Opportunity Rover Tops This Week's Internet News Roundup

Wired 17 Feb 2019 02:00 NASA Administrator Jim Bridenstine spoke at a press conference in Pasadena, California last week announcing the end of the mission of the Mars rover Opportunity. Mario Tama/Getty Images What kind of a week includes Congress turning on the president, Barbies in wheelchairs, adorable wire fox terriers, and actual, genuine discussion over where politicians fall on the moldy jam issue? Your answer, of course, should simply be, "Oh, just any ol' week in 2019." This, dear friends, is the hyper-accelerated world of today, and you just have to strap in and enjoy it. For those wondering about whether or not we should take a moment to reflect on anything, consider this: It's a year after the Parkland shootings, and it's not entirely clear if anything has substantially changed in the gun control debate over the past 12 months despite the best efforts of many. (Too depressing? Sorry; it’s still true, though.) When you're done contemplating that, here's everything else folks were talking about online in the last seven days. So Long, Opportunity What Happened: The internet shed more than a few tears last week in honor of the little Mars rover that could, and some final words that weren't. (They were, however, very poetic.) What Really Happened: A 15-year space mission finally came to an end last week when NASA announced that it had given up attempts to re-establish contact with Mars rover Opportunity after it fell silent last summer following a dust storm. https://twitter.com/NASA/status/1095512637675311104 https://twitter.com/tanyaofmars/status/1095580232797450241 https://twitter.com/Dr_ThomasZ/status/1095768653897916416 The end of the massively successful mission was big news as should be expected—it deserved no less—but what might have been more surprising was how many people were talking about it on social media. The reason for that may have been that so many were moved by one man's descriptions of the last transmission by the rover. https://twitter.com/JacobMargolis/status/1095436913173880832 https://twitter.com/TabletopAudio/status/1095676571875266561 https://twitter.com/BoomBoomBetty/status/1095673301257596935 https://twitter.com/PeachyBaws/status/1095672621964894208 https://twitter.com/MattSpaetzel/status/1095670045450092545 https://twitter.com/SeanMurricane/status/1095734500137545728 https://twitter.com/hailRyaan/status/1095677735228198912 There's only one problem: That's not exactly what happened. https://twitter.com/leighwalton/status/1095742079349780480 But, hey, cynics; here's something to really break your heart: Tweets from Scott Maxwell, one of the project leads in creating Opportunity in the first place. https://twitter.com/marsroverdriver/status/1095786996545404929 https://twitter.com/marsroverdriver/status/1095786999326208000 https://twitter.com/marsroverdriver/status/1095787001830199296 https://twitter.com/marsroverdriver/status/1095787003331805184 OK, if you're not crying now, what is wrong with you? The Takeaway: This feels appropriate, really. https://twitter.com/bafeldman/status/1095762759340433414 Shutdown Woes, Part 2 What Happened: Just when you thought there was going to be another government shutdown, victory was snatched from the jaws of defeat—at least until the jaws of defeat started snapping again. What Really Happened: As the week began, many people were worrying about the potential for another US government shutdown as the three-week window for bipartisan talks neared an end, seemingly without any movement. Then, on Monday, some hope. https://twitter.com/MariannaNBCNews/status/1095134367255543808 https://twitter.com/Phil_Mattingly/status/1095132859881476096 https://twitter.com/Phil_Mattingly/status/1095133138848759809 An agreement "in principal" is better than no agreement at all, and news that was quickly shared everywhere as people breathed a sigh of relief over the chaos that was about to be avoided. So, what was actually in the compromise, anyway? https://twitter.com/ChadPergram/status/1095147894997508096 https://twitter.com/nytimes/status/1095299023681044483 https://twitter.com/aabramson/status/1095144003870384128 That's not a win for the right, as they loudly grumbled. https://twitter.com/IngrahamAngle/status/1095300018674778113 https://twitter.com/WalshFreedom/status/1095164946957156355 https://twitter.com/thehill/status/1095174674718355456 Nonetheless, the Republican leadership in the Senate was all for it— https://twitter.com/senatemajldr/status/1095344789602664449 https://twitter.com/senatemajldr/status/1095344792115007488 https://twitter.com/AaronBlake/status/1095347422417637377 —and they weren't the only ones on the right hoping that President Trump would just suck it up and sign the bill, if only to avoid the disastrous scenes that had arrived as the result of the last shutdown, only weeks ago. https://twitter.com/JoyceKohTV/status/1095360260976660481 https://twitter.com/MariannaNBCNews/status/1095365951061671936 https://twitter.com/CNNnewsroom/status/1095366841298571264 https://twitter.com/costareports/status/1095365507253972997 It sure sounded like everyone wanted the president to just go along with the majority, but would Trump do that? https://twitter.com/dsupervilleap/status/1095372298549379072 https://twitter.com/esaagar/status/1095371563740917763 https://twitter.com/cherylbolen/status/1095372131473473537 Oh, of course. But, he remained as contrarian as ever right up until the very last minute. https://twitter.com/maggieNYT/status/1095534813304553472 One day, the president will remember that this is not a reality television show but instead, you know, the largest economy in the entire world. (Well, for awhile longer, at least.) The Takeaway: How'd it all shake out by the end of the week? Well... https://twitter.com/sahilkapur/status/1096138997041123331 Why Hello Again, Paul Manafort What Happened: Pity poor Paul Manafort, a man who, when given the opportunity by federal authorities to save his own skin, decided that it would be better if he just bit the hand that fed him, instead. What Really Happened: When it comes to former Trump campaign chairman Paul Manafort, it appears that he can’t help but appear in the news on a regular basis. This week’s installment? That would be a federal judge ruling on whether or not Manafort was lying to Special Counsel Robert Mueller when he cut a deal last year. It didn't go well for Manafort. https://twitter.com/ShimonPro/status/1095829291005341696 https://twitter.com/MarshallCohen/status/1095831134129274880 https://twitter.com/kenvogel/status/1095831249338490880 https://twitter.com/ScottMStedman/status/1095829498866524161 So, Manafort's plea deal is now null and void because he lied, again, basically. If you're still confused, perhaps we should just wait for the experts. https://twitter.com/salgentile/status/1095839879269879808 OK, maybe not. For those wondering how bad this is in real terms, the answer turns out to be, "pretty bad." https://twitter.com/renato_mariotti/status/1095833238881341440 One question was on a lot of people's lips after this ruling: What was Manafort thinking? https://twitter.com/WalshFreedom/status/1095835423128121346 https://twitter.com/tedlieu/status/1095842387685974016 https://twitter.com/JoyceWhiteVance/status/1095835311563780101 https://twitter.com/chrislhayes/status/1095834416545497088 https://twitter.com/chrislhayes/status/1095840817380581377 The Takeaway: Of course, perhaps there's a far more obvious reason behind the whole thing. https://twitter.com/ashleyfeinberg/status/1095833004327489537 El Chapo's Money What Happened: By the time that a jury had declared the result of the months-long El Chapo trial, lawmakers had declared a plan to spend his money in service of something that very few actually want. (No, not an El Chapo Teach-Yourself-Escapology Center, as much as we wish otherwise.) What Really Happened: It may not have been the trial of the century—admittedly, most people had entirely forgotten it was still going on—but last week, the trial of Joaquin Guzmán, aka "El Chapo," came to an end with a satisfying conclusion for everyone paying attention—well, except Guzmán. https://twitter.com/ZekeJMiller/status/1095375058221056000 https://twitter.com/marianowrites/status/1095378042845888512 https://twitter.com/CBSNewsRadio/status/1095377856157360128 https://twitter.com/CBSNewsRadio/status/1095377856924925952 The verdict was swiftly spread across the internet, as news of tense scenes in the courtroom started to emerge, adding to the intrigue. https://twitter.com/alanfeuer/status/1095378378482507781 https://twitter.com/keegan_hamilton/status/1095378762693259264 Why so tense? Perhaps because El Chapo was the kind of man to successfully bribe the Mexico's president and escape from prisons with his mistress, both of which are facts that make him sound more like a fictional crime lord or James Bond villain than a real person. No wonder jurors are reportedly hoping to stay anonymous after the trial. (For those wondering, El Chapo is reportedly going to "Supermax" in Colorado, and you can already take a tour, because the world is an incredible place.) But, wait. This story isn't strange enough yet, but it's about to be. https://twitter.com/SenTedCruz/status/1095392602369724416 https://twitter.com/SenTedCruz/status/1095393126758445065 Yes, that's right; Senator Ted Cruz, the internet's favorite Zodiac Killer tribute act, has decided that Guzmán and others like him should … pay for the wall? And he's far from alone. https://twitter.com/DrMarkGreen4TN/status/1095406621226868741 https://twitter.com/RepMarkGreen/status/1095384040990035969 https://twitter.com/RepTimBurchett/status/1095435426695917571 There was no small amount of irony in this announcement, as many pointed out. https://twitter.com/nytimes/status/1095429512660176896 https://twitter.com/CEDickson/status/1095396573826170880 https://twitter.com/adamcbest/status/1095465420977512448 https://twitter.com/TimOBrien/status/1095661495898251264 Before we all rush to judgment for it being a ridiculous idea when there are so, so many other ways to spend $14 billion in a more responsible manner, consider this: Given a choice between using El Chapo's money or stealing the money from Puerto Rico recovery efforts, suddenly Cruz's plan doesn't seem quite so dumb after all, does it? The Takeaway: Hey, look! The president's on board! Well, if by "on board," I mean, "clearly doesn’t know anything about it but isn’t willing to admit that publicly." https://twitter.com/RealSaavedra/status/1095744155576225792 Disappearing Bees—and Other Bugs What Happened: Just in case you forgot that bees were disappearing in record numbers, last week there was news that it wasn't just bees. What Really Happened: Good news for people who hate both creepy crawlies and life on Earth arrived early last week, thanks to a report that suggested that insects may vanish from the planet within the next century, based on current rates of attrition, which could lead to a—as the report calls it, understatedly, "catastrophic collapse of nature’s ecosystems." So, you know, nothing major. https://twitter.com/loubgray/status/1094903233137573890 https://twitter.com/kj_mestre/status/1094915008528568320 https://twitter.com/aliceross_/status/1094927990360821765 https://twitter.com/riptari/status/1094666738799702017 https://twitter.com/carolynlochhead/status/1094926950278287360 https://twitter.com/eugenegu/status/1094977428785512448 https://twitter.com/BostonJerry/status/1094982815022809091 The news of oncoming ecological disaster was something that went viral in a niche manner, even as the right wing fought back with blind denial. It wasn't all doom and gloom, however, as social media also tried to mobilize to recommend things that could be done to slow down the end of the world (as did CNN, unexpectedly). https://twitter.com/sigridellis/status/1095323886860808194 https://twitter.com/OakParkSoil/status/1095466704195338240 https://twitter.com/thisjulieday/status/1095329372699906048 https://twitter.com/inthebarberry/status/1095142046678089733 The response to the report, while widespread for about a day or so after its release, didn't last too long, because … well, everyone has such a short attention span and dumb politicians can be so entertaining sometimes. https://twitter.com/icelandrichard/status/1094895668563464192 Goodbye, insects. Goodbye, planet. The Takeaway: Let's just try and stay focused, shall we? Yes, we know that Brexit is continuing to be a disaster and people are selling secrets to Iran, but come on, people. https://twitter.com/ChuckWendig/status/1095023477193695233 More Great WIRED Stories Journalism isn't dying. It's returning to its roots A crypto CEO dies—with the only key to $137 million Probe your pupper’s genetic secrets with these DNA kits The WIRED guide to commercial human space flight Finding Lena, the patron saint of JPEGs 👀 Looking for the latest gadgets? Check out our latest buying guides and best deals all year round 📩 Want more? Sign up for our daily newsletter and never miss our latest and greatest stories
Read More

