Science

We All Nearly Missed the Largest Underwater Volcano Eruption Ever Recorded (sciencealert.com) 34

schwit1 quotes ScienceAlert: She was flying home from a holiday in Samoa when she saw it through the airplane window: a "peculiar large mass" floating on the ocean, hundreds of kilometres off the north coast of New Zealand. The Kiwi passenger emailed photos of the strange ocean slick to scientists, who realised what it was -- a raft of floating rock spewed from an underwater volcano, produced in the largest eruption of its kind ever recorded.

"We knew it was a large-scale eruption, approximately equivalent to the biggest eruption we've seen on land in the 20th Century," says volcanologist Rebecca Carey from the University of Tasmania, who's co-led the first close-up investigation of the historic 2012 eruption. The incident, produced by a submarine volcano called the Havre Seamount, initially went unnoticed by scientists, but the floating rock platform it generated was harder to miss. Back in 2012, the raft -- composed of pumice rock -- covered some 400 square kilometres (154 square miles) of the south-west Pacific Ocean, but months later satellites recorded it dispersing over an area twice the size of New Zealand itself... for a sense of scale, think roughly 1.5 times larger than the 1980 eruption of Mount St. Helens -- or 10 times the size of the 2010 Eyjafjallajokull eruption in Iceland.

When an underwater robot first sent back detailed maps, one volcanologist remembers that "I thought the vehicle's sonar was acting up... We saw all these bumps on the seafloor... It turned out that each bump was a giant block of pumice, some of them the size of a van."
Crime

Software 'No More Accurate Than Untrained Humans' At Predicting Recidivism (theguardian.com) 162

An anonymous reader quotes a report from The Guardian: The credibility of a computer program used for bail and sentencing decisions has been called into question after it was found to be no more accurate at predicting the risk of reoffending than people with no criminal justice experience provided with only the defendant's age, sex and criminal history. The algorithm, called Compas (Correctional Offender Management Profiling for Alternative Sanctions), is used throughout the U.S. to weigh up whether defendants awaiting trial or sentencing are at too much risk of reoffending to be released on bail. Since being developed in 1998, the tool is reported to have been used to assess more than one million defendants. But a new paper has cast doubt on whether the software's predictions are sufficiently accurate to justify its use in potentially life-changing decisions.

The academics used a database of more than 7,000 pretrial defendants from Broward County, Florida, which included individual demographic information, age, sex, criminal history and arrest record in the two year period following the Compas scoring. The online workers were given short descriptions that included a defendant's sex, age, and previous criminal history and asked whether they thought they would reoffend. Using far less information than Compas (seven variables versus 137), when the results were pooled the humans were accurate in 67% of cases, compared to the 65% accuracy of Compas. In a second analysis, the paper found that Compas's accuracy at predicting recidivism could also be matched using a simple calculation involving only an offender's age and the number of prior convictions.

The Courts

US Supreme Court Will Revisit Ruling On Collecting Internet Sales Tax (theverge.com) 178

An anonymous reader quotes a report from Bloomberg: The U.S. Supreme Court will consider freeing state and local governments to collect billions of dollars in sales taxes from online retailers, agreeing to revisit a 26-year-old ruling that has made much of the internet a tax-free zone. Heeding calls from traditional retailers and dozens of states, the justices said they'll hear South Dakota's contention that the 1992 ruling is obsolete in the e-commerce era and should be overturned. State and local governments could have collected up to $13 billion more in 2017 if they'd been allowed to require sales tax payments from online merchants and other remote sellers, according to a report from the Government Accountability Office, Congress's non-partisan audit and research agency. Other estimates are even higher. All but five states impose sales taxes.

The high court's 1992 Quill v. North Dakota ruling, which involved a mail-order company, said retailers can be forced to collect taxes only in states where the company has a "physical presence." The court invoked the so-called dormant commerce clause, a judge-created legal doctrine that bars states from interfering with interstate commerce unless authorized by Congress. South Dakota passed its law in 2016 with an eye toward overturning the Quill decision. It requires retailers with more than $100,000 in annual sales in the state to pay a 4.5 percent tax on purchases. Soon after enacting the law, the state filed suit and asked the courts to declare the measure constitutional.

