Archives for July 2018

Why These Contractors Had To Help Apple Cut Air Pollution In China

Visitors and buyers at the Apple store in Macau, China, on July 18, 2018. Apple is under pressure to treat workers, and the air, well in China. (Photo by S3studio/Getty Images)

China is clearly keen to stop the smog. State media just reported on Monday that Beijing will close down about 1,000 manufacturers by 2020 as part of its campaign toward curbing air pollution and boosting income in other regions.

The announcement follows Apple’s recent move to introduce a clean energy fund in China that’s aimed at increasing the use of renewable energy across it supply chain. The tech giant’s efforts to help the environment will no doubt be seen favourably by policymakers and consumers in the country where its contractors assemble so many iPhones and iPads.

So far, 10 Apple suppliers have shown their commitment to the company’s China Clean Energy Fund by co-investing, and that includes four Taiwan heavyweights: contract supplier Catcher Technology and device manufacturers Compal Electronics, Pegatron and  Wistron.

Analysts say that some of those contractors probably felt they had to join Apple’s fund to avoid the risk of being seen as unsupportive of a major client and China’s wider clean air ambitions.

“For the manufacturers, their concern is mostly future business from Apple,” says Lions Shih, research manager with Taipei-based market research firm EnergyTrend. “They join the fund to remain suppliers and partners of Apple on the one hand and show their contribution to environment as part of corporate social responsibility on the other hand.”

Growing pressure to help China

Apple held a 13% China market share in the first quarter of 2018, behind the Chinese firms Huawei, Oppo and Vivo, according to data from the research firm Counterpoint. Apple’s sales face potential headwinds as recent Sino-U.S. trade friction has likely soured many Chinese consumers’ perceptions of the American electronics brand, according to research and consulting firm Kapronasia. This month, Washington approved import tariffs of 25% approved this month on 800 Chinese goods. Apple also came under scrutiny from 2010 to 2012, when its Taiwanese assembler Foxconn Technology saw several workers commit suicide at its sprawling China plants and later faced labor protests at others.

Apple made the clean-air commitment clear in April. CEO Tim Cook said that Apple would “keep pushing the boundaries of what is possible with the materials in our products, the way we recycle them, our facilities and our work with suppliers,” Cook said then, as cited in this company statement.

Tim Cook, chief executive officer of Apple, photographed on July 12, 2018. (Photo by Drew Angerer/Getty Images)

“Apple has promised to use renewable energy in its global operating facilities earlier this year, so it seems like the next logical step to expand the use of renewable energy to its manufacturing facilities,” Shih says.

More on Forbes: Apple To Drop Qualcomm Modems From Next iPhones For Intel, Samsung Wireless Chips

The clean energy fund announced by Apple on July 12 complements its 3-year-old Supplier Clean Energy Program. The 23 suppliers in this program should make Apple gear with “100 per cent clean energy,” Apple says in a statement. Money from the new fund will go toward developing one gigawatt worth of renewable energy in China, the statement says.

“If you do business with (Apple), then you’d better cooperate,” says Tracy Tsai, research VP with tech market analysis firm Gartner in Taipei. “You are clients, so you must go along and protect relations.”

Bottom line consideration

Pegatron joined Apple’s fund because clean energy is a “trend” that has sparked growing “awareness” among consumers of electronics, a company media liaison said for this post. Apple also happens to be “a rather important client,” the liaison said. The 11-year-old firm participates in Apple’s Supplier Clean Energy Program, as well.

The Taipei-based manufacturer won’t disclose how much business it gets from Apple or discuss the financial importance of joining Apple’s clean-energy campaign in China. Contractors seldom give details of their work with Apple. “Apple’s revenue contribution to contractors is guarded pretty heavily,” says Jason Barry, data analyst with research firm Gap Intelligence in the United States.

A man walks past the gates of Pegatron, an Apple manufacturer, in China. Pegatron stopped recruitment three years ago in China due to a fall in iPhone orders, local media reported. (VCG/VCG via Getty Images)

Taiwanese contractors need to take China seriously also because Taiwan lacks a “coherent” policy to give them chances to prosper at home, says Liang Kuo-yuan, president of Taipei-based think tank Polaris Research Institute in Taipei.

Assemblers and suppliers were expected to grow their business last year on orders for the iPhones 8 and X. Pegatron reported 2017 operating revenue of NT$1.19 trillion ($38.9 billion), up from NT$1.16 trillion in 2016. Profits came to NT$16 billion, down from NT$22 billion a year before. Fellow assembler Wistron posted net income of NT$4.4 billion last year, up from $3 billion in 2016 following its own increase in revenue.