Digging Into Self-Driving Data and More Car News This Week

Wired 17 Feb 2019 02:00 This week, the California DMV released the disengagement reports every self-driving developer testing in state must file to the DMV. They're flawed documents, but a careful look yields some helpful insights. Aurora Innovation, Inc. The self-driving car world is a secretive one, where software, hardware, and testing methods are jealously guarded (and occasionally spark a major lawsuit). But this week, we got a glimpse into what these developers have actually been up to, thanks first to a newly released batch of “disengagement” reports every autonomous vehicle outfit testing in California provides to the state at the end of each year. The disengagement data isn’t too helpful, but the reports do reveal a serious spike in would-be AV testing, among other tidbits. More intel comes from SoftBank’s latest move in this space, a nearly $1 billion investment in AV startup Nuro. In non-robo news, we get a tour of all the tools and tricks that keep Nascar races racy, and bid adieu to the 380, Airbus’ freakishly large passenger jet. Let’s get you caught up. Headlines Stories you might have missed from WIRED this week Get a load of the Elroy Air Chaparral, a big ol’ cargo drone meant to carry 500 pounds and fly 300 miles on one electric charge. Uber sues New York to halt its freeze on the number of ride-hail vehicles in the city. Before the racing, there’s the meticulous organization. Join a Nascar pit crew for a tour of what it takes to get its cars and drivers ready for the green flag. A fond farewell to the Airbus 380, the jumbo-iest passenger jet roaming the skies. Farewell, too, to the work of the brilliant logisticians who figured out how to get the plane’s huge parts through the streets of a small French village on their way to the company’s assembly plant in Toulouse. California’s DMV released its annual report on autonomous vehicles “disengagements.” The reports aren’t nearly as transparent as they could be, but we still managed to learn something about the self-driving car companies testing in the state. The SoftBank Vision Fund dropped a cool $940 million into Nuro—a bet that the company’s small, toaster-like self-driving vehicles are the future of home delivery. A survey of transit riders in seven American metros found they’re taking fewer trips than they did two years ago. The culprit? Cars. The solution? A bit more complicated. Security researchers find the Xiaomi scooters once used by Bird and Lyft can be hacked to brake suddenly or accelerate a rider into traffic. Transportation and Logistics Superstars of the Week Give it up for agriculture specialists at the US Customs and Border Protection, who made sure that none of the millions of cut flowers shipped into the country for Valentine’s Day also carried hitchhiking pests and diseases. Stat of the Week 8.7% The year-over-year decline in truck speeds along the nation’s top 10 bottlenecks, including the stretch of I-95 in Fort Lee, New Jersey; the I-285 at I-85 (northbound) in Atlanta, Georgia; and SR 60 at SR 57 in Los Angeles. That’s according to the American Transportation Research Institute, a research organization that studies trucking issues. ATRI blames the slowdown on traffic. Required Reading News from elsewhere on the internet NHTSA, the US federal auto safety regulator, had already admitted a much-touted Tesla Autopilot safety stat was flawed. A new report shows why: terrible math. A record 7 million Americans are three months or more behind on their car loans. Amazon continued its push into the transpo space by leading a $700 million funding round for the American electric pickup truck and SUV company Rivian. GM may also invest in the EV company. Volkswagen may invest in and use technology from Ford-backed AV startup Argo AI as part of a larger cooperative agreement between the two global carmakers. Elon Musk’s Boring Company reportedly discussed building a tunnel in New York with city officials—but city engineers had a bunch of questions. We do, too. A Florida woman sued Lime after a scooter crash left her daughter in a vegetative state. Nothing like a little techno-dystopian sci-fi to get you hyped about the future of transportation!!!! In the Rearview Essential stories from WIRED’s canon That Rivian investment indicates Amazon may be as focused as ever on the future of earthbound, even as its founder and CEO works to liberate the human race from Planet Earth.
Read More