Medicine

Scientists Change Our Understanding of How Anaesthesia Messes With the Brain (sciencealert.com) 92

schwit1 shares a report from ScienceAlert: It's crazy to think that we still don't quite understand the mechanism behind one of the most common medical interventions -- general anaesthetic. But researchers in Australia just got a step closer by discovering that one of the most commonly used anesthetic drugs doesn't just put us to sleep; it also disrupts communication between brain cells. The team investigated the drug propofol, a super-popular option for surgeries worldwide. A potent sedative, the drug is thought to put us to sleep through its effect on the GABA neurotransmitter system, the main regulator of our sleep-and-wake cycles in the brain. But anyone who's been "put under" will know that waking up from a general anesthetic feels rather different from your usual morning grogginess. On top of that, some people can experience serious side-effects, so scientists have been trying to figure out what else the drugs might be doing in the brain.

Using live neuron cell samples from rats and fruit flies, the researchers were able to track neurotransmitter activity thanks to a super-resolution microscope, and discovered that propofol messes with a key protein that nerve cells use to communicate with each other. This protein, called syntaxin1A, isn't just found in animal models - people have it, too. And it looks like the anesthetic drug puts the brakes on this protein, making otherwise normal brain cell connections sluggish, at least for a while. The researchers think this disruption could be key to how propofol allows for pain-free surgery to take place - first it knocks us out as a normal sleeping pill would, and then takes things up a notch by disrupting brain connectivity.
The research has been published in Cell Reports.
Networking

Can Mesh Networks Save a Dying Web? (thenextweb.com) 201

From an anonymous reader: "The web is dying, but mesh networks could save it," writes open source hacker Andre Staltz. He warns that Facebook, Google, and Amazon plan to "grow beyond browsers, creating new virtual contexts where data is created and shared," and predicts the next wave of walled gardens will be a "social internet" bypassing the web altogether. "The Web may die like most other technologies do, simply by becoming less attractive than newer technologies."

He wants to build a mobile mesh web that works with or without internet access to reach the four billion people currently offline, adding that all the tools we need are already in our hands: smartphones, peer-to-peer protocols, and mesh networks. His vision? "Novel peer-to-peer protocols such as IPFS and Dat help replace HTTP and make the web a content-centered cyberspace... Browsers can be made to work like that, and although it's a small tweak to how the web works, it has massive effects on social structures in cyberspace... Now that we have experience with some of the intricacies of the social web, we can reinvent it to put people first without intermediate companies... We can actually beat the tech giants at this game by simply giving local and regional connectivity to people in developing countries. With mobile apps that are built mesh-first, the smartphones would make up self-organizing self-healing mobile ad-hoc networks... In internet-less regions, there is potential for scaling quickly, and through that, we can spawn a new industry around peer-to-peer wireless mesh networks."

He cites mega-projects "to rescue the web from the internet", which include progress on peer-to-peer and mesh networking protocols, followed by adoption on smartphones (and then a new wave of apps) -- plus a migration of existing web content to the new protocols, "to fix the overutilization of the wirenet and the underutilization of airnets, bringing balance to the wire-versus-air dichotomy, providing choice in how data should travel in each case...But it can only happen if the web takes a courageous step towards its next level."