Apple clean energy fund partners outside Taiwan include Corning, Golden Arrow, Jabil, Luxshare-ICT, Solvay and Sunway Communication.

India mulls single regulator for e-commerce sector: document

MUMBAI (Reuters) – India is mulling a single legislation to address all aspects of e-commerce regulation, and it is also exploring the idea of setting up a single regulator to consider all sector-related issues, according to a draft policy document seen by Reuters.

The federal government has indicated it aims to remove the legal fragmentation governing the e-commerce sector, according to the Draft National Policy Framework on e-commerce.

Some of the measures suggested in the draft include local data storage, mandating the use of state-run RuPay payments in online transactions and enhancing the participation of micro, small and medium enterprises in online retail.

Reporting by Sankalp Phartiyal; Editing by Euan Rocha

Google's Grand Plan To Make AI Accessible To Developers And Businesses

Artificial intelligence took center stage at Google’s annual user conference, Cloud Next 2018. The company made several announcements that make machine learning and artificial intelligence accessible to both developers and businesses.

Fei-Fei Li, Chief Scientist, Google AISource: Google

One of the first announcements came in the form of Cloud AutoML, a managed service that lets developers build machine learning models without requiring any specialized knowledge in machine learning or coding. AutoML Vision, along with other automated ML services became publicly available. According to Google, it is a suite of machine learning products that enables developers with limited machine learning expertise to train high-quality models specific to their business needs, by leveraging Google’s state-of-the-art transfer learning, and Neural Architecture Search technology.

With AutoML, developers use a simple graphical user interface (GUI) to train, evaluate, improve, and deploy models based on their own data. Apart from computer vision, AutoML also offers translation and natural language models. AutoML Natural Language helps customers to predict custom text categories specific to domains automatically. With AutoML Translation, they can upload translated language pairs to train custom translation models.

Google has also enhanced its cognitive computing APIs. Cloud Vision API now recognizes handwriting, supports additional file types (PDF and TIFF) and product search, and can identify where an object is located within an image. The improvements to Cloud Text-to-Speech include multilingual access to voices generated by DeepMind WaveNet technology and the ability to optimize for the type of speaker from which the speech is intended to play. Cloud Speech-to-Text added the ability to identify what language is spoken as well as different speakers in a conversation, word-level confidence scores, and multi-channel recognition. With this enhancement, customers can record each participant separately in multi-participant recordings.

Dialogflow, the platform to build bots, can now be used to build AI-powered virtual agents for the contact center, including phone-based conversational agents known as interactive voice response (IVR). Google Cloud Contact Center, an AI solution based on Dialogflow, includes new features alongside other tools to assist live agents and to perform analytics.

With Dialogflow Phone Gateway, customers can assign a working phone number to the virtual agent and begin taking calls. The dynamic platform can scale based on the utilization patterns. Behind the scenes, all of the telephony infrastructure, speech recognition, speech synthesis, natural language understanding and orchestration are managed automatically.

Another component of Dialogflow Enterprise, the Dialogflow Knowledge Connector understands unstructured documents like FAQs or knowledge base articles to automatically build intents with automated responses sourced from internal document collections, enriching the conversational experience with little extra effort. The added information extracted from the knowledge base is integrated with the Dialogflow agent to deliver conversational user experience.

Apart from the above enhancements, Dialogflow now includes automatic spelling correction, sentiment analysis and text-to-speech capabilities.

Google is integrating its cloud-based machine learning assets with Dialogflow to build an intelligent contact center. The platform includes an agent assist system to provide the call center agents with relevant information through suggested articles and shortcuts for fulfilling relevant tasks in real time. Another feature called the Conversational Topic Modeler uses Google AI to analyze historical audio and chat logs to uncover insights about topics and trends in customer interactions.

Google is working with several industry players to integrate Cloud Contact Center AI with mainstream contact center platforms.

From automated ML to AI-based contact center, Google wants AI to become accessible to both developers and enterprises.

YouTube plans original programing in India, Japan and other markets

BEVERLY HILLS, Calif. (Reuters) – Alphabet Inc’s YouTube is creating scripted series and other original programing for international markets including France, Germany, Japan, Mexico and India to try to draw new customers to its paid subscription service, a senior executive said on Friday.

