In the midst of this exhausting week in the news, many questions have been asked on this here internet: Are Bert and Ernie gay? Do we really need Space Jam 2? Is Rihanna really an ambassador of Barbados? The answer to all of these questions, dear friends, is yes. But there are other questions, and other topics, that require a little bit more focus, and that’s why we’re all here. Step right up to discover what the internet has been talking about over the past seven days.
Today’s artificial intelligence systems, including the artificial neural networks broadly inspired by the neurons and connections of the nervous system, perform wonderfully at tasks with known constraints. They also tend to require a lot of computational power and vast quantities of training data. That all serves to make them great at playing chess or Go, at detecting if there’s a car in an image, at differentiating between depictions of cats and dogs. “But they are rather pathetic at composing music or writing short stories,” said Konrad Kording, a computational neuroscientist at the University of Pennsylvania. “They have great trouble reasoning meaningfully in the world.”
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.
Cheat Sheet: What’s the Deal With Location-Based VR
Sorry, skeptics: Virtual reality is growing faster than you thought it would. You’re just looking in the wrong place. Millions of people are leaving their homes for the visceral experience of location-based VR, which has become a billion-dollar business.
You’ve read your last complimentary article this month. To read the full article, SUBSCRIBE NOW. If you’re already a subscriber, please sign in and and verify your subscription.
25 Years of WIRED
WIRED is turning 25! We are celebrating in San Francisco this October with four days of events honoring the ideas, innovations, and icons who have shaped the world we know today—and those who will shape it for the 25 years to come.
The Markup, dedicated to investigating technology and its effect on society, will be led by two former ProPublica journalists. Craig Newmark gave $20 million to help fund the operation.
When the investigative journalist Julia Angwin worked for ProPublica, the nonprofit news organization became known as “big tech’s scariest watchdog.”
By partnering with programmers and data scientists, Ms. Angwin pioneered the work of studying big tech’s algorithms — the secret codes that have an enormous impact on everyday American life. Her findings shed light on how companies like Facebook were creating tools that could be used to promote racial bias, fraudulent schemes and extremist content.
Now, with a $20 million gift from the Craigslist founder Craig Newmark, she and her partner at ProPublica, the data journalist Jeff Larson, are starting The Markup, a news site dedicated to investigating technology and its effect on society. Sue Gardner, former head of the Wikimedia Foundation, which hosts Wikipedia, will be The Markup’s executive director. Ms. Angwin and Mr. Larson said that they would hire two dozen journalists for its New York office and that stories would start going up on the website in early 2019. The group has also raised $2 million from the John S. and James L. Knight Foundation, and $1 million collectively from the Ford Foundation, the John D. and Catherine T. MacArthur Foundation, and the Ethics and Governance of Artificial Intelligence Initiative.
Ms. Angwin compares tech to canned food, an innovation that took some time to be seen with more scrutiny.
“When canned food came out, it was amazing,” said Ms. Angwin, who will be the site’s editor in chief. “You could have peaches when they were out of season. There was a whole period of America where every recipe called for canned soup. People went crazy for canned food. And after 30 years, 40 years, people were like, ‘Huh, wait.’
“That is what’s happened with technology,” Ms. Angwin said, calling the 2016 election a tipping point. “And I’m so glad we’ve woken up.”
The site will explore three broad investigative categories: how profiling software discriminates against the poor and other vulnerable groups; internet health and infections like bots, scams and misinformation; and the awesome power of the tech companies. The Markup will release all its stories under a creative commons license so other organizations can republish them, as ProPublica does.
Ms. Angwin, who was part of a Wall Street Journal team that won a Pulitzer Prize in 2003 for coverage of corporate corruption, said the newsroom would be guided by the scientific method and each story would begin with a hypothesis. For example: Facebook is allowing racist housing ads. At ProPublica, Ms. Angwin’s team bought ads on the site and proved the hypothesis.
At The Markup, journalists will be partnered with a programmer from a story’s inception until its completion.
“To investigate technology, you need to understand technology,” said Ms. Angwin, 47. “Just like I got an M.B.A. when I was a business reporter, I believe that technologists need to be involved from the very beginning of tech investigations.”