Technology

That '70s Show: the Conference That Predicted the Future of Work (wired.com) 40

theodp writes: Over at Wired, Leslie Berlin writes about Futures Day at the 1977 Xerox World Conference, an invitation-only demonstration of the Alto personal computer system developed at Xerox PARC. It's an excerpt from Troublemakers: How a Generation of Silicon Valley Upstarts Invented the Future. Both Berlin's book and Brian Dear's recent The Friendly Orange Glow: The Untold Story of the PLATO System and the Dawn of Cyberculture are shedding light on groundbreaking systems of the '70s that were ultimately done in by the less-featured but low-cost Apple II (yes, $2,638 for a system with 48 kB of RAM was 'low cost'!) and other personal computers. Interestingly, Dear notes that the Xerox Parc and PLATO teams sent people out to see and learn and exchange ideas with each other over the years. Their interactions included 'tremendous battles' over the advantages and disadvantages of mouse interfaces [Xerox] vs. touch screens [PLATO], as well as plasma displays [PLATO] vs. other, cheaper display solutions [Xerox]. As is the case with many debates, both teams proved to be "right." Apple wouldn't introduce the masses to a mouse interface until 1984 [Macintosh] and a touch screen interface until 2007 [iPhone].
Books

A Book Recommendation for Bill Gates: The Story of PLATO 59

Long-time Slashdot reader theodp writes: This holiday season, many Slashdot readers are likely to find gifts under the tree because of Bill Gates' book picks. Not that there's anything wrong with that, but it seems that turnabout is fair play -- what book recommendations do you have for Bill?

At the top of my pick list for personalized learning advocate Gates would be Brian Dear's remarkable The Friendly Orange Glow: The Untold Story of the PLATO System and the Dawn of Cyberculture, with its tale of how a group of visionary engineers and designers -- some of them only high school students -- created a shockingly little-known computer system called PLATO in the late 1960s and 1970s that was decades ahead of its time in experimenting with how people could learn, engage, communicate, and play through connected terminals and computers. After all, "we can't move forward," as Audrey Watters argued in The Hidden History of Ed-Tech, "til we reconcile where we've been before."
Electronic Frontier Foundation

EFF: Accessing Publicly Available Information On the Internet Is Not a Crime (eff.org) 175

An anonymous reader quotes a report from EFF: EFF is fighting another attempt by a giant corporation to take advantage of our poorly drafted federal computer crime statute for commercial advantage -- without any regard for the impact on the rest of us. This time the culprit is LinkedIn. The social networking giant wants violations of its corporate policy against using automated scripts to access public information on its website to count as felony "hacking" under the Computer Fraud and Abuse Act, a 1986 federal law meant to criminalize breaking into private computer systems to access non-public information.

EFF, together with our friends DuckDuckGo and the Internet Archive, have urged the Ninth Circuit Court of Appeals to reject LinkedIn's request to transform the CFAA from a law meant to target "hacking" into a tool for enforcing its computer use policies. Using automated scripts to access publicly available data is not "hacking," and neither is violating a website's terms of use. LinkedIn would have the court believe that all "bots" are bad, but they're actually a common and necessary part of the Internet. "Good bots" were responsible for 23 percent of Web traffic in 2016. Using them to access publicly available information on the open Internet should not be punishable by years in federal prison. LinkedIn's position would undermine open access to information online, a hallmark of today's Internet, and threaten socially valuable bots that journalists, researchers, and Internet users around the world rely on every day -- all in the name of preserving LinkedIn's advantage over a competing service. The Ninth Circuit should make sure that doesn't happen.

Transportation

Inside Faraday Future's Financial House of Cards (theverge.com) 44

Sean O'Kane, reporting for The Verge: When Faraday Future emerged from stealth mode in 2015, it promised to transform the car industry with an American-made luxury electric vehicle that would someday be fully autonomous, maybe even sold through a subscription service. As we learned at CES 2017, the company was taking aim at Tesla with a car -- the FF91 -- that was designed to dazzle, with a 0-60 time of 2.4 seconds as jaw-dropping as the proposed $180,000 price tag. Since then, though, Faraday Future has been more focused on survival than speed. The Verge has learned from multiple sources about the nature of the company's financial plight. While Faraday Future posed as the newest California electric car startup that attracted top auto industry talent, 10 former employees and one person close to the company say the behavior and business practices of its chief investor have brought business to a halt. Also read: Everything wrong with Faraday Future's "Tesla killer"
Space

