Slashdot Mirror

How long is the U.S. coastline? It's a straightforward question, and one that's important for scientists and government agencies alike. From a report: The U.S. Geological Survey could give you an answer, too, but I'm going to tell you right now that it's wrong. In fact, no one could give you the right answer, and if you look around, you'll find a number of estimations that differ by seemingly improbable amounts. One government report lists the number as 12,383 miles. The same report admits that a different government agency says the figure is actually 88,612 miles. That's an almost eight-fold disparity for a fact that seems simple to obtain. We all know how to use a ruler, right?

Well, we all know how to measure a straight line, but what about a curve? And what if that curve has curves? The crux of the problem comes down to geometry, and the fundamentally uneven nature of coastlines. Though the border between land and sea may look fairly straight when seen from far away, they're anything but. Coastlines jut and dip, curve and cut, and each deviation from a straight line adds distance. Some of these features are massive, like bays, while others are miniscule.

One of the poorest-kept secrets in Silicon Valley has been the huge salaries and bonuses that experts in artificial intelligence can command. Now, a little-noticed tax filing by a research lab called OpenAI has made some of those eye-popping figures public [Editor's note: the link may be paywalled]. From a report: OpenAI paid its top researcher, Ilya Sutskever, more than $1.9 million in 2016. It paid another leading researcher, Ian Goodfellow, more than $800,000 -- even though he was not hired until March of that year. Both were recruited from Google. A third big name in the field, the roboticist Pieter Abbeel, made $425,000, though he did not join until June 2016, after taking a leave from his job as a professor at the University of California, Berkeley. Those figures all include signing bonuses.

[...] Salaries for top A.I. researchers have skyrocketed because there are not many people who understand the technology and thousands of companies want to work with it. Element AI, an independent lab in Canada, estimates that 22,000 people worldwide have the skills needed to do serious A.I. research -- about double from a year ago.

Last week, President Donald Trump signed the Fight Online Sex Trafficking Act (FOSTA) into law. It's a bill that penalizes any platform found "facilitating prostitution," and has caused many advocacy groups to come out against the bill, saying that it undermines essential internet freedoms. The most recent entity to decry FOSTA is Cloudflare, which recently decided to terminate its content delivery network services for an alternative, decentralized social media platform called Switter. Motherboard talked to Cloudflare's general counsel, Doug Kramer, about the bill and he said that FOSTA was an ill-consider bill that's now become a dangerous law: "[Terminating service to Switter] is related to our attempts to understand FOSTA, which is a very bad law and a very dangerous precedent," he told me in a phone conversation. "We have been traditionally very open about what we do and our roles as an internet infrastructure company, and the steps we take to both comply with the law and our legal obligations -- but also provide security and protection, let the internet flourish and support our goals of building a better internet." Cloudflare lobbied against FOSTA, Kramer said, urging lawmakers to be more specific about how infrastructure companies like internet service providers, registrars and hosting and security companies like Cloudflare would be impacted. Now, he said, they're trying to figure out how customers like Switter will be affected, and how Cloudflare will be held accountable for them.

"We don't deny at all that we have an obligation to comply with the law," he said. "We tried in this circumstance to get a law that would make sense for infrastructure companies... Congress didn't do the hard work of understanding how the internet works and how this law should be crafted to pursue its goals without unintended consequences. We talked to them about this. A lot of groups did. And it was hard work that they decided not do." He said the company hopes, going forward, that there will be more clarity from lawmakers on how FOSTA is applied to internet infrastructure. But until then, he and others there are having to figure it out along with law enforcement and customers. "Listen, we've been saying this all along and I think people are saying now, this is a very bad law," Kramer said. "We think, for now, it makes the internet a different place and a little less free today as a result. And there's a real-world implication of this that people are just starting to grapple with."

