Group items tagged

"Researchers fed these algorithms (which function like autocomplete, but for images) pictures of a man cropped below his neck: 43% of the time the image was autocompleted with the man wearing a suit. When you fed the same algorithm a similarly cropped photo of a woman, it auto-completed her wearing a low-cut top or bikini a massive 53% of the time. For some reason, the researchers gave the algorithm a picture of the Democratic congresswoman Alexandria Ocasio-Cortez and found that it also automatically generated an image of her in a bikini. (After ethical concerns were raised on Twitter, the researchers had the computer-generated image of AOC in a swimsuit removed from the research paper.)"

"The disclosure points to a problem security researchers have been warning about for years: that despite its reputation for building what is seen by millions of customers as a secure product, some believe Apple's closed culture and fear of negative press have harmed its ability to provide security for those targeted by governments and criminals. "Apple's self-assured hubris is just unparalleled," said Patrick Wardle, a former NSA employee and founder of the Mac security developer Objective-See. "They basically believe that their way is the best way. And to be fair … the iPhone has had incredible success. "But you talk to any external security researcher, they're probably not going to have a lot of great things to say about Apple. Whereas if you talk to security researchers in dealing with, say, Microsoft, they've said: 'We're gonna put our ego aside, and ultimately realise that the security researchers are reporting vulnerabilities that at the end of the day are benefiting our users, because we're able to patch them.' I don't think Apple has that same mindset.""

"Researchers involved in a recent study trained an artificial intelligence (AI) model to diagnose type 2 diabetes in patients after six to 10 seconds of listening to their voice. Canadian medical researchers trained the machine-learning AI to recognise 14 vocal differences in the voice of someone with type 2 diabetes compared to someone without diabetes. The auditory features that the AI focussed on included slight changes in pitch and intensity, which human ears cannot distinguish. This was then paired with basic health data gathered by the researchers, such as age, sex, height and weight. Researchers believe that the AI model will drastically lower the cost for people with diabetes to be diagnosed."

"More than 150 artificial intelligence researchers have signed an open letter calling for future research in the field to focus on maximising the social benefit of AI, rather than simply making it more capable. The signatories, which include researchers from Oxford, Cambridge, MIT and Harvard as well as staff at Google, Amazon and IBM, celebrate progress in the field, but warn that "potential pitfalls" must be avoided."

"The US government is researching technology that it hopes will turn soldiers into cyborgs, allowing them to connect directly to computers. The US military's Defense Advanced Research Projects Agency (Darpa) has unveiled a research programme called Neural Engineering System Design (NESD) which aims to develop an implantable neural interface, connecting humans directly to computers."

" respite was thanks to a sloppy bit of programming from the worm's creator, who'd left a killswitch in the code: newly infected systems checked to see if a certain domain (iuqerfsodp9ifjaposdfjhgosurijfaewrwergwea.com) existed before attempting to spread the infection; by registering this domain, security researchers were able to freeze the worm.The respite was thanks to a sloppy bit of programming from the worm's creator, who'd left a killswitch in the code: newly infected systems checked to see if a certain domain (iuqerfsodp9ifjaposdfjhgosurijfaewrwergwea.com) existed before attempting to spread the infection; by registering this domain, security researchers were able to freeze the worm. But a day later, it's back, and this time, without the killswitch. Security researchers running honeypots have seen new infections by versions of the worm that can spread even when the iuqerfsodp9ifjaposdfjhgosurijfaewrwergwea.com domain is live."

"Researchers at Columbia University sought to shed light on the problem by tasking 400 AI engineers with creating algorithms that made over 8.2 million predictions about 20,000 people. In a study accepted by the NeurIPS 2020 machine learning conference, the researchers conclude that biased predictions are mostly caused by imbalanced data but that the demographics of engineers also play a role."

"Drowning in literature? Scientists often must manage research, teaching, and acquiring funding, and more. It can be hard to find time to read new papers in the field. It can also help non-specialists who are reading complicated papers and struggling to find the gist. Using this tool, you can enter a paper's abstract. The site will then generate a short summary. "The free tool, which creates what the team calls TLDRs (the common Internet acronym for 'Too long, didn't read'), was activated this week for search results at Semantic Scholar, a search engine created by the non-profit Allen Institute for Artificial Intelligence (AI2) in Seattle, Washington."