New Satellite Experiment Helps Confirm Einstein's Equivalence Principle (presse.cnes.fr) 71

Part of Einstein's theory of general relativity posits that gravity equals inertial mass -- and for the first time in 10 years, there's new evidence that he's right. Slashdot reader orsayman reports: Most stories around space today seem to revolve around SpaceX, but let's not forget that space is also a place for cool physics experiments. One such experiment currently running into low orbit is the MICROSCOPE satellite launched in 2016 to test the (weak) Equivalence Principle (also knows as the universality of free fall) a central hypothesis in General Relativity.

The first results confirm the principle with a precision ten times better than previous experiments. And it's just the beginning since they hope to increase the precision by another factor of 10. If the Equivalence Principle is still verified at this precision, this could constrain or invalidate some quantum gravity theories. For those of you who are more satellite-science oriented, the satellite also features an innovative "self destruct" mechanism (meant to limit orbit pollution) based on inflatable structures described in this paper.

"The science phase of the mission began in December 2016," reports France's space agency, "and has already collected data from 1,900 orbits, the equivalent of a free fall of 85 million kilometres or half the Earth-Sun distance."
The Almighty Buck

'We Could Fund a Universal Basic Income With the Data We Give Away To Facebook and Google' (thenextweb.com) 588

Tristan Greene reports via The Next Web: A universal basic income (UBI), wherein government provides a monthly stipend so citizens can afford a home and basic necessities, is something experts believe would directly address the issue of unemployment and poverty, and possibly even eliminate hundreds of other welfare programs. It may also be the only real solution to the impending automation bonanza. According to AI expert Steve Fuller, the problem is, giving people money when they lose jobs won't fix the issue, it's a temporary solution and we need permanent ones. Sounds fair, and he even has some ideas on how to accomplish this end: "We could hold Google and Facebook and all those big multinationals accountable; we could make sure that people, like those who are currently 'voluntarily' contributing their data to pump up companies' profits, are given something that is adequate to support their livelihoods in exchange."

It's an interesting idea, but difficult to imagine it's implementation. If the government isn't assigning a specific stipend value, we'll have to be compensated individually by companies. One way to do this, is by emulating the old coal mining company scrip scams of early last century. Employees working for companies would be paid in currency only redeemable at the company store. This basically created a system where a company could tax its own workers for profit. Google, for example, could use a system like that and say "opt-in for $10 worth of Google Play music for free," if they wanted to. Which doesn't help pay the bills when machines replace you at work, but at least you'll be able to voice search for your favorite songs. Another idea is to charge companies an automation tax, but again there's concerns as to how this would be implemented. A solution that combines government oversight with a tax on AI companies -- a UBI funded by the dividends of our data -- may be the best option. To be blunt: we should make Google, Microsoft, Facebook and other such AI companies pay for it with a simple data tax.

Facebook

Health Secretary Hits Out at Facebook's New App, Says 'Stay Away From My Kids' (theguardian.com) 113

Jeremy Hunt has publicly attacked Facebook for releasing a version of its Messenger app aimed at children, and called on the social media company to "stay away from my kids." From a report: The health secretary accused the company of "targeting younger children" after Facebook announced on Monday that it was conducting trials of an app called Messenger Kids in the US, which is designed to be used by pre-teens. He said the company was failing to act responsibly despite having assured the government that it would not target its service at children, who can only use the main social media website if they are over 13.
The Internet

Russia Wants To Launch Backup DNS System By August 1, 2018 (bleepingcomputer.com) 160