Web makers continue to ditch the infamous Flash for other safer, improved technologies. In 2011, more than 28.5 percent of websites used Flash in their code, a figure technology survey site W3Techs estimates to have dropped to 4.9 percent today. BleepingComputer: The number confirms Flash's decline, and a reason why Adobe has decided to retire the technology at the end of 2020. A decline from 28.5 percent to 4.9 percent doesn't look that bad, but we're talking about all Internet sites, not just a small portion of Top 10,000 or Top 1 Million sites. Taking into account the sheer number of abandoned sites on today's Internet, the decline is quite considerable, and W3Techs' findings confirm similar statistics put out by a Google security engineer in February.

An anonymous reader quotes a report from Ars Technica: Tesla has an Autopilot problem, and it goes far beyond the fallout from last month's deadly crash in Mountain View, California. Tesla charges $5,000 for Autopilot's lane-keeping and advanced cruise control features. On top of that, customers can pay $3,000 for what Tesla describes as "Full Self-Driving Capability." "All you will need to do is get in and tell your car where to go," Tesla's ordering page says. "Your Tesla will figure out the optimal route, navigate urban streets (even without lane markings), manage complex intersections with traffic lights, stop signs and roundabouts, and handle densely packed freeways with cars moving at high speed." None of these "full self-driving" capabilities are available yet. "Self-Driving functionality is dependent upon extensive software validation and regulatory approval, which may vary widely by jurisdiction," the page says. "It is not possible to know exactly when each element of the functionality described above will be available, as this is highly dependent on local regulatory approval."

But the big reason full self-driving isn't available yet has nothing to do with "regulatory approval." The problem is that Tesla hasn't created the technology yet. Indeed, the company could be years away from completing work on it, and some experts doubt it will ever be possible to achieve full self-driving capabilities with the hardware installed on today's Tesla vehicles. "It's a vastly more difficult problem than most people realize," said Sam Abuelsamid, an analyst at Navigant Research and a former auto industry engineer. Tesla has a history of pre-selling products based on optimistic delivery schedules. This approach has served the company pretty well in the past, as customers ultimately loved their cars once they ultimately showed up. But that strategy could backfire hugely when it comes to Autopilot.

Netflix expects to spend about $8 billion on content this year. For Hollywood studios, that's a reasonable figure. But for Netflix CEO Reed Hastings, it's not enough. From a report: The company is competing against a range of traditional entertainment companies around the globe, and of course, against the need to work, sleep and do other things. Speaking at TED in Vancouver, Hastings noted that $8 billion is about what Disney spends. "That's spread globally," he said. "It's not as much as it sounds." Hastings noted that House of Cards wasn't really the company's first effort at original content. It had tried back in the days when it was still mailing out DVDs. "It didn't work out because we were sub-scale," he said.

Like astronomers who discover new stars, security experts who first identify computer bugs, viruses, worms, ransomware and other coding catastrophes often get to name their finds. Such discoveries now number in the thousands each year, so crafting a standout moniker can be a serious challenge. From a report: Two years ago, German security firm SerNet GmbH figured a punchy name for their bug discovery would give the company a publicity jolt. They called it Badlock, designed a fractured-lock logo and set up a website. The marketing push backfired when some security experts decided Badlock wasn't that bad. Cynical hackers called it Sadlock. "We would not do this again," says SerNet Chief Executive Johannes Loxen of the branding blitz, which he says was overkill because a relatively small number of people were affected by Badlock. Hackers are no fans of marketing. They brand things in their own way. Puns and historic references are the name of the game. "They see it as a kind of grass-roots initiative," says Gabriella Coleman, an anthropologist who teaches courses on hacker culture at McGill University in Montreal.

Some venerable names that have stood the test of time: The Love Bug, for the worm that attacked millions of Windows personal computers in 2000, and Y2K, a turn-of-the-century programming scare that didn't live up to its hype. Many names tend more toward geekspeak. The title of hacker magazine 2600 is a tip of the hat to 2600 hertz, the frequency old-school hackers reproduced to trick AT&T phone lines into giving them free calls. Computer worm Conficker is an amalgam of "configure" and a German expletive. Code Red is named after the Mountain Dew drink researchers guzzled while investigating the worm.