How the Brain Keeps Its Memories in the Right Order

Wired 17 Feb 2019 01:00 Ashley Mackenzie/Quanta Magazine It began about a decade ago at Syracuse University, with a set of equations scrawled on a blackboard. Marc Howard, a cognitive neuroscientist now at Boston University, and Karthik Shankar, who was then one of his postdoctoral students, wanted to figure out a mathematical model of time processing: a neurologically computable function for representing the past, like a mental canvas onto which the brain could paint memories and perceptions. “Think about how the retina acts as a display that provides all kinds of visual information,” Howard said. “That’s what time is, for memory. And we want our theory to explain how that display works.” But it’s fairly straightforward to represent a tableau of visual information, like light intensity or brightness, as functions of certain variables, like wavelength, because dedicated receptors in our eyes directly measure those qualities in what we see. The brain has no such receptors for time. “Color or shape perception, that’s much more obvious,” said Masamichi Hayashi, a cognitive neuroscientist at Osaka University in Japan. “But time is such an elusive property.” To encode that, the brain has to do something less direct. Quanta Magazine About Original story reprinted with permission from Quanta Magazine, an editorially independent publication of the Simons Foundation whose mission is to enhance public understanding of science by covering research developments and trends in mathematics and the physical and life sciences. Pinpointing what that looked like at the level of neurons became Howard and Shankar’s goal. Their only hunch going into the project, Howard said, was his “aesthetic sense that there should be a small number of simple, beautiful rules.” They came up with equations to describe how the brain might in theory encode time indirectly. In their scheme, as sensory neurons fire in response to an unfolding event, the brain maps the temporal component of that activity to some intermediate representation of the experience—a Laplace transform, in mathematical terms. That representation allows the brain to preserve information about the event as a function of some variable it can encode rather than as a function of time (which it can’t). The brain can then map the intermediate representation back into other activity for a temporal experience—an inverse Laplace transform—to reconstruct a compressed record of what happened when. Cydney Scott/Boston University The cognitive neuroscientists Marc Howard (at left) and Karthik Shankar, now at Boston University, have devoted the better part of the past decade to developing a general mathematical framework for how the brain builds a temporal context for episodic memories. Courtesy Karthik Shankcar Just a few months after Howard and Shankar started to flesh out their theory, other scientists independently uncovered neurons, dubbed “time cells,” that were “as close as we can possibly get to having that explicit record of the past,” Howard said. These cells were each tuned to certain points in a span of time, with some firing, say, one second after a stimulus and others after five seconds, essentially bridging time gaps between experiences. Scientists could look at the cells’ activity and determine when a stimulus had been presented, based on which cells had fired. This was the inverse-Laplace-transform part of the researchers’ framework, the approximation of the function of past time. “I thought, oh my god, this stuff on the blackboard, this could be the real thing,” Howard said. “It was then I knew the brain was going to cooperate,” he added. Invigorated by empirical support for their theory, he and his colleagues have been working on a broader framework, which they hope to use to unify the brain’s wildly different types of memory, and more: If their equations are implemented by neurons, they could be used to describe not just the encoding of time but also a slew of other properties—even thought itself. But that’s a big if. Since the discovery of time cells in 2008, the researchers had seen detailed, confirming evidence of only half of the mathematics involved. The other half—the intermediate representation of time—remained entirely theoretical. Until last summer. Orderings and Timestamps In 2007, a couple of years before Howard and Shankar started tossing around ideas for their framework, Albert Tsao (now a postdoctoral researcher at Stanford University) was an undergraduate student doing an internship at the Kavli Institute for Systems Neuroscience in Norway. He spent the summer in the lab of May-Britt Moser and Edvard Moser, who had recently discovered grid cells—the neurons responsible for spatial navigation—in a brain area called the medial entorhinal cortex. Tsao wondered what its sister structure, the lateral entorhinal cortex, might be doing. Both regions provide major input to the hippocampus, which generates our “episodic” memories of experiences that occur at a particular time in a particular place. If the medial entorhinal cortex was responsible for representing the latter, Tsao reasoned, then maybe the lateral entorhinal cortex harbored a signal of time. The kind of memory-linked time Tsao wanted to think about is deeply rooted in psychology. For us, time is a sequence of events, a measure of gradually changing content. That explains why we remember recent events better than ones from long ago, and why when a certain memory comes to mind, we tend to recall events that occurred around the same time. But how did that add up to an ordered temporal history, and what neural mechanism enabled it? Tsao didn’t find anything at first. Even pinning down how to approach the problem was tricky because, technically, everything has some temporal quality to it. He examined the neural activity in the lateral entorhinal cortex of rats as they foraged for food in an enclosure, but he couldn’t make heads or tails of what the data showed. No distinctive time signal seemed to emerge. Tsao tabled the work, returned to school and for years left the data alone. Later, as a graduate student in the Moser lab, he decided to revisit it, this time trying a statistical analysis of cortical neurons at a population level. That’s when he saw it: a firing pattern that, to him, looked a lot like time. He, the Mosers and their colleagues set up experiments to test this connection further. In one series of trials, a rat was placed in a box, where it was free to roam and forage for food. The researchers recorded neural activity from the lateral entorhinal cortex and nearby brain regions. After a few minutes, they took the rat out of the box and allowed it to rest, then put it back in. They did this 12 times over about an hour and a half, alternating the colors of the walls (which could be black or white) between trials. What looked like time-related neural behavior arose mainly in the lateral entorhinal cortex. The firing rates of those neurons abruptly spiked when the rat entered the box. As the seconds and then minutes passed, the activity of the neurons decreased at varying rates. That activity ramped up again at the start of the next trial, when the rat reentered the box. Meanwhile, in some cells, activity declined not only during each trial but throughout the entire experiment; in other cells, it increased throughout. Based on the combination of these patterns, the researchers—and presumably the rats—could tell the different trials apart (tracing the signals back to certain sessions in the box, as if they were timestamps) and arrange them in order. Hundreds of neurons seemed to be working together to keep track of the order of the trials, and the length of each one. “You get activity patterns that are not simply bridging delays to hold on to information but are parsing the episodic structure of experiences,” said Matthew Shapiro, a neuroscientist at Albany Medical College in New York who was not involved in the study. The rats seemed to be using these “events”—changes in context—to get a sense of how much time had gone by. The researchers suspected that the signal might therefore look very different when the experiences weren’t so clearly divided into separate episodes. So they had rats run around a figure-eight track in a series of trials, sometimes in one direction and sometimes the other. During this repetitive task, the lateral entorhinal cortex’s time signals overlapped, likely indicating that the rats couldn’t distinguish one trial from another: They blended together in time. The neurons did, however, seem to be tracking the passage of time within single laps, where enough change occurred from one moment to the next. Tsao and his colleagues were excited because, they posited, they had begun to tease out a mechanism behind subjective time in the brain, one that allowed memories to be distinctly tagged. “It shows how our perception of time is so elastic,” Shapiro said. “A second can last forever. Days can vanish. It’s this coding by parsing episodes that, to me, makes a very neat explanation for the way we see time. We’re processing things that happen in sequences, and what happens in those sequences can determine the subjective estimate for how much time passes.” The researchers now want to learn just how that happens. Lucy Reading-Ikkanda/Quanta Magazine Howard’s mathematics could help with that. When he heard about Tsao’s results, which were presented at a conference in 2017 and published in Nature last August, he was ecstatic: The different rates of decay Tsao had observed in the neural activity were exactly what his theory had predicted should happen in the brain’s intermediate representation of experience. “It looked like a Laplace transform of time,” Howard said—the piece of his and Shankar’s model that had been missing from empirical work. “It was sort of weird,” Howard said. “We had these equations up on the board for the Laplace transform and the inverse around the same time people were discovering time cells. So we spent the last 10 years seeing the inverse, but we hadn’t seen the actual transform. … Now we’ve got it. I’m pretty stoked.” “It was exciting,” said Kareem Zaghloul, a neurosurgeon and researcher at the National Institutes of Health in Maryland, “because the data they showed was very consistent with [Howard’s] ideas.” (In work published last month, Zaghloul and his team showed how changes in neural states in the human temporal lobe linked directly to people’s performance on a memory task.) “There was a nonzero probability that all the work my colleagues and students and I had done was just imaginary. That it was about some set of equations that didn’t exist anywhere in the brain or in the world,” Howard added. “Seeing it there, in the data from someone else’s lab — that was a good day.” Building Timelines of Past and Future If Howard’s model is true, then it tells us how we create and maintain a timeline of the past—what he describes as a “trailing comet’s tail” that extends behind us as we go about our lives, getting blurrier and more compressed as it recedes into the past. That timeline could be of use not just to episodic memory in the hippocampus, but to working memory in the prefrontal cortex and conditioning responses in the striatum. These “can be understood as different operations working on the same form of temporal history,” Howard said. Even though the neural mechanisms that allow us to remember an event like our first day of school are different than those that allow us to remember a fact like a phone number or a skill like how to ride a bike, they might rely on this common foundation. The discovery of time cells in those brain regions (“When you go looking for them, you see them everywhere,” according to Howard) seems to support the idea. So have recent findings—soon to be published by Howard, Elizabeth Buffalo at the University of Washington and other collaborators—that monkeys viewing a series of images show the same kind of temporal activity in their entorhinal cortex that Tsao observed in rats. “It’s exactly what you’d expect: the time since the image was presented,” Howard said. He suspects that record serves not just memory but cognition as a whole. The same mathematics, he proposes, can help us understand our sense of the future, too: It becomes a matter of translating the functions involved. And that might very well help us make sense of timekeeping as it’s involved in the prediction of events to come (something that itself is based on knowledge obtained from past experiences). Howard has also started to show that the same equations that the brain could use to represent time could also be applied to space, numerosity (our sense of numbers) and decision-making based on collected evidence — really, to any variable that can be put into the language of these equations. “For me, what’s appealing is that you’ve sort of built a neural currency for thinking,” Howard said. “If you can write out the state of the brain … what tens of millions of neurons are doing … as equations and transformations of equations, that’s thinking.” He and his colleagues have been working on extending the theory to other domains of cognition. One day, such cognitive models could even lead to a new kind of artificial intelligence built on a different mathematical foundation than that of today’s deep learning methods. Only last month, scientists built a novel neural network model of time perception, which was based solely on measuring and reacting to changes in a visual scene. (The approach, however, focused on the sensory input part of the picture: what was happening on the surface, and not deep down in the memory-related brain regions that Tsao and Howard study.) LEARN MORE The WIRED Guide to Artificial Intelligence But before any application to AI is possible, scientists need to ascertain how the brain itself is achieving this. Tsao acknowledges that there’s still a lot to figure out, including what drives the lateral entorhinal cortex to do what it’s doing and what specifically allows memories to get tagged. But Howard’s theories offer tangible predictions that could help researchers carve out new paths toward answers. Of course, Howard’s model of how the brain represents time isn’t the only idea out there. Some researchers, for instance, posit chains of neurons, linked by synapses, that fire sequentially. Or it could turn out that a different kind of transform, and not the Laplace transform, is at play. Those possibilities do not dampen Howard’s enthusiasm. “This could all still be wrong,” he said. “But we’re excited and working hard.” Original story reprinted with permission from Quanta Magazine, an editorially independent publication of the Simons Foundation whose mission is to enhance public understanding of science by covering research developments and trends in mathematics and the physical and life sciences. More Great WIRED Stories A scary map shows how climate change will alter cities Strava has a new way to build routes with a finger swipe What happens if Russia cuts itself off from the internet Ride with the guy who builds roller coasters in his yard Captain Marvel has the best movie site since Space Jam 👀 Looking for the latest gadgets? Check out our latest buying guides and best deals all year round 📩 Want more? Sign up for our daily newsletter and never miss our latest and greatest stories
Read More