An anonymous reader shares a report from BleepingComputer: The Russian government plans to build its own "independent internet infrastructure" that will be used by BRICS member states -- Brazil, Russia, India, China, and South Africa. The plan was part of the topic list at the October meeting of the Russian Security Council, and President Vladimir Putin approved the initiative with a completion deadline of August 1, 2018, according to Russian news agency RT. The Russian Security Council has today formally asked the country's government to start the building of a backup global DNS system that Russia and fellow BRICS member states could use. The Russian Security Council cited the "increased capabilities of western nations to conduct offensive operations in the informational space." Russia, China, and many other countries have criticized the U.S. for hoarding control over the domain naming system (DNS), a position they claim has allowed the U.S. to intercept and tap global internet traffic. The U.S. has relinquished control over the DNS system last year.
Privacy

There's Now a Dark Web Version of Wikipedia (vice.com) 20

An anonymous reader shares a report: In many parts of the world, like North America, using Wikipedia is taken for granted; hell, there are even Twitter accounts to track government employees editing the internet's free encyclopedia while on the clock. But in other places, like Turkey or Syria, using Wikipedia can be difficult, and even dangerous. To make using Wikipedia safer for at-risk users, former Facebook security engineer Alec Muffett has started an experimental dark net Wikipedia service that gives visitors some strong privacy protections. The project is unofficial; for now, Wikipedia isn't involved. So it's a bit janky. The service uses self-signed certificates that may trigger a security warning in Tor, so you have to manually white-list the addresses, which takes a couple minutes.
Businesses

Thank You, Phish Fans, For Caring About Net Neutrality (theoutline.com) 79

If you venture over to Battle For the Net, which encourages internet users to call Congress to advocate for the preservation of net neutrality rules, you'll find something peculiar: Several of the top sites that direct calls are Phish-related. (Phish is an American rock band.) From a report: As someone on Twitter pointed out, the traffic from phish.net -- which describes itself as "a non-commercial project run by Phish fans and for Phish fans" -- appears to be coming from a pop-up message that greets visitors to the site. The same pop-up, which directs to www.battleforthenet.com, appears when you visit the site's forums and setlist pages. So, it appears that Phish fans, while in the midst of discussing their favorite extended noodling sessions, are leading the charge to save us from our impending telecom-dominated hellscape. Thanks, guys!" Phish.net sees over 400,000 unique visitors each month, according to web analytics firm SimilarWeb. In July, the website served over one million unique visitors.
Privacy

How a Wi-Fi Pineapple Can Steal Your Data (And How To Protect Yourself From It) (vice.com) 46

An anonymous reader writes: The Wi-Fi Pineapple is a cheap modified wireless router enables anyone to execute sophisticated exploits on Wi-Fi networks with little to no networking expertise. A report in Motherboard explains how it can be used to run a Wall of Sheep and execute a man-in-the-middle attack, as well as how you can protect yourself from Pineapple exploits when you're connected to public Wi-Fi. "... it's important that whenever you are done connecting to a public Wi-Fi network that you configure your phone or computer to 'forget' that network. This way your device won't be constantly broadcasting the SSIDs of networks it has connected to in the past, which can be spoofed by an attacker with a Pineapple," reports Motherboard. "Unfortunately there is no easy way to do this on an Android or an iPhone, and each network must be forgotten manually in the 'Manage Network' tab of the phone's settings. Another simple solution is to turn off your Wi-Fi functionality when you're not using it -- though that isn't as easy to do on some devices anymore -- and don't allow your device to connect to automatically connect to open Wi-Fi networks."
Censorship

Hitler Quote Controversy In the BSD Community 500

New submitter Seven Spirals writes: Recently, the FreeBSD folks have removed Fortune with a fairly predictable far right 4chan condemnation. Then last weekend saw a lively debate on NetBSD's current-users mailing list about the inclusion of Hitler quotes in the Fortune database with dozens of posts falling on the left and right. The quotes themselves are fairly tame material probably intended as cautionary. However, the controversy and the reaction of BSD users has been real and very diverse. So far, the result has been to pull Fortune out of FreeBSD and to relocate the quotes into the "offensive" database in NetBSD's case.
Businesses