According to a recently-shared press release, Apple has finally hit its goal of running its own operations off 100% renewable energy. "All Apple facilities, from Apple Park to its data centers to worldwide fleet of Apple retail stores, are now solely powered by green energy," reports 9to5Mac. From the report: This figure does not include Apple's third-party suppliers or manufacturers, although the company is convincing many of those to switch to 100% renewable sources too. Apple's environment VP Lisa Jackson discussed the news in an interview with Fast Company. Jackson highlights how Apple has not only focused on reducing emissions but also contributed to the availability of green energy on the grid. Apple has gone from 16% renewable energy to 100% in eight years, with CO2 emissions falling by 58%. The company has built numerous wind and solar farms in cooperation with local institutions, as well as intense focus on environmental sustainability during development of its new buildings like Apple Park. Its data centers are flanked by fields of solar panels. Filling out the last 4% required Apple to find renewable energy sources in some of its more remote retail stores and offices. It has signed power purchase agreements in Brazil, India, Israel, Mexico and Turkey.

T-Mobile Austria admitted on Twitter that it stores at least part of their customer's passwords in plaintext. What this means is that "if anyone breaches T-Mobile (it's only a matter of time), they could likely guess or brute-force every user's password," reports Motherboard. "If the passwords were fully encrypted or hashed, it wouldn't be that easy. But having a portion of the credential in plaintext reduces the difficulty of decoding the hashed part and obtaining the whole password." From the report: "Based on what we know about how people choose their passwords," Per Thorsheim, the founder of the first-ever conference dedicated to passwords, told me via Twitter direct message, "knowing the first 4 characters of your password can make it DEAD EASY for an attacker to figure out the rest." T-Mobile doesn't see that as a problem because it has "amazingly good security." On Thursday, a T-Mobile Austria customer support employee made that stunning revelation in an incredibly nonchalant tweet. Twitter user Claudia Pellegrino was quick to point out that storing passwords in plaintext is wrong, but another T-Mobile customer rep didn't see it that way. "I really do not get why this is a problem. You have so many passwords for every app, for every mail-account and so on. We secure all data very carefully, so there is not a thing to fear," the rep wrote back.

An anonymous reader writes: Within the past week, two Tesla crashes have been reported while Autopilot was engaged, and both involved a Tesla vehicle slamming into a highway divider. One of the crashes resulted in the death of Walter Huang, a Tesla customer with a Model X. The other crash resulted in minor injuries to the driver, thanks largely to a working highway safety barrier in front of the concrete divider. Ars Technica reports on the growing evidence that Tesla's Autopilot handles lane dividers poorly: "The September crash isn't the only evidence that has emerged that Tesla's Autopilot feature doesn't deal well with highway lane dividers. At least two people have uploaded videos to YouTube showing their Tesla vehicles steering toward concrete barriers. One driver grabbed the wheel to prevent a collision, while the other slammed on the brakes. Tesla argues that this issue doesn't necessarily mean that Autopilot is unsafe. 'Autopilot is intended for use only with a fully attentive driver,' a Tesla spokesperson told KGO-TV. Tesla argues that Autopilot can't prevent all accidents but that it makes accidents less likely. There's some data to back this up. A 2017 study by the National Highway Transportation Safety Administration (NHTSA) found that the rate of accidents dropped by 40 percent after the introduction of Autopilot. And Tesla argues that Autopilot-equipped Tesla cars have gone 320 million miles per fatality, much better than the 86 million miles for the average car. These figures don't necessarily settle the debate. That NHTSA figure doesn't break down the severity of crashes -- it's possible that Autopilot prevents relatively minor crashes but is less effective at preventing the most serious crashes. And as some Ars commenters have pointed out, luxury cars generally have fewer fatalities than the average vehicle. So it's possible that Tesla cars' low crash rates have more to do with its wealthy customer base than its Autopilot technology. What we can say, at a minimum, is that there's little evidence that Autopilot makes Tesla drivers less safe. And we can expect Tesla to steadily improve the car's capabilities over time."