"It forecast 14 earthquakes within a 200-mile area of the estimated epicenter and also made a very accurate forecast regarding their intensity, a report on the university's website said. It failed to warn of just one earthquake and gave eight false predictions. The research team trained the AI to detect statistical bumps in real-time seismic data that the research team had paired with previous earthquakes, the report explained. Once trained, the AI monitored for signs of approaching earthquakes. "Predicting earthquakes is the holy grail," said Sergey Fomel, a professor at UT's Bureau of Economic Geology and a member of the research team, adding: "What we achieved tells us that what we thought was an impossible problem is solvable in principle.""

""As expected, we found evidence of a performance improvement over generations due to social learning," the researchers wrote. "Adding an algorithm with a different problem-solving bias than humans temporarily improved human performance but improvements were not sustained in following generations. While humans did copy solutions from the algorithm, they appeared to do so at a lower rate than they copied other humans' solutions with comparable performance." Brinkmann told Motherboard that while they were surprised superior solutions weren't more commonly adopted, this was in line with other research suggesting human biases in decision-making persist despite social learning. Still, the team is optimistic that future research can yield insight into how to amend this."

"Facebook disabled our personal accounts, obstructing the research we lead at New York University to study the spread of disinformation on the company's platform. The move has already compromised our work - forcing us to suspend our investigations into Facebook's role in amplifying vaccine misinformation, sowing distrust in our elections and fomenting the violent riots at the US Capitol on 6 January."

"Up to 10 million Android smartphones have been infected by malware that generates fake clicks for adverts, say security researchers. The software is also surreptitiously installing apps and spying on the browsing habits of victims. The malware is currently making about $300,000 (£232,000) a month for its creators, suggests research."

"Security researcher Troy Hunt reports that the snuggly spies, from Spiral Toys, Security researcher Troy Hunt reports that the snuggly spies, from Spiral Toys, "represents the nexus" of the problem with internet-connected appliances and toys: children being recorded, data being leaked, and the technical possibility of surreptitious access to children through networked toys. "The best way to understand what these guys do is to simply watch the video [advertisement for the toy].""

""The Internet was a way to connect computers, and now it can be a way to connect brains," said researcher Andrea Stocco, assistant professor in psychology at the University of Washington."

""Each time we build new ways of doing something close to human level, it allows us to automate or augment human labor," said Jeremy Howard, founder of Fast.ai, an independent lab based in San Francisco that is among those at the forefront of this research. "This can make life easier for a lawyer or a paralegal. But it can also help with medicine." It may even lead to technology that can - finally - carry on a decent conversation. But there is a downside: On social media services like Twitter, this new research could also lead to more convincing bots designed to fool us into thinking they are human, Howard said."

"The researchers, Tero Karras, Samuli Laine, and Timo Aila, came up with a new way of constructing a generative adversarial network, or GAN. GANs employ two dueling neural networks to train a computer to learn the nature of a data set well enough to generate convincing fakes. When applied to images, this provides a way to generate often highly realistic fakery. The same Nvidia researchers have previously used the technique to create artificial celebrities (read our profile of the inventor of GANs, Ian Goodfellow)."

"This way of thinking about search results is wrong. Recent studies suggest that search engines, rather than providing a neutral way to find information, may actually play a major role in shaping public opinion on political issues and candidates. Some research has even argued that search results can affect the outcomes of close elections. In a study aptly titled In Google We Trust participants heavily prioritized the first page of search results, and the order of the results on that page, and continued to do so even when researchers reversed the order of the actual results."

"But researchers have approximately zero clues as to why it's so good at it, and it doesn't help that the AI essentially taught itself to make these predictions. According to one researcher involved in the project, "We can build these models, but we don't know how they work.""

"the facial recognition system was tested by disney research using infrared hi-def cameras that capture people's faces while watching movies like 'big hero 6', 'the jungle book' and 'star wars: the force awakens'. the results showcased 16 million facial landmarks from 3,179 viewers demonstrating a 'very strong predictive performance'. to do so, the AI software takes the faces of people and understands how many of them are laughing, how wide are their eyes, and the different expressions they make."

"Research by Imperial College in London and Samsung's AI research center showed how a single photo and audio file can be used to generate a singing video portrait. "