FILE PHOTO: Susanne, Daniels Global Head, Original Content, YouTube, speaks at the Milken Institute 21st Global Conference in Beverly Hills, California, U.S., May 1, 2018. REUTERS/Mike Blake/File Photo

The programing will come in the form of multiple genres such as music documentaries, reality series, talk shows and scripted series, Susanne Daniels, YouTube’s global head of original programing, said in an interview. It will be produced in local languages and subtitled or dubbed for other markets.

Some of the programing will appear on YouTube Premium, the monthly subscription service formerly called YouTube Red. Other content will be available on YouTube’s free service with advertising.

“We are targeting markets where we believe we have a tremendous upside in potential subscribers,” Daniels said.

YouTube already has released a handful of original shows in South Korea and one in India, a talk show in Hindi about cricket. Called “UnCricket,” the show has performed “beyond expectations,” Daniels said.

Daniels also said a reality show starring South Korean pop band Big Bang had boosted subscriptions, and that 50 percent of the new customers came from outside of Korea.

More details about the new international slate will be released in the coming weeks, she said.

YouTube will be competing with companies including Netflix Inc and Inc that are investing in local language programing for online audiences around the world.

The first original shows from YouTube debuted on its premium service in 2016, starting with series from some of the platform’s most popular video creators. It added programing from Hollywood stars and also released a batch of children’s shows including Emmy-winning “Fruit Ninja Frenzy Force.”

There are no current plans for more original children’s programing, Daniels said. The YouTube Kids app serves that younger audience, she said, and the company does not believe children’s content will drive subscriptions to YouTube Premium at this time.

Reporting by Lisa Richwine; Editing by Will Dunham

Facebook’s Stock Just Took a Massive Hit, Wiping Off As Much As $145 Billion in Market Cap

Facebook’s problems have reached a boiling point. After months of questions and, often reluctant, disclosures about massive information leaks and about how it handles false information on its site seen by hundreds of millions of people, disappointing user growth caused the social network’s stock to plummet in after-hours trading on Wednesday, shedding over $145 billion in market cap.

Investors’ alarm was likely triggered by a failure in growth in its most important markets, the combined U.S. and Canada segment and Europe. U.S. and Canadian traffic was flat from the previous quarter, while Europe shed 3 million average daily users quarter over quarter, down to 279 million.

U.S. and Canadian Facebook visitors provided an average revenue per user (ARPU) in the latest quarter of $25.91, the vast majority from advertising, while the ARPU of Europeans was $8.76, according to figures provided by Facebook. Other markets offer much less value: Asia-Pacific users rack up just $2.61 in revenue, and the rest of the world lumped together, a mere $1.91.

The drop in European visitors was potentially due to the continuous revelations highlighted there about Facebook’s breaches and weaknesses, and the implementation of the European Union and related entities’ General Data Protection Regulation (GDPR) in late May. The GDPR requires more disclosure and opting in to many tracking and ad-related behaviors that aren’t related to the core function of a website.

While the company saw revenue up 42% year-over-year to $13.2 billion in its second quarter, that was short of what Wall Street expected. Net income was similarly up, to $5.1 billion from $3.9 billion the year-ago quarter, but that didn’t assuage investors and institutions. The after-hours plunge came despite Facebook also beating a consensus estimate of earnings per share of $1.72 by two cents.

This slowing growth in valuable markets may have provided the jitters that led investors to significant after-hours profit taking. The company had a nearly unbroken steady climb in its stock price since mid-2014, with a blip shedding 15% in a matter of days in March when revelations about alleged data misuse by Cambridge Analytica emerged. Facebook stock recovered gradually, and was up 29% in the last year and 21% in 2018 through the close of regular trading today, rising to a new high of 217.50, before the after-hours tumble. Nearly the last year’s gains have now been lost.

Facebook has no end in sight for scrutiny and oversight, with regulators, prosecutors, and other public and private parties in multiple countries examining the company’s actions, those of nation states allegedly manipulating news and advertising, and that of firms like Cambridge Analytica, which obtained massive amounts of information that many Facebook users likely considered private.

Yesterday, BuzzFeed published a memo by chief security officer Alex Stamos written to staff in March after the initial Cambridge Analytica stories broke in which he urged the company to pick sides on important issues. Stamos reportedly still plans to leave the company next month, following a reorganization that the New York Times said earlier this year took away 98% of the group he managed. Today, Facebook’s chief legal officer announced he’s departing at the end of this year for family reasons.

Animators Push Tech Boundaries In Hotel Transylvania 3

Drac and his merry band of classic monsters is at it again with Hotel Transylvania 3: Summer Vacation. The movie is in theaters now and finished its opening weekend on top of the box office charts. The movie is great entertainment for the whole family—but I’m not here to play movie critic. I’m focused on the technology behind the movie.

