of the companies interested in manufacturing augmented reality contacts is Magic Leap. Sony, for its part, applied a few years ago for a patent for lenses that can record videos, and Samsung did the same for lenses equipped with a camera and a display that projects images directly into the user’s eye. Some researchers are trying to create robotic lenses that can zoom in and out with the blink of an eye, and yet others are working on night vision contact lenses, which could be useful in military applications.
Their research was published Sept. 26 in the journal Optics Express.
The five dimensions in question aren’t new or hidden spatial dimensions. Instead, a team headed by Tingting Wu, a Ph.D. student in the McKelvey School of Engineering’s imaging sciences program, was able to design a system that could tell the orientation of a molecule in 3D space as well as its position in 2D: five parameters from a single, noisy, pixelated image.
Anthony Levandowski, the engineer at the heart of the Uber/Google lawsuit, is serious about his artificial intelligence religion.