What Is Credential Stuffing?

Wired 17 Feb 2019 12:00 Attackers take a massive trove of usernames and passwords and try to "stuff" those credentials into the login page of other digital services. Emily Waite You may have noticed this happening more and more lately: Online accounts get taken over in droves, but the companies insist that their systems haven't been compromised. It's maddening, but in many cases, technically they're right. The real culprit is a hacker technique known as "credential stuffing." The strategy is pretty straightforward. Attackers take a massive trove of usernames and passwords (often from a corporate megabreach) and try to "stuff" those credentials into the login page of other digital services. Because people often reuse the same username and password across multiple sites, attackers can often use one piece of credential info to unlock multiple accounts. In the last few weeks alone, Nest, Dunkin' Donuts, OkCupid, and the video platform DailyMotion have all seen their users fall victim to credential stuffing. "With all of the massive credential dumps that have happened over the past few years, credential stuffing has become a serious threat to online services," says Crane Hassold, a threat intelligence manager at the digital fraud defense firm Agari. "Most people don't change their passwords regularly, so even older credential dumps can be used with relative success. And since password reuse is rampant, cybercriminals will generally test a set of credentials against numerous different websites." Credential Craze Credential stuffing has been a problem for years now, as troves of credentials from seminal breaches like LinkedIn and Dropbox in 2012 and Myspace in 2013 have been used—to great effect!—in countless credential stuffing campaigns. But one trend in particular has fueled a recent rise in successful campaigns. Recently hackers have posted more gigantic, aggregated credential collections that comprise multiple data breaches. One of the most wild recent examples is known as Collection #1-5, a "breach of breaches" that totaled 2.2 billion unique username and password combinations, all available to download in plaintext—for free. LEARN MORE The WIRED Guide to Data Breaches “With Collections 1 through 5 we have actually seen spikes in credential stuffing recently, immediately after that news came out,” says Shuman Ghosemajumder, chief technical officer at the corporate digital fraud defense firm Shape Security. “In fact, we saw some of the largest credential stuffing attacks across several customers in just that week. And that makes sense because you’ve got all these plaintext usernames and passwords available through a torrent. It democratizes credential stuffing.” The Collection credentials are mostly a few years old, meaning many were already in broad circulation and not worth much. But over the last week, another outlandish trove has provided exactly the type of fresh, high-quality credentials hackers cherish. Posted on the Dream Market dark web marketplace, the collection includes a total of roughly 841 million records, released in three batches, from 32 web services, including MyFitnessPal, MyHeritage, Whitepages, and the file-sharing platform Ge.tt. The first part of the dump costs about $20,000 in bitcoin, the second about $14,500, and the third roughly $9,350. A few of the breaches don’t include passwords, and some that do are protected by cryptographic scrambling that buyers will need to decode, but overall these are top-shelf troves ripe for use in credential stuffing. Hot Stuff As you've probably guessed, credential stuffing relies on automation; hackers aren't literally typing in hundreds of millions of credential pairs across hundreds of sites by hand. Credential stuffing attacks also can't try massive numbers of logins on a site with all the tries coming from the same IP address, because web services have basic rate-limiting protections in place to block floods of activity that could be destabilizing. So hackers use credential stuffing tools, available on malicious platforms, to incorporate "proxy lists" to bounce the requests around the web and make them look like they're coming from all different IP addresses. They can also manipulate properties of the login requests to make it look like they come from a diverse array of browsers, because most websites will flag large amounts of traffic all coming from the same type of browser as suspicious. Credential stuffing tools will even offer integrations with platforms built to defeat Captchas. Credential stuffing campaigns ultimately try to get the malicious requests to blend into the noise of all the legitimate logins happening on a service at any given time, or "simulate the activity of a large population of humans," as Shape Security's Ghosemajumder puts it. It also requires patience; Shape estimates that typically attackers find matches between their test credentials and an account on the platform they are attacking 0.1 to 2 percent of the time. This is why attackers need hundreds of thousands or millions of credential pairs to make credential stuffing attacks worth it. And once they've gotten into some accounts, attackers still need a way to monetize what they find there—either by stealing more personal data, money, gift card balances, credit card numbers, and so on—to make the whole thing worthwhile. Stuff It Th best way to protect against credential stuffing attacks is to use unique passwords for each of your digital accounts—ideally by using a password manager—and turn on two-factor authentication when it's available. But it's not entirely on you. Companies, too, are increasingly attempting to detect and block credential stuffing attempts. And some like Google (which also owns Nest) have started initiatives to proactively check whether users' account credentials have been compromised in breaches and trigger password resets if they discover a match. But the trick is to do all of this without blocking or hindering legitimate activity. One strategy companies can deploy is to track logins that ultimately result in fraud, then blacklist the associated IP address. Over time, this can erode the effectiveness of the proxy lists attackers rely on to mask their mass login attempts. This doesn't completely stop credential stuffing, but does make it more difficult and potentially costly for hackers to carry out the attacks. Services whose users are mainly in specific geographic regions can also establish geofences, blocking proxy traffic that comes in from elsewhere in the world. Once again, though, attackers can ultimately adapt to this restriction as well by switching to using proxy IPs within those areas. A recent credential stuffing attack against the productivity and project management service Basecamp helps illustrate the problem. The company reported recently that it had faced 30,000 malicious login attempts from a diverse set of IP addresses in a single hour. The company began blocking the IPs as quickly as possible, but needed to implement a Captcha to ultimately end the attack. When the barrage died down, Basecamp found that the attackers had only succeeded in penetrating 124 accounts; the company quickly reset those account passwords to revoke the attackers' access. Many companies aren't as prepared to handle the scale of the credential stuffing threat. Shape Security's Ghosemajumder says that it's pretty typical at this point for corporate clients to see 90 percent of their logins come from malicious attacks. He has even worked with customers who deal with credential stuffing in 99.9 percent of login attempts to their service. And while credential dumps from leaks and breaches are the primary fuel for these attacks, criminals can also diversify their approach by using credential pairs gathered from phishing attacks. "Most credential stuffing uses information obtained from the major data breaches," Agari's Hassold says. "But over the past few years there has been a shift in the credential phishing landscape to target generic account credentials that are then 'stuffed' into a number of different websites." Though it is frustrating when companies