For those that aren’t familiar with the Hotel Transylvania movies, though, I’ll start with a brief synopsis. The movies have an all-star cast of voices, including Adam Sandler, Selena Gomez, Kevin James, Mel Brooks, Keegan-Michael Key and others. Drac—short for Dracula—runs a hotel for monsters and mythic creatures. In the first movie, things get crazy when a human boy discovers the hotel and falls in love with Drac’s daughter. In Hotel Transylvania 3: Summer Vacation, Drac wants a break from managing everyone else’s vacation and indulges in taking one himself—with the usual cast of classic monsters—aboard a cruise ship.

The inspiration for this movie came from Tartakovsky’s actual life. “After I finished the second film, I needed a vacation – and my in-laws surprised us with a family cruise,” he notes. “I’m sure everyone loves their in-laws as much as I do, but the cruise forced us all to be together in a small space for a week. That inspired me—what would happen if you put Drac’s Pack in that situation?”

As with previous iterations of the Hotel Transylvania franchise, director Genndy Tartakovsky laid out an ambitious vision and the team of animators working on the movie had to push the boundaries to make it happen. Most CG (computer graphic) animated films attempt to keep the character consistent using a base model—treating the character rig like a puppet. Tartakovsky encourages his animators to bring an exaggerated version of the characters to life, but without sacrificing what defines the characters.

The characters in Hotel Transylvania are pliable—displaying a rang of emotions from cartoony to subtle and realistic. Tartakovsky and some of the animators explain the challenges and thinking used to create the movie in this exclusive character shot build video:

“The Hotel Transylvania films really are a chance for the animators to live inside Genndy’s head for a little while,” says producer Michelle Murdocca. “The computer doesn’t always see things the way that Genndy sees them in his mind’s eye, but I think our animators love the challenge of breaking the mold and showing the wide range of expressions that are possible in the name of bigger emotion and bigger laughs.”

The computer technology and graphics rendering used to create the movie is crucial to bringing Tartakovsky’s vision to life. It is a cartoon and it’s a fictional world—but it still has to maintain some sense of realism for the audience to suspend disbelief and become immersed in the story. The animators strive to develop characters that are engaging and entertaining, while also paying attention to detail—like how hair blows in the wind, or how water drops or splashes, or how clothing lays on the frame of the character. The attention to realistic detail on otherwise completely unrealistic characters is part of what makes Hotel Transylvania so appealing and helps the franchise continue to be a hit at the box office.

How to Secure Your Accounts With Better Two-Factor Authentication

Hopefully by now you’ve heeded the repeated warnings from your friends and loved ones (and friendly, beloved internet writers) to use two-factor authentication to secure your digital accounts. That’s where access to Facebook or Twitter or your online bank—anything that supports it, really—requires not just a password but also a special code. Not all two-factor is created equal, however. For better protection, you’re going to want an authenticator app.

Yes, the easiest way to implement two-factor is with SMS, receiving a text with an access code every time you try to log into a secured account. While certainly better than nothing, getting your 2FA from SMS has plenty of potential downside. Specifically, it leaves you exposed if someone hijacks your smartphone’s SIM, a longtime problem that has only gotten worse of late. By stealing your phone number, hackers can redirect any two-factor notifications to their own devices, allowing them much easier entry to your accounts.

“Unfortunately, it isn’t that hard for thieves to impersonate you to your mobile phone carrier and hijack your mobile phone number—either with a phone call to customer support or walking into a phone store,” says Lorrie Cranor, a computer scientist at Carnegie Mellon University and former FTC technologist who had her own SIM stolen in 2016. Authenticator apps are not vulnerable to this problem, and thus are a more secure way to do two-factor verification.

Instagram, in particular, has seen a surge of troubling SIM attacks, largely because it only supports text-based two-factor for now. The company confirmed that it’s working on the obvious solution: Letting you use an authenticator app instead.

“Authenticator apps are not vulnerable to this problem” of SIM hijacking, says Cranor. “They’re a more secure way to do two-factor verification.”

The good news? Most of the sensitive accounts you use today already offer stronger 2FA. And there’s no shortage of third-party authenticator apps that’ll enable it for you. Here’s how to get set up, and make your sign-ins that much more stress-free.

The Basics

The most popular authenticator apps are Google Authenticator and Authy, but password managers 1Password and LastPass offer the service as well, if that helps you streamline. If you’re heavy into Microsoft’s ecosystem, you might want Microsoft Authenticator. While they all differ somewhat in features, the core functionality is the same no matter which one you use.