Facebook was in talks with top hospitals and other medical groups as recently as last month about a proposal to share data about the social networks of their most vulnerable patients, CNBC reported on Thursday. From the story: Facebook was intending to match it up with user data it had collected, and help the hospitals figure out which patients might need special care or treatment. The proposal never went past the planning phases and has been put on pause after the Cambridge Analytica data leak scandal raised public concerns over how Facebook and others collect and use detailed information about Facebook users. "This work has not progressed past the planning phase, and we have not received, shared, or analyzed anyone's data," a Facebook spokesperson told CNBC. But as recently as last month, the company was talking to several health organizations, including Stanford Medical School and American College of Cardiology, about signing the data-sharing agreement.

In a blog post today, Microsoft said that it has created what it believes is the "first technological breakthrough" toward making conversations with chatbots more like speaking to another person. Windows Central reports: Microsoft says that it has figured out how to make chatbots talk and listen at the same time, allowing them to operate in "full duplex," to use telecommunications jargon. The company says this allows chatbots or assistants to have a flowing conversation with humans, much more akin to how people talk to one another. That stands in contrast to how digital assistants and bots currently work, where only one side can talk at any given time. The technology is already up and running in Xiaolce, Microsoft's AI chatbot currently operating in China. Using "full duplex voice sense," as Microsoft calls it, Xiaolce can more quickly predict what the person it is speaking to will say. "That helps her make decisions about both how and when to respond to someone who is chatting with her, a skill set that is very natural to people but not yet common in chatbots," Microsoft says. Another bonus of the breakthrough is that people interacting with chatbots don't have to use a "wake word" every time they speak during a conversation.

An anonymous reader quotes a report from TechCrunch: Without warning, Instagram has broken many of the unofficial apps built on its platform. This weekend it surprised developers with a massive reduction in how much data they can pull from the Instagram API, shrinking the API limit from 5,000 to 200 calls per user per hour. Apps that help people figure out if their followers follow them back or interact with them, analyze their audiences or find relevant hashtags are now quickly running into their API limits, leading to broken functionality and pissed off users. Two sources confirmed the new limits to TechCrunch, and developers are complaining about the situation on StackOverflow. In a puzzling move, Instagram is refusing to comment on what's happening while its developer rate limits documentation site 404s. All it would confirm is that Instagram has stopped accepting submissions of new apps, just as Facebook announced it would last week following backlash over Cambridge Analytica. Developers tell me they feel left in the dark and angry that the change wasn't scheduled or even officially announced, preventing them from rebuilding their apps to require fewer API calls.

An anonymous reader writes: The last male northern white rhinoceros died just last week, and a total of just 29,000 rhinoceroses now remain on earth. But National Geographic reports that "the genetic material of several northern white rhinos has been stored away," and scientists hope to give birth to another using in vitro fertilization -- or to breed a hybrid using a genetically similar southern white rhino.
Meanwhile, a postdoctoral fellow in ecology and evolutionary biology reports that scientists are seriously considering the possibility of "de-extincting" the Carolina parakeet, America's only native parrot, which became extinct 100 years ago. Thanks to the data I compiled as well as cutting-edge machine learning approaches to analyze those data, my colleagues and I were able to reconstruct the Carolina parakeets' likely range and climate niche, [which] turned out to be much smaller than previously believed... While this may seem rather minor, some scientists consider the Carolina parakeet one of the top candidates for 'de-extinction', a process in which DNA is harvested from specimens and used to "resurrect" extinct species... If someone were to spend millions of dollars doing all of the genetic and breeding work to bring back this species, or any other, how will they figure out where to release these birds...? Whether or not de-extinction is a worthwhile use of conservation effort and money is another question, best answered by someone other than me. But this is just an example of one potential use of this type of research. "
It seems like all kinds of havoc could ensue if we released a resurrected species back into the modern ecosystem. And yet Harvard researchers are already working to breed a new creature that's half-elephant, half Wooly Mammoth.