insist that they haven't been breached and deny responsibility for protecting their users from credential stuffing attacks, the truth is that service providers don't have a foolproof way of defending against this threat. As Basecamp's CTO and co-founder David Heinemeier Hansson put it after the service's recent incident, "Our ops team will continue to monitor and fight any future attacks. ... But if someone has your username and password, and you don’t have 2FA protection, there are limits to how effective this protection can be." For such a simple technique, credential stuffing is frustratingly difficult to quash. So keep your passwords as diverse as possible and use two-factor whenever you can. And complain loudly on social media about any web service that isn't offering it. More Great WIRED Stories A scary map shows how climate change will alter cities Strava has a new way to build routes with a finger swipe What happens if Russia cuts itself off from the internet Ride with the guy who builds roller coasters in his yard Captain Marvel has the best movie site since Space Jam 👀 Looking for the latest gadgets? Check out our latest buying guides and best deals all year round 📩 Want more? Sign up for our daily newsletter and never miss our latest and greatest stories
Read More

This Company Takes the Grunt Work Out of Using the Cloud

Wired 17 Feb 2019 12:00 HashiCorp cofounder Mitchell Hashimoto wrote course registration software in college so he didn't have to wake up early to sign up for classes. HashiCorp Like most 12-year-old boys, Mitchell Hashimoto played a lot of videogames. But he never liked the repetitive parts of games like Neopets, where players feed and care for virtual animals. "I used a lot of bot software that other people wrote to play the more mundane parts for me, so I could do the fun stuff," he says. Those bots were often blocked by gamemakers, so Hashimoto taught himself to program and created his own bot. When the creators of Neopets ordered him to stop using that bot, he was done with the game. Along the way, he discovered that creating bots was more fun. "It's the dream of every child programmer to create an army of robots," Hashimoto, now 29, says. Soon he was writing scripts to automatically set up web forum software. As a student at the University of Washington in Seattle, he wrote course registration software so he didn't have to wake up early to sign up for classes. But even as automation let Hashimoto avoid some tedious tasks, he learned that programming came with its own time-consuming drudgery. So in 2012 Hashimoto and college friend Armon Dadgar cofounded HashiCorp, which makes open-source software designed to free programmers and system administrators from grunt work. "The consistent thread of everything I've ever done is automating the things I don't want to do," Hashimoto says. "Humans are good at creativity; computers should be doing the repetitive work." HashiCorp's flagship product, Terraform, has become the de facto standard for setting up, or "provisioning," cloud infrastructure since the product’s launch in 2014, says Forrester analyst Charles Betz. Many software development tools simply assume that you use Terraform. The software is used by companies like Barclays, Capital One, and General Motors' self-driving car company, GM Cruise. Along the way, HashiCorp has grown to more than 400 employees, raised $174.2 million, and was most recently valued at $1.9 billion. Building and running applications requires programmers and system administrators to install and configure programming languages, database systems, and a host of other tools. Cloud computing made some of this easier, but there’s still plenty of rote work involved in setting up and configuring cloud servers and ensuring that applications have all the components they need to function. Terraform automates these sorts of tasks. Manuel Kiessling, a software architect in Cologne, Germany, likens the experience of using Terraform to ordering food from a restaurant: You don't have to give the chef explicit instructions on how to cook it. The upshot is that it's much easier to get cloud applications up and running. "We've gone from minutes rather than days to provision infrastructure," says Kieran Broadfoot, head of developer experience at Barclays. Developer Focus Much of Terraform's success stems from HashiCorp's focus on developers’ experience. HashiCorp releases open source versions of its products that anyone can use without charge. The open source versions are usually used by individuals, smaller companies, or tests. The company sells versions of its products with advanced features for teams in larger organizations. HashiCorp's first product, which Hashimoto built before starting the company, was a tool called Vagrant. Vagrant helps developers build ready-to-use "virtual machines" that bundle up all the software a developer needs for a particular project. Once a virtual machine is built, it can be reused for other projects: A developer doesn't need to again install or configure the software it contains. Vagrant was an instant hit with programmers, who shared virtual machines to save each other the effort, and helped HashiCorp attract a loyal following of developers who were happy to check out subsequent HashiCorp products like Terraform or its security product Vault. "It's like Apple devices," Kiessling says. "You hold them in your hand and you're not sure if they have the features you need, but you can immediately sense that someone has put a lot of effort and love into them. You really sense that HashiCorp are people who know their stuff, people who care about quality, about technology." Kiessling started with Vagrant, then used Terraform for some personal projects. So far he's only used it for tests for his work at German retail giant Galeria Kaufhof. But that sort of grassroots interest helped HashiCorp land customers like Barclays. "We knew many of our employees were using these technologies, so rather than go against the grain, we went with the tools our developers love," Broadfoot says. Potential Threats Cloud providers typically offer their own provisioning tools—but they tend to work only with that company’s technology. Amazon's CloudFormation tool, for example, only works with Amazon services. Terraform, by contrast, works with many cloud services. It can be configured to run an application's main code from, say, Amazon, but access data from Microsoft Azure. Forrester’s Betz says there's a need for more of these sorts of "multi-cloud" setups. Many companies fear being locked in to a single cloud, he says. "There are people out there saying 'I just got out from under IBM 10 years ago, there's no way in hell I'm going to go all in on Amazon,'" he says. Others need tools that can work with so-called "hybrid clouds," which combine private data centers with public cloud services from companies like Amazon and Google. Acquisitions can also result in companies having software that runs in multiple clouds. For now, Terraform has few direct competitors, Betz says. But it could eventually be displaced by software that accomplishes the same ends in a new way. That’s what happened to Vagrant. It wasn't displaced by a better virtual machine, but by Docker, which uses a potentially more efficient technology called "containers" to create bundles of ready-to-use, self-contained software without the need to virtualize an entire operating system. Even if Terraform is eventually displaced, HashiCorp has developers' attention. Kiessling mostly uses Docker instead of Vagrant now, but he's an advocate for Terraform. Whatever HashiCorp does next, he and countless other developers will be watching. More Great WIRED Stories Monkeys with super-eyes could help cure color blindness All the times Facebook moved fast (and broke things) How to make your home more energy-efficient Twitter still can't keep up with its flood of junk accounts The world might actually run out of people 👀 Looking for the latest gadgets? Check out our latest buying guides and best deals all year round 📩 Get even more of our inside scoops with our weekly Backchannel newsletter
Read More