Rather than send you an SMS, each of these apps shows you a randomly generated six-digit code that refreshes roughly every 30 seconds, and stays constantly synced with whichever service you’re trying to log into. The benefits of tying those codes to a physical device, rather than your phone number, extend beyond security; apps like Google Authenticator generally continue to work even without an internet or cell connection. If 2FA has ever locked you out of Facebook on a flight, here’s some relief.

Most services you would want to secure offer this type of token-based 2FA; Instagram is more of the exception than the rule at this point. You can see a comprehensive list for yourself here. As for which app to use, Google Authenticator offers a barebones experience backed by a company with a sterling security record, while Authy offers more features, like being able to pull codes from not just your smartphone but your desktop or tablet. It also lets you back up your codes to the cloud, enabling a seamless migration when you inevitably upgrade your smartphone. With Google Authenticator, when you switch your main device, you have to sync your accounts over again.

For that reason, we’ll use Authy for a quick walkthrough of how to actually use a more secure 2FA app. The steps are basically the same on Google Authenticator, but it covers a little more ground.

Lock It Down

Step one: Download the app. See? This is easy. No sweat.

Once you open Authy, it’ll ask for your phone number, and then send you a registration code via either phone call, SMS, or another device. From there, it’s a blank slate until you start pairing it with the accounts you want to secure.

Here comes the drudgery. You’ll need to go to every single account you want to pair your authenticator app with; there’s no omnibus route, and no automated way to transition from SMS to Authy or Authenticator. The silver lining: While you have to repeat the set-up process many, many times across all corners of the internet, it’s quick and relatively painless.

Let’s use Dropbox as an example. Once you’re signed in on the web on your desktop, click the ID icon in the upper right corner. From there, go to Settings, then Security. Toggle on Two-step verification, then head to Edit, under Preferred Method. Click Use a mobile app, and you’ll see a QR code. Tap Add Account on Authy, point your smartphone at the screen, and congrats! Your Dropbox account is locked down tight.

Now onto the rest: Twitter, Facebook, Gmail, Evernote, and on and on. Each uses slightly different wording for its menus, but go to the settings and click on words like “privacy” and “security” until you find the available two-factor options.

If you’re using Google Authenticator, that’s basically all you need to know. And to be absolutely clear, that no-frills approach works great for most people. If you want more features, though, you can take some extra steps with Authy.

For instance! Go to Settings and tap Accounts, then toggle on Authenticator Backups if you want to create encrypted backups in the cloud. The extra cautious may prefer to keep their codes on a single device, but the cloud backup makes it possible to use Authy on more than just your smartphone—there’s even a Chrome extension—and also makes switching to a phone much more seamless.

Speaking of which, to add more devices to your Authy account, go to Settings, then Devices, and tap Allow Multi-device. From there, you can authenticate whatever else you need. Authy also lets you protect the app with a 4-digit PIN, to keep people from accessing your tokens even if they steal your device.

One more miscellaneous tip: The services that offer two-factor will also generally offer one-time use backup codes. Print these out, especially if you’re traveling, and keep them in a safe place. If for whatever reason you can’t access your app or an SMS, it’s your last, best bet to keep from getting locked out of your account.

The 2FA Spectrum

Using an authenticator app for two-factor beats SMS, but it’s still not the absolute most secure way to go. To lock even your online accounts down even further, consider stepping up to a YubiKey, which adds a hardware layer of protection. (You can get a free YubiKey 4 with a new WIRED subscription.) If you’re an activist, journalist, or other potential target of attacks, Google Advanced Protection is the most secure option around.

As with so many things, it’s a matter of balancing security and convenience. But for most people, the few minutes it takes to set up an authenticator app are more than worth the benefit over sticking with SMS—especially once Instagram and other stragglers get around to offering it.

More Great WIRED Stories

How Agile Helped Elect Donald Trump

The idea of democracy is that voters will compare the candidates and select the better on the merits of their respective campaigns. But what if the outcome reflects better use of technology, independent of the merits? In 2012, Democrats celebrated the Obama campaign’s smarter use technology that helped lead to victory. But in 2018, the shoe was on the other foot, as the Trump campaign is being hailed for nimbler use of technology, specifically the “test, learn, adapt” approach that exemplifies Agile management.

