cyber ami Le moteur de recherche achète DeepMind, pour plus de 400 millions de dollars. Cette start-up est spécialiste de l'intelligence artificielle. Google prépare-t-il un robot intelligent? Le groupe Internet a confirmé dimanche avoir acheté l'entreprise DeepMind, une société londonienne qui travaille sur l'intelligence artificielle. L'acquisition a été menée par Larry Page, le cofondateur de Google, et aurait coûté 400 millions de dollars selon Re/code ou plus de 500 millions de dollars selon The Information. DeepMind est une entreprise très discrète. DeepMind a été fondée en 2011 par Demis Hassabis, joueur prodige d'échec et neuroscientifique, Shane Legg et Mustafa Suleyman. Google s'est déjà intéressé à la question de l'intelligence artificielle dans le passé. Google, avec ses 56 milliards de dollars, peut investir partout.
cognition artifice In the 1950s and '60s, artificial-intelligence researchers saw themselves as trying to uncover the rules of thought. But those rules turned out to be way more complicated than anyone had imagined. Since then, artificial-intelligence (AI) research has come to rely, instead, on probabilities -- statistical patterns that computers can learn from large sets of training data. The probabilistic approach has been responsible for most of the recent progress in artificial intelligence, such as voice recognition systems, or the system that recommends movies to Netflix subscribers. Early AI researchers saw thinking as logical inference: if you know that birds can fly and are told that the waxwing is a bird, you can infer that waxwings can fly. The problem with this approach is, roughly speaking, that not all birds can fly. Embracing uncertainty “With probabilistic reasoning, you get all that structure for free,” Goodman says. Modeling minds
gestion learning After 24 hours of staring at their screens, the teams that participated in our Disrupt NY 2013 Hackathon have now finished their projects and are currently presenting them onstage. With more than 160 hacks, there are far too many cool ones to write about, but one that stood out to me was NewsRel, an iPad-based news app that uses machine-learning techniques to understand how news stories relate to one other. The app uses Google Maps as its main interface and automatically decides which location is most appropriate for any given story. The app currently uses Reuters‘ RSS feed and analyzes the stories, looking for clusters of related stories and then puts them on the map. Say you are looking at a story about the Boston Marathon bombings. In addition to this, the team built an algorithm that picks the most important sentences from each story to summarize it for you. The team members have a background in machine learning and iOS engineering.
hacked brain By Peter V. Milo August 25, 2012 1:56 AM News Get Breaking News First Receive News, Politics, and Entertainment Headlines Each Morning. Sign Up BERKELEY, Calif. Researchers from the University of California and University of Oxford in Geneva figured out a way to pluck sensitive information from a person’s head, such as PIN numbers and bank information. The scientists took an off-the-shelf Emotiv brain-computer interface, a device that costs around $299, which allows users to interact with their computers by thought. The scientists then sat their subjects in front of a computer screen and showed them images of banks, people, and PIN numbers. The P300 signal is typically given off when a person recognizes something meaningful, such as someone or something they interact with on a regular basis. Scientists that conducted the experiment found they could reduce the randomness of the images by 15 to 40 percent, giving them a better chance of guessing the correct answer.
carreau de_peau Touch sensitivity on gadgets and robots is nothing new. A few strategically placed sensors under a flexible, synthetic skin and you have pressure sensitivity. Add a capacitive, transparent screen to a device and you have touch sensitivity. The brainchild of Stanford University Associate Professor of chemical engineering Zhenan Bao, this “super skin” employs a transparent film of spray-on, single-walled carbon nanotubes that sit in a thin film of flexible silicon, which is then sandwiched between more silicon. After an initial stretch, which actually aligns the randomly sprayed-on conductive, carbon nanotubes into microscopic spring-like forms, the skin can be stretched and restretched again to twice its original size, without the springs or skin losing their resiliency. SEE ALSO: Humanoid Robot Charges Up, Takes a Load Off [VIDEOS] This unique makeup allows the malleable skin to measure force response even as it’s being stretched, or “squeezed like a sponge.”
insects drones A conceptual insect robot If you are talking about big boys’ toys, then surely being part of the military’s hush-hush research and development team would place you squarely in the front line of being able to play with such new technology. While UAVs (Unmanned Aerial Vehicles) have done their bit in wars across the world to be able to scout enemy territory without putting human lives at risk, they are large – and unwieldy at times. What happens when one wants to perform a delicate operation that has a far more covert objective? This is where miniature sized robots come in handy – and insect drones could be the answer to this question. The extremely tiny remote controlled vehicles that are based on insects will most likely have been deployed in sensitive areas to date, where these are called the micro air vehicles (MAVs), and will share similar physics as that employed by flying insects. .