What do Slashdot's readers think? Should we revive extinct species?

The founder of Rendition Security believes his daughter "is more safe on a Chromebook than a Windows laptop," and he's not the only one. CNET's staff reporter argues that Google's push for simplicity, speed, and security "ended up playing off each other." mspohr shared this article: Heading to my first security conference last year, I expected to see a tricked-out laptop running on a virtual machine with a private network and security USB keys sticking out -- perhaps something out of a scene from "Mr. Robot." That's not what I got. Everywhere I went I'd see small groups of people carrying Chromebooks, and they'd tell me that when heading into unknown territory it was their travel device... "If you want prehardened security, then Chromebooks are it," said Kenneth White, director of the Open Crypto Audit Project. "Not because they're Google, but because Chrome OS was developed for years and it explicitly had web security as a core design principle...." Drewry and Liu focused on four key features for the Chromebook that have been available ever since the first iteration in 2010: sandboxing, verified boots, power washing and quick updates. These provided security features that made it much harder for malware to pass through, while providing a quick fix-it button if it ever did.

That's not to say Chrome OS is impervious to malware. Cybercriminals have figured out loopholes through Chrome's extensions, like when 37,000 devices were hit by the fake version of AdBlock Plus. Malicious Android apps have also been able to sneak through the Play Store. But Chrome OS users mostly avoided massive cyberattack campaigns like getting locked up with ransomware or hijacked to become part of a botnet. Major security flaws for Chrome OS, like ones that would give an attacker complete control, are so rare that Google offers rewards up to $200,000 to anyone who can hack the system.
The article argues that "Fewer software choices mean limited options for hackers. Those are some of the benefits that have led security researchers to warm up to the laptops...

"Chrome OS takes an approach to security that's similar to the one Apple takes with iOS and its closed ecosystem."

dryriver shares a report from Nature about a neural network-based, deep-learning software that is as good as trained chemists in figuring out what reagents and reactions may lead to the successful creation of a desired organic molecule: Chemists have a new lab assistant: artificial intelligence. Researchers have developed a "deep learning" computer program that produces blueprints for the sequences of reactions needed to create small organic molecules, such as drug compounds. The pathways that the tool suggests look just as good on paper as those devised by human chemists. The tool, described in Nature on March 28, is not the first software to wield artificial intelligence (AI) instead of human skill and intuition. Yet chemists hail the development as a milestone, saying that it could speed up the process of drug discovery and make organic chemistry more efficient. Chemists have conventionally scoured lists of reactions recorded by others, and drawn on their own intuition to work out a step-by-step pathway to make a particular compound. They usually work backwards, starting with the molecule they want to create and then analyzing which readily available reagents and sequences of reactions could be used to synthesize it -- a process known as retrosynthesis, which can take hours or even days of planning. The new AI tool, developed by Marwin Segler, an organic chemist and artificial-intelligence researcher at the University of Munster in Germany, and his colleagues, uses deep-learning neural networks to imbibe essentially all known single-step organic-chemistry reactions -- about 12.4 million of them. This enables it to predict the chemical reactions that can be used in any single step. The tool repeatedly applies these neural networks in planning a multi-step synthesis, deconstructing the desired molecule until it ends up with the available starting reagents.

An anonymous reader writes: China's relatively young internet industry is facing a mature-market problem: User growth for popular online services such as instant messaging, search, online news and video has fallen to single digits. Online population growth has hovered around 5% to 6% annually since 2014, which is only slightly higher than in mature economies. Unlike in many developed markets, a vast number of Chinese are unconnected. As they slowly come online, they're creating a sizable market that companies can tap into -- if they can figure out how.