Ms. Angwin has known Mr. Newmark since 1997, when she wrote about him while a reporter at The San Francisco Chronicle.
“Craig is ideal for us because he has no interest or temperament for trying to interfere in coverage,” she said.
Mr. Newmark, who splits his time between San Francisco and New York, has for years kept a low profile. But he worries about what he sees as a lack of self-reflection among engineers.
“Sometimes it takes an engineer a while to understand that we need help, then we get that help, and then we do a lot better,” Mr. Newmark said. “We need the help that only investigative reporting with good data science can provide.”
Craigslist, which Mr. Newmark founded in the mid-1990s, helped to decimate print newspapers’ main source of revenue at the time: classified advertising. Recently, he has given several substantial donations to journalistic institutions, including $20 million to the CUNY Graduate School of Journalism.
“We’re in an information war now,” Mr. Newmark said.
For many years, the outrageous success of Silicon Valley companies — and the aggressive public relations teams who worked for them — kept many journalists at a remove.
The societal effects of tech were hard to quantify, and moral responsibility was often sloughed off on something called an algorithm, which most people could not quite explain or examine. Even if, as in the case of Facebook, it influenced around 2.5 billion people.
At ProPublica, Ms. Angwin and Mr. Larson subverted the traditional model of tech reporting altogether. They did not need access. With the right tools, they could study impact.
“There’s an opportunity for more reporters to use statistics to uncover societal harms,” said Mr. Larson, who has been doing data-driven journalism for a decade. “And then Julia’s gift is she takes data journalism and doesn’t make it like an academic report.”
Some of Ms. Angwin and Mr. Larson’s reporting tactics may violate tech platform terms of service agreements, which ban people from performing automated collection of public information and prohibit them from creating temporary research accounts. Ms. Angwin has been a strong defender of these practices and has argued that tech companies ought to allow reporters to be an exception to their rules.
“Without violating those rules, journalists can’t investigate our most important platform for public discourse,”Ms. Angwin wrote in August.
Mr. Larson, who will be The Markup’s managing editor, said the result was just as much a surprise to readers as it was to those who had made the biased algorithm.
“Increasingly, algorithms are used as shorthand for passing the buck,” said Mr. Larson, 36. “We don’t have enough people to look at parole decisions, so we’re going to pass it on to the computer and the computer is going to decide, and once they go into production, there’s no oversight.”
“There are unintended consequences,” Mr. Larson said. “In all three of those cases, it was a complete surprise to the people who made those algorithms as well.”
Engineers being surprised by the tools they have made is, to the Markup team, part of the problem.
“Part of the premise of The Markup is the level of understanding technology and its effects is very, very low, and we would all benefit from a broader understanding,” Ms. Gardner said. “And I would include people who work for the companies.”
Ms. Angwin said part of her goal was to help readers understand what exactly they should be worried about when it comes to tech.
“We’re all a little uncertain,” Ms. Angwin said. “The evidence isn’t in. I want to be providing the evidence.”
She hopes the stories they take on will lead to better government and corporate policies.
“We are a numbers-driven data society,” Ms. Angwin said. “That’s the price of entry these days for political change — a data set.”
And searching for that information, Ms. Angwin said she was not worried about getting Facebook or Google to return her phone calls.
“I’ve never been on Google’s or Facebook’s campus and I imagine I’ll never be invited,” she said. “I’m kind of a dorky scientist just over here measuring stuff.”
Nellie Bowles covers tech and internet culture. Follow her on Twitter: @nelliebowles
A version of this article appears in print on
, on Page
of the New York edition
with the headline:
A News Site Dedicated To Peering at Big Tech And Its Surprise Effects
Modifications to Google Chromium for removing Google integration and enhancing privacy, control, and transparency
Bringing back the “Don’t” in “Don’t be evil”
ungoogled-chromium is Google Chromium, sans integration with Google. It also features some changes to enhance privacy, control, and transparency.