living neural Researchers in Korea have taken a leaf out of the microengineer's book, and used lithographic techniques to build live neural circuits in a petri dish. They hope the technique could be exploited one day to build neural tissue scaffolds, to help regenerate neurons in damaged areas of the body, including the spinal cord. The development is not strictly materials science, but fascinating nonetheless — especially in a world where there is an increasing interplay between biology and technology, with proteins being used as the building blocks for circuits, and graphene proving itself adept at protein detection. The researchers printed a variety of single-cell-sized shapes — including triangles, circles, hexagons, squares and stars — onto a culture substrate using microcontact printing, a form of soft lithography, with a mixture of poly-L-lysine and laminin A chain synthetic peptide.
maths models The folding of an Origami crane Origami (折り紙?, from ori meaning "folding", and kami meaning "paper" (kami changes to gami due to rendaku) is the traditional Japanese art of paper folding, which started in the 17th century AD at the latest and was popularized outside of Japan in the mid-1900s. The principles of origami are also being used in stents, packaging and other engineering structures.[1] History There is much speculation about the origin of Origami. In China, traditional funerals include burning folded paper, most often representations of gold nuggets (yuanbao). The earliest evidence of paperfolding in Europe is a picture of a small paper boat in Tractatus de sphaera mundi from 1490. In the early 1900s, Akira Yoshizawa, Kosho Uchiyama, and others began creating and recording original origami works. Techniques and materials Origami Basic Folds Techniques Many origami books begin with a description of basic origami techniques which are used to construct the models. Origami paper Tools
brain trouble By Rick Nauert PhD Senior News Editor Reviewed by John M. Grohol, Psy.D. on October 8, 2012 UK researchers report the discovery of a neural mechanism that protects individuals from stress and trauma turning into post-traumatic stress disorder. Investigators from the University of Exeter Medical School began with the knowledge of the brain’s “plasticity,” its unique capability to adapt to changing environments. The receptors (called protease-activated receptor 1 or PAR1) act in the same way as a command center, telling neurons whether they should stop or accelerate their activity. Normally, PAR1s tell amygdala neurons to remain active and produce vivid emotions. This adaptation helps us to keep our fear under control, and not to develop exaggerated responses to mild or irrelevant fear triggers. In the study, researchers used a mouse model in which the PAR1 receptors were genetically de-activated. The study has been published in the journal Molecular Psychiatry.
minded computer Scientists believe they have found a way to read people's minds in what could be the first step towards helping brain-damaged patients who cannot speak. US researchers used a computer programme to decode brain activity and put it into words using a form of electronic telepathy. Experts described the breakthrough, unveiled in the journal Public Library of Science Biology, as "remarkable" and believe it could ultimately be possible to decipher people's thoughts. Researchers at the University of California in Berkeley used the programme to predict what spoken words volunteers had listened to by analysing their brain activity. Previous research has shown imagined words activate similar brain areas as words that are said aloud, raising hopes they can also be uncovered by "reading" brain waves. Professor Robert Knight, who worked on the study, said: "This is huge for patients who have damage to their speech mechanisms because of a stroke or Lou Gehrig's disease and can't speak.
brain computer What's the Latest Development? Russian entrepreneur Dmitry Itskov is courting the world's richest individuals to help him in conquering death. Itskov, a 33 year-old, can afford to wait but the billionaires he approaches have an average age of 66, meaning they may be looking for shorter-term solutions to living longer—much longer. What's the Big Idea? Preserving the brain and placing it in a host container, so that the spark of consciousness could outlive the body's organ failure, may be "just a way station to Nirvana, which would ultimately involve downloading the brain’s contents into a computer." Photo credit: Shutterstock.com
neurocell language Imagine if we could understand the language two neurons use to communicate. We might learn something about how thoughts and consciousness are formed. At the very least, our improved understanding of neuron communication would help biologists study the brain with more precision than ever before. Heather Clark, an associate professor of pharmaceutical sciences at Northeastern University, has received a $300,000 Young Faculty Award from the Defense Advanced Research Projects Agency to explore neural cell communication using her expertise in nanosensors. "We were interested in looking into neural cells because of the need to measure chemicals in the brain," she explained. In separate work, Clark had already been developing nanosensors to measure the biochemical environment inside a single neuron. The other DARPA project, she noted, enabled the team to incorporate enzymes into their sensor format.