Group items tagged

"FOLLOW US Twitter / Facebook / RSS Machine learning image classifiers use context clues to help understand the contents of a room, for example, if they manage to identify a dining-room table with a high degree of confidence, that can help resolve ambiguity about other objects nearby, identifying them as chairs. The downside of this powerful approach is that it means machine learning classifiers can be confounded by confusing, out-of-context elements in a scene, as is demonstrated in The Elephant in the Room, a paper from a trio of Toronto-based computer science academics."

"A new method to crack open locked iPhones is so promising that US government officials have classified it, the Guardian has learned."

"To demonstrate how reality may differ for different Facebook users, The Wall Street Journal created two feeds, one "blue" and the other "red." If a source appears in the red feed, a majority of the articles shared from the source were classified as "very conservatively aligned" in a large 2015 Facebook study. For the blue feed, a majority of each source's articles aligned "very liberal." These aren't intended to resemble actual individual news feeds. Instead, they are rare side-by-side looks at real conversations from different perspectives. "

"Computers can classify skin cancers as successfully as human experts, according to the latest research attempting to apply artificial intelligence to health."

"The FBI can issue national security letters to conduct online surveillance in the US without court approval, but requests often come with a gagging order, which prevents websites from publicly disclosing them. To get around this, many websites state that they have not received any classified requests."

"showing how they can create a 40cm x 40cm "patch" that fools a convoluted neural network classifier that is otherwise a good tool for identifying humans into thinking that a person is not a person -- something that could be used to defeat AI-based security camera systems. They theorize that the could just print the patch on a t-shirt and get the same result."

""What is emerging is that [Aadhaar] is being used to create a panopticon, a centralized database that's linked to every aspect of our lives - finances, travel, birth, deaths, marriage, education, employment, health, etc.," Reetika Khera, an Indian economist and social scientist, told VICE News. Security concerns have plagued the system for years, but in recent weeks criticism has grown deafeningly loud. Earlier this month, as part of the Supreme Court case on privacy, an activist's freedom of information request suggested that foreign firms were being given "full access" to the classified data - including fingerprints and iris scans."

"They then analysed how diverse the individuals' work was at different points in their careers. This was assessed using an artificial intelligence system that was trained, in the case of art, to "recognise" different styles by features such as the brush strokes, shapes and objects in a piece, while in the case of film, it was trained to classify a director's work based on plot and cast information. For science, the system identified different research topics based on the papers cited within a researcher's publications."

"Open AI researchers said that while it was "impossible to reliably detect all AI-written text", good classifiers could pick up signs that text was written by AI. The tool could be useful in cases where AI was used for "academic dishonesty" and when AI chatbots were positioned as humans, they said."

"OpenAI has "discussed and debated quite extensively" when to release a tool that can determine whether an image was made with DALL-E 3, OpenAI's generative AI art model, or not. But the startup isn't close to making a decision anytime soon. That's according to Sandhini Agarwal, an OpenAI researcher who focuses on safety and policy, who spoke with TechCrunch in a phone interview this week. She said that, while the classifier tool's accuracy is "really good" - at least by her estimation - it hasn't met OpenAI's threshold for quality."

"The Open AI researchers were intrigued by a claim that self-driving cars would be intrinsically hard to fool (tricking them into sudden braking maneuvers, say), because "they capture images from multiple scales, angles, perspectives, and the like.""