Trump Administration Tightens Scrutiny of Skilled Worker Visa Applicants (inc.com) 263

wyattstorch516 writes: The Trump administration is tightening the scrutiny on the H-1B visa program (Warning: paywalled; alternative source). Changes would undo actions by the Obama administration. There are two big regulatory changes looming that would undo actions by the Obama administration. "The first change allowed spouses of H-1B workers the right to work. That regulation is being challenged in court and the Trump administration is expected to eliminate the provision rather than defend it," reports WSJ. "The second change affects the Optional Practical Training program, which allows foreign graduates from U.S. colleges in science and technology an extra two years of work authorization, giving them time to win an H-1B visa. The Trump administration could kill that benefit or reduce the two-year window, according to people familiar with the discussions." The Journal highlights a "series of more modest changes that have added scrutiny to visa processing":

- "USCIS directed last month that adjudicators no longer pay 'deference' to past determinations for renewal applications. This means an applicant's past approval won't carry any weight if he or she applies for a renewal.

- The agency is conducting more applicant interviews, which critics say slows the system. The agency spokesman says this process will ramp up over several years and is needed to detect fraud and make accurate decisions.

- In the spring, the agency suspended premium processing, which allowed for fast-track consideration to those who paid an extra fee. This option wasn't resumed until October, meaning many workers who qualified for a coveted H-1B visa had to wait months for a decision.

- State Department officials have been told to consider that Mr. Trump's 'Buy American, Hire American' executive order directs visa programs must 'protect the interests of United States workers.' And the Foreign Affairs Manual now instructs officers to scrutinize applications of students to ensure they plan to return to their home countries. A State Department official said the official rules haven't changed but said a 'comprehensive' review is under way."
AI

Deep Learning Is Eating Software (petewarden.com) 147

Pete Warden, engineer and CTO of Jetpac, shares his view on how deep learning is already starting to change some of the programming is done. From a blog post, shared by a reader last week: The pattern is that there's an existing software project doing data processing using explicit programming logic, and the team charged with maintaining it find they can replace it with a deep-learning-based solution. I can only point to examples within Alphabet that we've made public, like upgrading search ranking, data center energy usage, language translation, and solving Go, but these aren't rare exceptions internally. What I see is that almost any data processing system with non-trivial logic can be improved significantly by applying modern machine learning. This might sound less than dramatic when put in those terms, but it's a radical change in how we build software. Instead of writing and maintaining intricate, layered tangles of logic, the developer has to become a teacher, a curator of training data and an analyst of results. This is very, very different than the programming I was taught in school, but what gets me most excited is that it should be far more accessible than traditional coding, once the tooling catches up. The essence of the process is providing a lot of examples of inputs, and what you expect for the outputs. This doesn't require the same technical skills as traditional programming, but it does need a deep knowledge of the problem domain. That means motivated users of the software will be able to play much more of a direct role in building it than has ever been possible. In essence, the users are writing their own user stories and feeding them into the machinery to build what they want.
Education

Microsoft Debuts Minecraft-Themed Coding Tutorial 24

theodp writes: In a few weeks, writes Microsoft Corporate VP Mary Snapp, "millions of kids and others will participate in an Hour of Code, a global call to action to spend an hour learning the basics of coding. Today, it's my privilege to announce that Microsoft has released a new Minecraft tutorial for Hour of Code, called Hero's Journey." The release of the new Code.org-touted flagship Hour of Code tutorial -- the third since Microsoft purchased Minecraft Maker Mojang for $2.5B in 2014 -- comes as Microsoft celebrates Minecraft: Education Edition reaching a milestone of 2 million users.

Microsoft boasts that nearly 70 million of its Minecraft Hour of Code sessions have been launched to-date, which is certainly impressive from an infomercial or brand awareness standpoint. But does [adding a Scratch block to] move a Minecraft character forward 7 times on an $800 Microsoft Surface offer all that much more educational value than, say, moving a peg forward 5 times on a $10.99 Pop-O-Matic Trouble board game?

Slashdot Top Deals