'We proved Amazon wrong': activists celebrate Bezos retreat from Queens

Guardian Technology 16 Feb 2019 09:00 Protesters rally against the proposed Amazon headquarters in Queens, New York in November. Photograph: Bebeto Matthews/AP Amazon picked a tough moment to come to New York. Progressive groups were in the ascendant and they turned their fire on an obvious target: a plan to lure a company run by the world’s richest man with $3bn in subsidies and tax breaks. The abrupt collapse of Amazon’s plan for a new headquarters, or HQ2, in Long Island City was a milestone victory for leftwing insurgents over establishment Democrats who backed the deal. “This is a new day in New York City politics,” said Sasha Wijeyeratne, executive director of Caaav: Organizing Asian Communities. “They sauntered in here and said HQ2 was inevitable,” she said. “Queens and New York City rose up, and we proved Amazon wrong.” Grassroots activists joined forces with labor unions and leftist groups such as the Democratic Socialists of America, marching on Amazon’s bricks and mortar stores and going door to door to gather signatures in opposition to the project. Local politicians, well aware of the influence of activists who helped unseat a longtime incumbent and elect Alexandria Ocasio-Cortez to the House of Representatives last year, joined the resistance. City councilman Jimmy Van Bramer and state senator Michael Gianaris became leaders in the fight to stop the project. Jimmy Van Bramer speaks during a conference in Queens, New York following Amazon’s announcement it would abandon its proposed headquarters on 14 February. Photograph: Bebeto Matthews/AP They were girding for a years-long fight but it turned out to be unnecessary. Amazon was dragged before the city council for hostile questions and opponents went beyond the New York deal to pound the global company over its labor practices and work with Immigration and Customs Enforcement on facial recognition technology. Amazon opted not to stand and fight. With remarkable speed, executives called off their plans and left. New York City’s mayor, Bill de Blasio, supported the project but thought Amazon should have anticipated the onslaught. This was New York, after all. “Yeah, we are a tough city, but they should have known that coming in,” De Blasio told reporters, saying he was “flabbergasted” when an Amazon executive called to tell him the company was pulling out, not long before the decision was made public. He had expected the company to negotiate with its critics. “There wasn’t a shred of dialogue,” he said. “Out of nowhere they just took their ball and went home.” In Queens, activists were celebrating complete with a mariachi band, a piñata with a likeness of Amazon chief executive Jeff Bezos, and choruses of: “Hey, hey, hey, goodbye.” “If they had good intentions, they would have stayed and figured it out,” said Maritza Silva-Farrell, executive director of Align. “I don’t think they ever would have, because of who they are.” Amazon came promising jobs: 25,000 paying an average of $150,000 a year. But New York, unlike some HQ2 contenders that bet on Amazon to turn around slumping economies, has enjoyed years of economic growth, including a booming tech sector. In a bulletin after the collapse of the deal, S&P Global Ratings called it a “nonevent” for the local economy. Opponents were more concerned about the cost of housing, saying rising rents in Queens would drive residents out of their homes. When pressed to remain neutral if their New York workers attempted to unionize, Amazon executives flatly refused. “The tide is turning,” said Fahd Ahmed, head of the group Desis Rising Up and Moving. “People need good jobs and they can’t afford their rent.” ‘We took it and we destroyed it,’ Frank Raffaele, founder of a Long Island City-based coffee shop chain, said of the deal. Photograph: Bebeto Matthews/AP While opponents celebrated, some residents and business owners rued the death of the deal and expressed anger at what they saw as a betrayal by officials who bowed to the loudest voices. “This was a gift,” said Josh Bowen, who owns John Brown Smokehouse near the site now dominated by warehouses and old industrial buildings where Amazon planned to build its campus. “Have you seen this neighborhood? It’s a dump over here.” “They want to out-AOC AOC,” he said of local politicians channeling the rhetoric of Ocasio-Cortez. “One of the things I thought this new wave would do is breed a lot of leaders. It bred a lot of lemmings, is what it did. It bred a lot of robots.” Polls showed that solid majorities of New Yorkers supported an Amazon campus in Queens. Tenant leaders at public housing developments spoke in favor of the project and unions for construction and building service workers supported the plan. But supporters never mobilized at the scale opponents did. In a statement, Queensbridge Houses tenant association president April Simpson and other tenant leaders called the failure of the deal a “disaster” and blamed “grandstanding politicians” who “spread misinformation to whip up the small band of opponents”. Frank Raffaele, founder of Coffeed, a Long Island City-based coffee shop chain, said he was “overjoyed” when he learned Amazon would be coming to Queens. “We took it and we destroyed it. For no reason,” he said. “These socialists, whoever they are, these activists, they’re not doing the math on this. Their whole mindset is ‘No, no, no, no’.” Raffaele said he understood Amazon’s decision to pull out in face of the torrent of criticism because: “They are not going to be bullied. They don’t want to be bashed all the time.”
Read More