"The Chinese internet is experiencing the third wave of [a] demographic dividend," said Wang Hua, a partner at venture-capital firm Sinovation Ventures, at a speech in December. The first wave, he said, were early adopters, while the second was driven by young people in major cities. "About half of the Chinese population is not yet heavy internet users, and they're the third wave of the demographic dividend," he says. "And they're usually the ones that are in charge of a family's daily consumption." Only 56% of 1.4 billion Chinese -- about 772 million people -- use the internet, according to official data. The U.S. reached that level of penetration in 2002, according to the United Nations. Interest in the lower end of the internet market has been building; live-streaming services have managed to attract working-class Chinese. This time around, the spread of e-commerce and new business models are unlocking more potential.

Ever since the Tempe police released a video of Uber's self-driving car hitting and killing a pedestrian, experts have been racing to analyze the footage and determine what exactly went wrong. (If you haven't watched the video, you can do so here. Warning: it's disturbing, though the actual impact is removed.) In a blog post, software architect and entrepreneur Brad Templeton highlights some of the big issues with the video:
1. On this empty road, the LIDAR is very capable of detecting her. If it was operating, there is no way that it did not detect her 3 to 4 seconds before the impact, if not earlier. She would have come into range just over 5 seconds before impact.
2.On the dash-cam style video, we only see her 1.5 seconds before impact. However, the human eye and quality cameras have a much better dynamic range than this video, and should have also been able to see her even before 5 seconds. From just the dash-cam video, no human could brake in time with just 1.5 seconds warning. The best humans react in just under a second, many take 1.5 to 2.5 seconds.
3. The human safety driver did not see her because she was not looking at the road. She seems to spend most of the time before the accident looking down to her right, in a style that suggests looking at a phone.
4.While a basic radar which filters out objects which are not moving towards the car would not necessarily see her, a more advanced radar also should have detected her and her bicycle (though triggered no braking) as soon as she entered the lane to the left, probably 4 seconds before impact at least. Braking could trigger 2 seconds before, in theory enough time.)

To be clear, while the car had the right-of-way and the victim was clearly unwise to cross there, especially without checking regularly in the direction of traffic, this is a situation where any properly operating robocar following "good practices," let alone "best practices," should have avoided the accident regardless of pedestrian error. That would not be true if the pedestrian were crossing the other way, moving immediately into the right lane from the right sidewalk. In that case no technique could have avoided the event. The overall consensus among experts is that one or several pieces of the driverless system may have failed, from the LIDAR system to the logic system that's supposed to identify road objects, to the communications channels that are supposed to apply the brakes, or the car's automatic braking system itself. According to Los Angeles Times, "Driverless car experts from law and academia called on Uber to release technical details of the accident so objective researchers can help figure out what went wrong and relay their findings to other driverless system makers and to the public."

Robots have taken over many of America's factories. They can explore the depths of the ocean, and other planets. They can play ping-pong. But can they pick a strawberry? From a report: "You kind of learn, when you get into this -- it's really hard to match what humans can do," says Bob Pitzer, an expert on robots and co-founder of a company called Harvest CROO Robotics. (CROO is an acronym. It stands for Computerized Robotic Optimized Obtainer.) Any 4-year old can pick a strawberry, but machines, for all their artificial intelligence, can't seem to figure it out. Pitzer says the hardest thing for them is just finding the fruit. The berries hide behind leaves in unpredictable places. "You know, I used to work in the semiconductor industry. I was a development engineer for Intel, and it was a lot easier to make semiconductor chips," he says with a laugh.

As smartphones and other mobile devices have become more widespread, 26 percent of American adults now report that they go online "almost constantly," up from 21 percent in 2015, according to a Pew Research Center survey conducted in January 2018. From the study: Overall, 77 percent of Americans go online on a daily basis. That figure includes the 26 percent who go online almost constantly, as well as 43 percent who say they go online several times a day and 8 percent who go online about once a day. Some 11 percent go online several times a week or less often, while 11 percent of adults say they do not use the internet at all.