Hexbyte Hacker News Computers Motivation and Description
A number of features or background services communicate with Google servers despite the absence of an associated Google account or compiled-in Google API keys. Furthermore, the normal build process for Chromium involves running Google’s own high-level commands that invoke many scripts and utilities, some of which download and use pre-built binaries provided by Google. Even the final build output includes some pre-built binaries. Fortunately, the source code is available for everything.
ungoogled-chromium is a set of configuration flags, patches, and custom scripts. These components altogether strive to accomplish the following:
Disable or remove offending services and features that communicate with Google or weaken privacy
Strip binaries from the source tree, and use those provided by the system or build them from source
Disable features that inhibit control and transparency, and add or modify features that promote them (these changes are minor and do not have significant impacts on the general user experience)
ungoogled-chromium should not be considered a fork of Chromium. The main reason for this is that a fork is associated with more significant deviations from the Chromium, such as branding, configuration formats, file locations, and other interface changes. ungoogled-chromium will not modify the Chromium browser outside of the project’s goals.
Since these goals and requirements are not precise, unclear situations are discussed and decided on a case-by-case basis.
Disable functionality specific to Google domains (e.g. Google Host Detector, Google URL Tracker, Google Cloud Messaging, Google Hotwording, etc.)
Add Omnibox search provider “No Search” to allow disabling of searching
Disable automatic formatting of URLs in Omnibox (e.g. stripping http://, hiding certain parameters)
Added menu item under “More tools” to clear the HTTP authentication cache on-demand
Add new command-line switches and chrome://flags entries:
--disable-search-engine-collection – Disable automatic search engine scraping from webpages.
--enable-stacked-tab-strip and --enable-tab-adjust-layout – These flags adjust the tab strip behavior. --enable-stacked-tab-strip is also configurable in chrome://flags Please note that they are not well tested, so proceed with caution.
--extension-mime-request-handling – Change how extension MIME types (CRX and user scripts) are handled. Acceptable values are download-as-regular-file or install-always. Leave unset to use normal behavior. It is also configurable under chrome://flags
--fingerprinting-canvas-measuretext-noise (Added flag to Bromite feature) – Scale the output values of Canvas::measureText() with a randomly selected factor in the range -0.0003% to 0.0003%, which are recomputed on every document initialization.
--fingerprinting-client-rects-noise (Added flag to Bromite feature) – Implements fingerprinting deception of JS APIs getClientRects() and getBoundingClientRect() by scaling their output values with a random factor in the range -0.0003% to 0.0003%, which are recomputed for every document instantiation.
--fingerprinting-canvas-image-data-noise (Added flag to Bromite feature) – Implements fingerprinting deception for Canvas image data retrieved via JS APIs. In the data, at most 10 pixels are slightly modified.
--max-connections-per-host (from Bromite) – Configure the maximum allowed connections per host.
--set-ipv6-probe-false – (Not in chrome://flags) Forces the result of the browser’s IPv6 probing (i.e. IPv6 connectivity test) to be unsuccessful. This causes IPv4 addresses to be prioritized over IPv6 addresses. Without this flag, the probing result is set to be successful, which causes IPv6 to be used over IPv4 when possible.
Please submit feedback (i.e. problems, suggestions, and questions) to the Issue Tracker. The Issue Tracker is the main hub for development activity.
There are chat room options available via Gitter and Matrix.org (name ungoogled-software/lobby, which has bidirectional connection with Gitter). They are optional; it is only provided for those who prefer this format.
Unless there are significant benefits, adding the setting to chrome://settings is not recommended due to the additional maintenance required (caused by the infrastructure that backs preferences).
NOTE: In the event that the codebase changes significantly for a non-essential patch (i.e. a patch that does not contribute to the main objectives of ungoogled-chromium), it will be removed until someone updates it.
In the midst of my September job hop I headed to Kazan for the weekend. I don’t know exactly why — probably because I could. I had a hotel booked via booking.com, but once I arrived there the receptionist told me it was the first time he’s heard of my booking, and he told me that they had no more rooms, and he told me that I’d rather find another place to stay.
But why, I always see these things for what they are. Some sneaky caffeinated programmer kids at work with the booking system integrations, you should always double-check after us! After what came next, I started suspecting there’s some particularly defective developer marketing himself as a “Chief Hotel Booking Management System (HBMS) Professional with 10+ years experience in the field”.