Five of the best top-end smartphones

Guardian Technology 16 Feb 2019 07:00 There’s a smartphone for everyone in our guide to the best top-end mobiles. Photograph: franckreporter/Getty Images There’s never been a better time to buy a top-of-the-line smartphone, with a large range of truly excellent phones available from a variety of different manufacturers starting at £500 and stretching to a wallet-busting £1,000-plus. But with all that top-quality choice, it’s difficult to know which one to buy. Here’s a quick guide to get the best flagship phone for you right now. Huawei Mate 20 Pro Photograph: Samuel Gibbs/The Guardian Huawei Mate 20 Pro RRP: £899 - deals from £650 The Chinese firm might be in the middle of a political storm, but that didn’t stop Huawei from producing the most cutting-edge, feature-packed phone of 2018: the Mate 20 Pro. Headline features include 3D IR-based (the most secure kind) face recognition matching the iPhone XS, a pressure-sensitive in-display fingerprint scanner and a brilliant triple-camera system on the back that is the best in the business. If that wasn’t enough the Mate 20 Pro also has a huge and beautiful 6.38in OLED screen squeezed into a slim and curved body, which keeps the whole phone relatively easy to handle. Huawei’s own Kirin 980 processor matches or bests rivals, while a large battery means it running for close to two days. When it does run out, it has really fast cable charging, fast wireless charging and it can even wirelessly charge another device with the flick of a switch. Verdict If you’re going to blow the budget, this is the one to do it on, unless you’re an iPhone diehard. Apple iPhone XS Photograph: Tatyana Makeyeva/Reuters Apple iPhone XS RRP: £999 - deals from £749 (with trade-in) The best iPhone available isn’t the most expensive, nor is it, sadly, the cheapest. The iPhone XS is the middle of the road with best combination of features, size and that premium feel that makes and iPhone special, which it has to have costing this much. The 5.8in screen is stunning and both big enough for watching videos on the commute and small enough to easily hold and fit in your pocket. The dual camera system on the back is one of the best, the Face ID system works great (there’s no Touch ID or home button anymore) and Apple’s gesture navigation is the best in the business. Performance is top-notch, but battery life is only about a day compared to almost two for competitors. It still has Apple’s Lightning connector, not the newer universal USB-C standard, but does have wireless charging and will get software support longer than any other phone here. Verdict If you want an Apple smartphone, the iPhone XS is the best, but it costs a pretty penny. OnePlus 6T Photograph: OnePlus OnePlus 6T RRP: £499 - deals from £440 OnePlus has made a name for itself by offering top phones that undercut the competition on price, and the latest OnePlus 6T is no exception. For around half the cost of a Huawei Mate 20 Pro or an iPhone XS you get a phone that’s 90% as good. It has the big and beautiful 6.41in full HD+ OLED screen, slim bezels and a tiny notch for the selfie camera. It has a cutting-edge in-display fingerprint scanner. It even has Qualcomm’s top of the line processor and a battery that will last 1.5 days, plus super fast charging via cable. OnePlus doesn’t skimp on the software either, with one of the best Android experiences you can get and three years of software updates from release. The only downsides are a camera that isn’t quite as good as the top-priced models, no wireless charging and not formal water resistance. Verdict With a great Android experience second only to Google’s, good battery, top performance and a gorgeous screen, the OnePlus 6T is hard to beat for £500. Honor View20 Photograph: Samuel Gibbs/The Guardian Honor View20 RRP: £499 Until recently the OnePlus 6T was the king of the cut-price flagship. Now it has some real competition in the form of the Honor View20. Made by Huawei’s e-brand Honor, the View20 has a big, 6.4in full HD+ LCD screen that takes up the entire front with just a small hole in the top left corner for a selfie camera. It also has good battery life, super fast charging, a premium feel, a great rear-mounted fingerprint scanner and the same processor as Huawei’s Mate 20 Pro making it fast and capable. On the back it has a new generation of 48-megapixel camera, which is really very good and a cut above the OnePlus 6T. The Android software experience is solid, but not on the same level as the OnePlus, but you get a similar two to three years of updates from release. It’s even still got a headphone socket - a rarity in 2019. Verdict If you want the best camera for under £500, the Honor View20 is a cracking performer that feels and operates like a phone that’s much more expensive. Google Pixel 3 Photograph: Samuel Gibbs/The Guardian Google Pixel 3 RRP: £739 - deals from £550 As the Android-marker, the Pixel 3 is Google showing what can be done when you create both the hardware and software together, just as Apple does. And as a result it has the best, most polished Android experience going. The hardware more or less matches the competition on specifications, but lacks any truly cutting-edge additions. Instead you get advanced software, a brilliant single camera and Android updates first with software support guaranteed through October 2021. Features such as Gmail’s Smart Compose are fun, while the recent update with Night Sight produces some amazing and unparalleled low-light images. The downside is you have to put up with battery that doesn’t quite match up to the competition lasting only about a day, but at least it has fast cable and wireless charging. There’s a bigger sibling, the Pixel 3 XL, but the Huawei Mate 20 Pro is better all round at that size. Verdict The Google Pixel 3 offers a first-class experience and is the best smaller Android phone currently available. The runners up - watch out for deals Samsung Galaxy S9 Samsung is due to replace the Galaxy S9 next week, but if you can find it for £400 or less it could be a good deal. The larger S9+ could also be a worth looking at. Samsung Galaxy Note 9 The Galaxy Note 9 is Samsung’s huge, stylus-equipped productivity beast that’s overkill for most, but if you need a phone to do everything, this is it. Huawei P20 Pro Huawei’s original hit, the P20 Pro is due to be replaced in the next couple of months, but is still a great phone with a camera that rivals the excellent one on the Mate 20 Pro and can be picked up for as little as £430. Dual-sim versions are available. Apple iPhone X If you don’t want to spend £999 on the iPhone XS and the £749 iPhone XR just doesn’t cut it, Apple’s selling last year’s smash hit, the iPhone X for £769 refurbished. Check Apple’s online store for stock. Samuel Gibbs is the Guardian’s consumer technology editor This article contains affiliate links, which means we may earn a small commission if a reader clicks through and makes a purchase. All our journalism is independent and is in no way influenced by any advertiser or commercial initiative. The links are powered by Skimlinks. By clicking on an affiliate link, you accept that Skimlinks cookies will be set. More information.
Read More

Revealed: Facebook enables ads to target users interested in 'vaccine controversies'