The Trump Campaign on Facebook

In the 2016 U.S. presidential election, the Trump campaign spent about 50% more in Facebook ads, but more importantly, tested almost 90 times the number of ad variations. Thus “Clinton spent $28 million from June to November 2016, while testing 66,000 different ads,” Bloomberg has reported. “In comparison, Trump spent $44 million in that period and tested 5.9 million versions of ads, suggesting his campaign’s Facebook strategy “was more complex than Clinton’s and better leveraged Facebook’s ability to optimize for outcomes.”

Trump 2016 campaign digital director Brad Parscale explained on 60 Minutes last year that the campaign “obsessively tested ad creative and messaging — up to 100,000 different versions per day of different ads. That involved “changing language, words, colors, changing things because certain people like a green button better than a blue button,” he explained, outlining the automated alterations that were made to spit out different versions of an ad. “Some people like the word ‘donate’ or ‘contribute.’”

The Trump campaign not only tested many more variants of its messages: the campaign was more tightly focused on achieving impact. Thus, “Trump’s campaign was more focused on finding new donors, according to Bloomberg, while Clinton’s campaign concentrated more on enhancing Clinton’s appeal as a candidate. “84% of Trump’s budget asked people on Facebook to take an action, like donating, compared with 56% of Clinton’s.”

In a statement to BuzzFeed News, Facebook said it “provided advice on best practices including insights on which ad formats were generating the best performance results and how to use their insights to determine best strategies.” The company also noted that most of the work of building and scaling ad campaigns was executed by “third-party vendors.”

Facebook told Buzzfeed that “it offered the same level of support to both campaigns.” But Facebook’s internal memos suggest that the Trump campaign used those tools to greater effect. In fact, Facebook embraced the methods it learned from the Trump campaign to enhance the Agile marketing model that it calls “Test, Learn, Adapt” to assess its own advertising.

For instance, Facebook has adopted lessons from the Trump campaign for its current “Here Now” effort, a multimillion-dollar advertising push to alleviate users’ concerns about privacy and the misinformation that clutters the platform. In addition to primetime television spots and “False news is not your friends” ads on Facebook itself have been fed to users calibrated with the help of the TLA ad-testing methodology.

The 2012 Obama Campaign

In the 2012, presidential campaign, it was the Obama campaign that was more adept in the use of technology. “The Obama 2012 campaign used data analytics and the experimental method to assemble a winning coalition vote by vote.

In 2012, the Obama campaign was widely celebrated in news stories for its mastery of Big Data, and its genius at data mining it to get out the vote. The press was full of stories about how the Obama campaign “won the race for voter data,” and how it “connected with young voters.” His data analytics gurus were treated as heroes.

Thus the Obama 2012 campaign used data analytics and Agile thinking to assemble a winning coalition vote by vote. In doing so, it overturned the dominance of TV advertising in U.S. politics and created something new in the world: a national campaign run like a local ward election, where the interests of individual voters were known and addressed.

To pull it off, the Obama team used Amazon’s cloud computing services for computing and storage power. At its peak, the IT infrastructure for the Obama campaign took up a significant amount of resources in AWS’s Northern Virginia data center.

“Atop Amazon’s services, the Obama team built Narwhal—a set of services that acted as an interface to a single shared data store for all of the campaign’s applications, making it possible to quickly develop new applications and to integrate existing ones into the campaign’s system. Those apps include sophisticated analytics programs like Dreamcatcher, a tool developed to ‘microtarget’ voters based on sentiments within text. And there’s Dashboard, the “virtual field office” application that helped volunteers communicate and collaborate.”

After the Obama campaign used technology to help it win in 2012, it was publicly celebrated as an act of political genius. The Trump campaign has not received the same public acclaim. The misuse of Facebook data, its relationship with Cambridge Analytica and the possible involvement with Russian meddling in the election have all tended to distract from the Trump campaign’s success in the more adept use of technology generally.

Agile Management In The Trump Campaign

Traditional management approaches problems by developing a comprehensive plan and then implementing the plan, making adjustments to the extent possible along the way. That works well enough when you are management something like constructing a building or building a bridge. But in today’s world where most activities are taking place in a world that is volatile, uncertain, complex and ambiguous—the so-called VUCA world—big plans are suspect. Instead, plans need to be implemented in increasingly-short cycles, with direct feedback from customers and end-users and evidence on impact, i.e. whether their behavior is changing. Rapid learning and impact are facets of rapidly changing plans.

Agile management, which took off in software development over the last 15 years, has now spread—in forward looking organizations—to the management of everything, including political campaigns. Organizations implementing traditional management find they can’t keep up.