I booked the next hotel for the two nights, with a discount, and walked there. Guess what? The room I booked was occupied, but they had another one, but not for two night, just one. In the end it worked out all right because there was that woman using the responsive, fully interactive, enterprise-ready ERP system “sheet of paper and a pencil” that allows you to reschedule bookings at will, with no programmers involved.
The rest of my trip went well, probably because I had little interaction with software stuff. The pizza place charged me twice, because they thought the first payment didn’t pass, but they gave the money back. Human problem, wasn’t it? (Of course not completely, some usability failure out there, but let’s leave it for what it is, shall we?)
I was riding the train from the airport, almost eager to get back to work writing things that mostly function. I was a bit worried they didn’t give me a call — was I supposed to just come to the office and figure out what to do next? Anyways, I made some calls and found out that the information about me starting my work got lost somewhere along the course of a month it’s been running around the infrastructure.
This got me a little worried: I haven’t spent a week in Moscow without work in over three years, and, besides, I was running low on money. I wasn’t angry with anyone in particular: who should I blame for the state of the industry? I’m just as guilty as anyone for letting it fall that low. it turned out not to be taht bad after all: the weather was good, and I got a chance to walk the empty city — everyone here in Moscow is too busy working hard to be out of office in the middle of the day.
My accidental vacation started along the fanfare of Nikita Prokopov’s software disenchantment. It’s depressing but, and even more depressive for being true. Let me pick a quote for you:
We cover shit with blankets just not to deal with it.
I’ve been subscribed to this blog for several years, and never knew the man was Russian. I should have guessed — where else can a man that depressed and dissatisfied with the state-of-the-world come from? The part that caught my atteniton was the Russian apps on the screenshots. From there I got onto Nikita’s russian-language blog.
Why must it be like this?
It’s a good the blog is in Russian — serves well to save the good people of the world from moscow-grade sadness. Somewhere in the comments on the fourth page I found a neat explanation for why the things in software are the way they are. I can’t find the particular comment any more, but I remember the gist and even spent some time elaborating on it.
Once a problem is solved by software at a minimum viable level, you don’t need programmers (we are crazy expensive) to actually make it work. The solution gets canned and distributed in a bundle that usually works, as long as you’re not getting too tricky:
Edit a table of numbers? Excel!
Write a text? Word!
Send some data to another person? E-mail!
Make a website? WordPress!
Edit an image? Photoshop!
The programmers are there to stick them into the holes of the abstraction. If you’re generating excel tables programatically from a database, or want your website to synchronize with a price list from a Googledoc, you need some programmers to duct-tape the systems together in unexpected ways (I’m working hard to stay away from shit-based metaphors, but feel free to make one yourself). These systems have a long train of backwards compatibility, they are a massive overkill for the job, but hey, they work just well enough.
I’m enraged by the lack of attention to static website generators. The idea is beautiful: pipe some markdowns through a template, get a set of static HTMLs and resources that can be host anywhere for, like, free! Why do people make their websites in WordPress — have a MySQL database at work pushing the limits of storing static data, and PHP, rendering the pages at every request. Oh dear, the system was supposed to write blogs — how did they ever manage to bend it into e-commerce, landing pages and ERP front-ends?
No, I do know the answer to that — just hit a freelance farm to find a bunch of schoolchildren who can make you something that looks like a website in a week, for $50. I wouldn’t consider disrupting a business with that low profit margin — and who would?
On Thursday I was watching my TV when the electricity died. It spent the evening randomly switching it on and off, then got stuck loading forever. The philips tech support was very sympathetic. “There’s some very high tech in there — a whole embedded system! How in the world did you expect it to survive the electricity shutdown?” they said. And yeah, sure, what did I expect from a system that had a piece of software in it?
Really, I should assume I write code at least marginally better-than-average and get into programming education. There’s a horrible shortage of material on learning actual programming (not “How to Use Technology X to Solve Problem Y”-style things) out there. I must to fight as well as I possibly can. You must, too (yes, you, the guy who actually read this all).
You’ve just seen a weekly snapshot of my thoughts pile. While taking another look at this website’s design to see if the the navigation is obscure enough to prevent the users from noticing the other posts (it probably is; I don’t think i’m changing it just yet) I noticed something horrible.
The blog’s address has thoughts pile in it. How is that name appropriate for the smart, boring and technical stuff I’ve come to post? I need to give the place some wildness it deserves. Klyukovka, ay, igray Balalaechka! Writing messy speculative posts is far more entertaining!
A new Windows version for multiple users was spotted last month, and now we know what it’s for: Windows Virtual Desktop (WVD) is a new service providing multi-user remote desktop and VDI in the Azure cloud.
WVD combines three things. Using the new Windows 10 version, WVD can be used to provide remote desktop sessions with multiple users remotely logged in to the same Windows 10 virtual machine (or, alternatively, a Windows Server virtual machine). This can provide both remoting of a full desktop session and of individual applications, serving as a replacement for the RemoteApp service that Microsoft cancelled last year. The service also supports full VDI, with remote users each having their own single-user virtual machine while both persistent and non-persistent VMs are supported. This is supported both with Windows 10 and with Windows 7.
Licenses for WVD will be an integrated, no-additional-cost part of Windows Enterprise E3 licenses. This will enable, for example, a local Windows 10 installation that uses WVD for remote access to a couple of legacy applications running on Windows 7 on Azure with no additional Windows licensing requirements.
This service addresses certain gaps that have previously existed in Windows virtualization. While Windows Server has long supported this multi-user style of operation, offering many concurrent remote desktop sessions at a time, the Windows Server desktop experience isn’t the same as the one in Windows 10. It doesn’t have Edge, Store, or Cortana, for example, and it gets updated only every three years (on Microsoft’s LTSC update cycle) rather than every six months (on the SAC update cycle). This means tha
Microsoft’s unveiling of the Surface Hub 2 in May this year provoked a very positive reaction. So much so that the company says it’s going to shake up its plans with a staggered launch.
The Surface Hub 2 as originally envisaged is a 50.5-inch 4K display with a 3:2 aspect ratio, a webcam, and a touchscreen. It builds on the existing Surface Hub in a couple of key areas: the whole screen can rotate to be used in either orientation; multiple devices can be tiled together to build a giant screen that operates as one; and its software will support multiple user accounts.
That’s still ultimately the plan, but the way it’s going to be delivered is a little more complicated. In the second quarter of 2019, Microsoft will release the Surface Hub 2S. That offers the new display and form factor but running the current Surface Hub software—meaning no rotation, no tiling, and no multiuser. It’s essentially a sleeker, faster versi
Applications using Azure Active Directory (AD) to authenticate—a category that includes Office 365, among other things—will soon be able to stop using passwords entirely.
Azure AD accounts can already use the Microsoft Authenticator app for two factor authentication, combining a password with a one-time code. With the new passwordless support, authentication is handled entirely by the app; the app itself represents “something you have,” and this is combined with either biometric authentication or a PIN. Passwords have a long, problematic history; while they can be very strong, if suitably long and suitably random, human passwords are often short, non-random, and reused across multiple sites. App-based authentication avoids this long-standing weakness.
Enabling two-factor authentication is just one of the things that organizations can do to improve their security. To that end, Microsoft has extended “Microsoft Security Score,” a tool used to assess organizational policy and provide guidance on measures that can b
Roku announced two new streaming devices today that sit in the middle of its device lineup. The Roku Premiere and Premiere+ set-top boxes are barely “boxes” at all; instead, they resemble the company’s streaming sticks more than any of its other devices.
If you took the Roku Ultra, the company’s top-tier device, and slashed it in half and shrank it a bit, you’d get the Roku Premiere and Premiere+. The streaming devices are about the length of your index finger and the width of two fingers, making them lightweight and nearly invisible when sitting on an entertainment console while connected to a TV. The front side is a glossy black while the flat back side holds an HDMI port and the power port.
The Premiere devices basically replace the old Roku 2 and Roku 3 devices, all of which took up more space with their boxier designs. For those who don’t want a device as big or expensive as the Roku Ultra, they can get either the Premiere or the Premiere+, both of which have quad-core processors and support HD, 4K, and 4K HDR streaming.
The ability to accept voice commands differentiates the Premiere+ from the Premiere: the former comes with Roku’s voice remote, allowing users to press a button and say things like “show me stand-up comedy” and “show me TV dramas.” Voice control is a feature that has become ubiquitous across entertainment tech, pa