Guardian Technology 15 Feb 2019 06:59 Facebook has been pressured to stop promoting anti-vaccine propaganda. Photograph: Facebook Facebook enables advertisers to promote content to nearly 900,000 people interested in “vaccine controversies”, the Guardian has found. Other groups of people that advertisers can pay to reach on Facebook include those interested in “Dr Tenpenny on Vaccines”, which refers to anti-vaccine activist Sherri Tenpenny, and “informed consent”, which is language that anti-vaccine propagandists have adopted to fight vaccination laws. Facebook’s self-serve advertising platform allows users to pay to promote posts to finely tuned subsets of its 2.3 billion users, based on thousands of characteristics, including age, location, gender, occupation and interests. In some cases, users self-identify their interests, but in other cases, Facebook creates categories based on users’ online activity. In 2017, after a controversy involving antisemitic interest categories, Facebook vowed to build “new guardrails” on its targeting categories. Facebook is already facing pressure to stop promoting anti-vaccine propaganda to users amid global concern over vaccine hesitancy and a measles outbreak in the Pacific north-west. On Thursday, California congressman Adam Schiff, the chair of the House intelligence committee, cited the Guardian’s reporting on anti-vaccine propaganda on Facebook and YouTube in letters to Mark Zuckerberg and Google CEO Sundar Pichai urging them to take more responsibility for health-related misinformation on their platforms. “The algorithms which power these services are not designed to distinguish quality information from misinformation or misleading information, and the consequences of that are particularly troubling for public health issues,” Schiff wrote. “I am concerned by the report that Facebook accepts paid advertising that contains deliberate misinformation about vaccines,” he added. Facebook’s ad-targeting tools are highly valued by businesses because they enable, for example, a pet supply store in Ohio to show its advertising exclusively to pet owners in Ohio. But the tools have also spurred controversy. A Russian influence operation took advantage of the self-service platform to promote divisive content during the 2016 US presidential election. In 2017, ProPublica revealed that the platform included targeting categories for people interested in a number of antisemitic phrases, such as “How to burn Jews” or “Jew hater”. While the antisemitic categories found by ProPublica were automatically generated and were too small to run effective ad campaigns by themselves, the “vaccine controversies” category contains nearly 900,000 people, and “informed consent” about 340,000. The Tenpenny category only includes 720 people, which is too few to run a campaign. Facebook’s ad platform enables targeting of people interested in “vaccine controversies” Photograph: Facebook Following the ProPublica report, Facebook removed many automatically generated targeting categories and said it was “building new guardrails in our product and review processes to prevent other issues like this from happening in the future.” Facebook declined to comment on the ad targeting categories, but said it was looking into the issue. “We’ve taken steps to reduce the distribution of health-related misinformation on Facebook, but we know we have more to do,” a Facebook spokesperson said in a statement responding to Schiff’s letter. “We’re currently working on additional changes that we’ll be announcing soon.” The changes under consideration include removing anti-vaccine misinformation from recommendations and demoting it in search results, the spokesperson said. The steps they have already taken include having third-party fact-checkers review health-related articles. A YouTube spokesperson also declined to comment on Schiff’s letter, but noted the company’s recent changes to its recommendation algorithm to reduce the spread of misinformation, including some anti-vaccine videos. “We’ve done a number of things within this realm,” the spokesperson said. “But these are still early days. And our systems will get better and more accurate.” Competing echo chambers In the past, Facebook has suggested that simply censoring anti-vaccine propaganda might be less effective than counter-speech providing accurate information. But a pro-vaccination activist whose non-profit organization, Voices for Vaccines, uses Facebook to promote positive messages about vaccination to parents, questioned the efficacy of counter-speech on the topic. “I’m not interested in promoting the idea that vaccines are controversial,” said Karen Ernst, who runs Voices for Vaccines from her home in St Paul, Minnesota, and budgets from $50 to $100 each month to advertise pro-vaccine content on Facebook. Ernst said that she was well-aware of the “vaccine controversies” interest category, but never uses it. Targeting people who see vaccines as controversial “gets toward making social media a place where vaccines are fought over, which feels really counterproductive to public health,” she said. “It’s making the echo chamber more echo-y.” Ernst said that while she believes Facebook advertising can be an effective means of promoting vaccination, she feels outgunned by anti-vaxxers. Many of her advertisements have been rejected because she has not registered as a political advertiser, she said, in part out of concern that Facebook’s new transparency tools for political advertising will make her a target for harassment from anti-vaxxers. (The tools would identify Ernst as having paid to promote the posts, and maintain an archive of her ads.) A Facebook ad promoting anti-vaccine propaganda. Photograph: Facebook Larry Cook, a Californian who runs an anti-vaccine Facebook group with more than 150,000 members, has raised more than $7,700 on GoFundMe to fund anti-vaccine misinformation Facebook advertisements since 29 January. Cook has raised nearly $80,000 total on GoFundMe for anti-vaccine promotion since 2015. On GoFundMe, he claims that he has spent more than $35,000 on Facebook advertising over four years. Cook’s advertisements have been censured by the UK advertising regulator, the Advertising Standards Authority, but continue to run in the US, where he is currently focused on targeting mothers in Washington state – the site of a current measles outbreak. Ernst said one way Facebook could help combat anti-vaccine propaganda would be to provide free advertising to public health organizations promoting sound medical advice. Schiff first introduced a House resolution declaring “unequivocal congressional support for vaccines” in 2015. He told the Guardian by phone that he plans to introduce a similar resolution again this year, but that he may update it to feature “the role that these social media companies are playing in the propagation of this bad information”. “It’s difficult to understand why, when this problem has been raised, why either company would take advertising dollars to promote dangerous and misleading information,” he said. “I think our chances of passage are far better than they have been in the past, and tragically that’s because we’ve seen the problem just grow and grow.”
Read More

AI can write just like me. Brace for the robot apocalypse | Hannah Jane Parkinson

Guardian Technology 15 Feb 2019 03:36 ‘AI like the GPT2 system could exacerbate the already massive problem of fake news.’ Photograph: Alamy Elon Musk, recently busying himself with calling people “pedo” on Twitter and potentially violating US securities law with what was perhaps just a joke about weed – both perfectly normal activities – is now involved in a move to terrify us all. The non-profit he backs, OpenAI, has developed an AI system so good it had me quaking in my trainers when it was fed an article of mine and wrote an extension of it that was a perfect act of journalistic ventriloquism. As my colleague Alex Hern wrote yesterday: “The system [GPT2] is pushing the boundaries of what was thought possible, both in terms of the quality of the output, and the wide variety of potential uses.” GPT2 is so efficient that the full research is not being released publicly yet because of the risk of misuse. And that’s the thing – this AI has the potential to absolutely devastate. It could exacerbate the already massive problem of fake news and extend the sort of abuse and bigotry that bots have already become capable of doling out on social media (see Microsoft’s AI chatbot, Tay, which pretty quickly started tweeting about Hitler). It will quash the essay-writing market, given it could just knock ‘em out, without an Oxbridge graduate in a studio flat somewhere charging £500. It could inundate you with emails and make it almost impossible to distinguish the real from the auto-generated. An example of the issues involved: in Friday’s print Guardian we ran an article that GPT2 had written itself (it wrote its own made-up quotes; structured its own paragraphs; added its own “facts”) and at present we have not published that piece online, because we couldn’t figure out a way that would nullify the risk of it being taken as real if viewed out of context. (Support this kind of responsible journalism here!) The thing is, Musk has been warning us about how robots and AI will take over the world for ages – and he very much has a point. Though it’s easy to make jokes about his obsession with AI doom, this isn’t just one of his quirks. He has previously said that AI represents our “biggest existential threat” and called its progression “summoning the demon”. The reason he and others support OpenAI (a non-profit, remember) is that he hopes it will be a responsible developer and a counter to corporate or other bad actors (I should mention at this point that Musk’s Tesla is, of course, one of these corporate entities employing AI). Though OpenAI is holding its system back – releasing it for a limited period for journalists to test before rescinding access – it won’t be long before other systems are created. This tech is coming. Traditional news outlets – Bloomberg and Reuters, for example – already have elements of news pieces written by machine. Both the Washington Post and the Guardian have experimented – earlier this month Guardian Australia published its first automated article written by a text generator called ReporterMate. This sort of reporting will be particularly useful in financial and sports journalism, where facts and figures often play a dominant role. I can vouch for the fact newsrooms have greeted this development with an element of panic, even though the ideal would be to employ these auto-generated pieces to free up time for journalists to work on more analytical and deeply researched stories. But, oh my God. Seeing GPT2 “write” one of “my” articles was a stomach-dropping moment: a) it turns out I am not the unique genius we all assumed me to be; an actual machine can replicate my tone to a T; b) does anyone have any job openings? A glimpse at GPT’s impressiveness is just piling bad news on bad for journalism, which is currently struggling with declining ad revenues (thank you, Google! Thank you, Facebook!); the scourge of fake news and public distrust; increasingly partisan readerships and shifts in consumer behaviour; copyright abuses and internet plagiarism; political attacks (the media is “the enemy of the people”, according to Donald Trump) and, tragically, the frequent imprisonment and killings of journalists. The idea that machines may write us out of business altogether – and write it better than we could ourselves – is not thrilling. The digital layoffs are already happening, the local papers are already closing down. It’s impossible to overstate the importance of a free and fair press. In a wider context, the startling thing is that once super-intelligent AI has been created and released it is going to be very hard to put it back in the box. Basically, AI could have hugely positive uses and impressive implications (in healthcare, for instance, though it may not be as welcomed in the world of the Chinese game Go), but could also have awful consequences. Take a look at this impressive/horrifying robot built by Boston Dynamics, which keeps me from sleeping at night. We’ve come a long way from Robot Wars. Play Video 0:45 New dog-like robot from Boston Dynamics can open doors – video The stakes are huge, which is why Musk – again, in one of his more sensible moods – is advocating for greater oversight of companies well on their way in the AI race (Facebook, Amazon and Alphabet’s DeepMind to take just three examples. AND TESLA). Others have also stressed the importance of extensive research into AI before it’s too late: the late Stephen Hawking even said AI could signal “the end of the human race” and an Oxford professor, Nick Bostrom, has said “our fate would be sealed” once malicious machine super-intelligence had spread. At least as we hurtle towards this cheering apocalypse we’ll have the novels and poetry that GPT2 also proved adept at creating. Now you just need to work out whether it was actually me who wrote this piece. • Hannah Jane Parkinson is a Guardian columnist
Read More