The “Test Learn, Adapt” approach is an example of the Agile mindset that underlies all Agile management. While the approach had been used by Facebook with other customers before the Trump campaign, Facebook has embraced the Trump campaign as a demonstration of the method’s efficacy. It goes beyond “simple A/B testing” — a common method of running experiments by changing one variable —and develops many combinations using different “messaging,” “creative,” “format,” and “delivery.” The variations are then then tested on a tiny segment of the audience, with the one generating the most engagement being shared with the full target audience.

According to Buzzfeed, the approach was employed on Facebook’s advertising tests for News Feed within the US as part of the current “Here Now” campaign, with Facebook users now opening up the social network’s app or website to see ads proclaiming “Clickbait is not your friend” or “Data misuse is not your friend.”

The Need For Greater Transparency In The Use Of Technology

The idea of democracy, that voters will select the better candidate on the merits of their respective campaigns, has long been recognized to be under siege from the impact of money in elections. Steps, albeit only partially effective, have been taken to make the role of money more transparent.

Now given the increasing evidence that the use of technology itself may make a significant difference between campaigns, quite independent of the candidates’ merits, it is becoming increasingly urgent to have greater transparency in the use of technology in elections.

The message is also clear for candidates in upcoming elections. Success will depend not just on strength of their messaging and their fund-raising: it will also depend on how agile is the use of technology in their campaigns. What was agile enough in 2012 was not good enough in 2016. What was good enough in 2016 won’t be good enough in 2018 or 2020. Agile is in effect eating the world.

And read also

The Five Whys Of The Trump Surprise

Trump And Authoritarian Propaganda

The Political Fabric Unravels

Space Photos of the Week: Sweeping the Clouds Away on Titan

You know that hazy, cloudy, yellowish moon of Saturn called Titan? This is the same moon! Using years’ worth of infrared data from Cassini’s Visual and Infrared Mapping Spectrometer, scientists have created the most detailed view of Titan we’ve ever seen. Because of the thick atmosphere enveloping this moon, it’s ordinarily a challenge to see through the clouds and haze. With infrared? A whole new world, or moon, opens up.

Sometimes the Earth, seen from space, can have features more like those you’d expect on alien planets. This particular photo is of the Chukchi Sea, a stretch of Arctic water north of the Bering Strait between Alaska and Russia. NASA’s Landsat8 satellite captured these lovely blue and green swirls—blooms of phytoplankton—just a few weeks ago. Looks almost warm enough for a summer swim!

The Pillars of Creation, part of the Eagle nebula, are perhaps the most renowned nebula feature of our time, and for good reason: They’re beautiful. These towers got their name because they were a stellar nursery, a place where new stars are born. Yes, the key word is “were.” Scientists think a supernova explosion felled the pillars some years ago. However, the light from that cataclysm hasn’t reached us yet; what we’re seeing in this photo is about 5,700 years old, since that’s how long it took the nebula’s light to reach the Hubble Space Telescope.

The dwarf planet Ceres, the closest such planet to Earth, resides in the asteroid belt between our world and Mars. NASA’s Dawn spacecraft has been orbiting Ceres for over three years and has spotted some pretty interesting things close up—including these bright, salty regions in the impact crater known as Occator Crater. Dawn’s mission is due to end sometime in the next couple of months as it runs out of fuel, though it will continue forever orbiting Ceres. We just won’t be able to communicate with it anymore.

Saturn, you’re so awesome. The Cassini spacecraft took this stunning pic of the ringed planet in January 2010, with the light from the night side enhanced so we can better see the demarcations in the rings. Although Cassini ended its mission when it entered Saturn’s atmosphere last September, we’ll always have these grand views of the most remarkable planet (other than ours) in our solar system.

This sparkling cluster of stars is actually an irregular galaxy called UGC 12682. The bright blue regions indicate star formation, and their contrast against the dark fabric of space makes this photo all the more alluring. Galaxy UGC 12682 is 70 million light years from Earth, so all the baby stars here probably have grown up by now.

?How to add Linux to your Chromebook

It’s long been possible to run Linux on a Chromebook. That’s no surprise. After all, Chrome OS is a Linux variant. But, doing it by using either Crouton in a chroot container or Gallium OS, a Xubuntu Chromebook-specific Linux variant, wasn’t easy. Then, Google announced it was bringing a completely integrated Linux desktop to the Chromebook.

Today, with a properly-equipped Chromebook and the bravery to run canary code, you can run Debian Linux on your Chromebook. Here’s how to do it.

This new Chromebook Linux feature is Crostini, the umbrella technology for getting Linux running with Chrome OS. Crostini gets enough Linux running to run KVM, Linux’s built-in virtual machine (VM). On top of this, Crostini starts and runs LXC containers. You won’t see it, unless you look closely, but it’s in those containers that your Debian Linux instances are running.

Eventually, anyone with a newer Chromebook will be able to run Linux. Specifically, if your Chromebook’s operating system is based on the Linux 4.4 kernel, you’ll be supported. But we’re not there yet. It’s also possible that older Chromebooks, running Linux 4.14, will be retrofitted with Crostini support.

Officially, you need a Pixelbook, Google’s top-of-the-line Chromebook, to run Linux. But, users have found a dozen other models can run Crostini with half-a-dozen others expected to be supported soon. Chromebooks that can already use Crostini include newer Intel-powered Chromebooks from Acer, Asus, HP, Lenovo, and Samsung. Dell models will start getting supported later this year.

I used my best-of-breed Pixelbook with its 1.3GHz quad-core Intel Core i7-7Y75 processor, 512GB SSD, with 16GBs of RAM for my tests. This is the fastest Chromebook on the market. It’s not cheap, at a list price of $1,399, but it’s worth it if you want to push Linux on the Chromebook’s limits.

Once you have the hardware you need, you must switch your Chromebook from the stable update channel to the dev channel. This is alpha software and it updates about once a week. Let me make this absolutely clear: This is not stable software. It will blow up at times. But, faint heart never won fair technology discoveries.

This is a bigger decision than it looks at first. You’ll lose all your local data if you try to go back to the stable, or even beta, channels. With a Chromebook that’s not much of a problem since most of your data and settings are kept on the Google Cloud, it’s still worth keeping in mind.

If you want to wait and be safe, Crostini support is expected to enter the stable channel with Chrome OS 69 in mid-September.

To make the switch to dev, take the following steps:

  1. Sign in to your Chromebook with the owner account.
  2. Click your account photo.
  3. Click Settings.
  4. At the top left, click Menu.
  5. Scroll down and click About Chrome OS.
  6. Click Detailed build information.
  7. Next to “Channel” click Change channel.
  8. Pick a channel.
  9. Click Change Channel.
  10. Your Chromebook will download the dev channel update. It will then ask you to restart your Chromebook.

Once that’s done, if you’re not using a Pixelbook you may need to set a Chrome flag to access Linux. You do this by entering: chrome://flags on the Chrome browser’s address line. This command displays all of Chrome’s experimental features. Scroll down the list until you find:


Activate this, and your system may be ready to go. I say “may” because to run Crostini your Chromebook must not only be on the dev channel, but Google must also have enabled the Linux VM for your hardware.

The easiest way to confirm that a particular Chromebook works with Linux is to follow the above steps and then open Chrome OS’s built-in shell, crosh, and run the shell command:

vmc start termina

If you get a message such as “ERROR: command ‘vmc’ is not available”, you’re out of luck. But, if you see a terminal, congrats, you’ve just found a new Chromebook that’s Linux-ready.

Next, head to Chrome OS settings (chrome://settings), scroll down to to “Linux (Beta)” and activate it.

Now, open the app switcher by pressing the Search/Launcher key and type “Terminal”. This launches the Termina VM, which will start running a Debian 9.0 Stretch Linux container.

Congratulations! You’re now running Debian Linux on your Chromebook.

From here you can install and run programs using Debian’s normal software commands. For example, to update my new Linux system and install the lynx web browser, I’d run:

$ sudo apt-get update

$ sudo apt-get upgrade

$ sudo apt install lynx

Crostini Linux running Lynx

With Crostini, you can now run Linux and Linux applications, such as Lynx, on a Chromebook.


While you could install pretty much any program on your new Linux instance, I gave a shell-based program example because accelerated graphics and audio aren’t working yet. So, while you could install Cinnamon, my favorite Linux interface, or Steam for games, it’s not fast enough even on a maxed out Pixelbook to be that enjoyable. Not yet anyway.

In addition, many graphics-based programs, such as the photo-editor Gimp, won’t run yet on Crostini. Give it time to mature before trying to get too fancy with heavy graphics software.

Soon, though, Linux and Chrome will be a matched pair. Come that day, I see high-end Chromebooks becoming the laptops of choice for developers.

In the meantime, if you want to do more with Linux and Chromebook, check out the excellent Reddit Crostini Wiki. For up-to-the-minute hands-on information about Crostini, its parent Reddit forum is the best resource on the web.


Related Stories: