Connect with us

Technology

Censoring China’s Internet, for Stability and Profit

Published

on

[ad_1]

Li Chengzhi had a lot to learn when he first got a job as a professional censor.

Like many young people in China, the 24-year-old recent college graduate knew little about the 1989 Tiananmen Square crackdown. He had never heard of China’s most famous dissident, Liu Xiaobo, the Nobel Peace Prize laureate who died in custody two years ago.

Now, after training, he knows what to look for — and what to block. He spends his hours scanning online content on behalf of Chinese media companies looking for anything that will provoke the government’s wrath. He knows how to spot code words that obliquely refer to Chinese leaders and scandals, or the memes that touch on subjects the Chinese government doesn’t want people to read about.

Mr. Li, who still has traces of youthful acne on his face, takes his job seriously. “It helps cleanse the online environment,” he said.

For Chinese companies, staying on the safe side of government censors is a matter of life and death. Adding to the burden, the authorities demand that companies censor themselves, spurring them to hire thousands of people to police content.

That in turn has created a growing and lucrative new industry: censorship factories.

Mr. Li works for Beyondsoft, a Beijing-based tech services company that, among other businesses, takes on the censorship burden for other companies. He works in its office in the city of Chengdu. Located in the heart of a high-tech industrial area, the space is bright and new enough that it resembles the offices of well-funded start-ups in tech centers like Beijing and Shenzhen. It recently moved to the space because customers complained that its previous office was too cramped to allow employees to do their best work.

“Missing one beat could cause a serious political mistake,” said Yang Xiao, head of Beyondsoft’s internet service business, including content reviewing. (Beyondsoft declined to disclose which Chinese media or online companies it works for, citing confidentiality.)

China has built the world’s most extensive and sophisticated online censorship system. It grew even stronger under President Xi Jinping, who wants the internet to play a greater role in strengthening the Communist Party’s hold on society. More content is considered sensitive. Punishments are getting more severe.

Once circumspect about its controls, China now preaches a vision of a government-supervised internet that has surprising resonance in other countries. Even traditional bastions of free expression like Western Europe and the United States are considering their own digital limits. Platforms like Facebook and YouTube have said that they would hire thousands more people to better keep a handle on their content.

Workers like Mr. Li show the extremes of that approach — one that controls what more than 800 million internet users in China see every day.

Beyondsoft employs over 4,000 workers like Mr. Li at its content reviewing factories. That is up from about 200 in 2016. They review and censor content day and night.

“We’re the Foxconn in the data industry,” said Mr. Yang, comparing his firm to the biggest contract manufacturer that makes iPhones and other products for Apple.

Many online media companies have their own internal content review teams, sometimes numbering in the thousands. They are exploring ways to get artificial intelligence to do the work. The head of the A.I. lab at a major online media company, who asked for anonymity because the subject is sensitive, said the company has 120 machine learning models.

But success is spotty. Users can easily fool algorithms.

“The A.I. machines are intelligent but they aren’t as clever as human brains,” Mr. Li said. “They miss a lot of things when reviewing content.”

Beyondsoft has a team of 160 people in Chengdu working four shifts a day to review potentially politically sensitive content on a news aggregating app.

For the same app, Beyondsoft has another team in the western city of Xi’an reviewing potentially vulgar or profane content. Like the rest of the world, China’s internet is rife with pornography and other material that many users might find offensive.

In the Chengdu office, workers must put their smartphones in hallway lockers. They can’t take screenshots or send any information from their computers.

The workers are almost all young college graduates in their 20s. They are often unaware of, or indifferent to, politics. In China, many parents and teachers tell the young that caring about politics leads only to trouble.

To overcome that, Mr. Yang and his colleagues developed a sophisticated training system. New hires start with weeklong “theory” training, during which senior employees teach them the sensitive information that they didn’t know before.

“My office is next to the big training room,” Mr. Yang said. “I often hear the surprised sounds of ‘Ah, ah, ah.’”

“They didn’t know things like June 4,” he added, referring to the 1989 Tiananmen Square crackdown. “They really didn’t know.”

Beyondsoft has developed an extensive database based on such information that Mr. Yang calls one of its “core competencies.” They also use anti-censorship software to regularly visit what they call anti-revolutionary websites that are blocked by the Chinese government. They then update the database.

New employees study the database much like preparing for college entrance exams. After two weeks, they have to pass a test.

The screen saver on each computer is the same: photos and names of current and past members of the Politburo Standing Committee, the Communist Party’s top leadership. Workers must memorize those faces: Only government-owned websites and specially approved political blogs — a group on what’s called a whitelist — are allowed to post photos of top leaders.

Workers are briefed at the beginning of their shift on the newest censoring instructions sent by clients, which the clients themselves receive from government censors. Workers then must answer about 10 questions designed to test their memory. The results of the exam affect the workers’ pay.

One question on a recent Friday: Which one of the following names is the daughter of Li Peng, China’s former premier? The correct answer is Li Xiaolin, a longtime target of online ridicule for her expensive fashion taste and for being one of many children of senior officials who come into high positions or wealth.

That’s a relatively easy one. A tougher test is parsing out the roundabout ways that China’s internet users evade stringent censorship to talk about current affairs.

Take, for example, a Hong Kong news site’s 2017 commentary that compared the six Chinese leaders since Mao Zedong to emperors during the Han dynasty. Some Chinese users started using the emperors’ names when referring to the leaders. Beyondsoft’s workers have to know which emperor’s name is associated with which leader.

Then there are the photos of an empty chair. They refer to Mr. Liu, the Nobel laureate, who wasn’t allowed to leave China to attend the award ceremony and was represented by an empty chair. References to George Orwell’s novel “1984” are also forbidden.

Beyondsoft’s software trawls through web pages and marks potentially offensive words in different colors. If a page is full of color-coded words, it usually requires a closer look, according to the executives. If there are only one or two, it’s pretty safe to let it pass.

According to Beyondsoft’s website, its content monitoring service, called Rainbow Shield, has compiled over 100,000 basic sensitive words and over three million derivative words. Politically sensitive words make up one-third of the total, followed by words related to pornography, prostitution, gambling and knives.

Workers like Mr. Li make $350 to $500 a month, about average pay in Chengdu. Each worker is expected to review 1,000 to 2,000 articles during a shift. Articles uploaded to the news app must be approved or rejected within an hour. Unlike Foxconn workers, they don’t work much overtime because longer hours could hurt accuracy, said Mr. Yang, the executive.

It’s easy to make mistakes. One article about Peng Liyuan, China’s first lady, mistakenly used the photo of a famous singer rumored to be linked to another leader. It was caught by someone else before it went out, Mr. Yang said.

Mr. Li, the young censor, said the worst mistakes are almost all related to senior leaders. He once missed a tiny photo of Mr. Xi on a website not on the whitelist because he was tired. He still kicks himself for it.

When asked whether he had shared with family and friends what he learned at work, such as the Tiananmen crackdown, Mr. Li vehemently said no.

“This information is not for people outside to know,” he said. “Once many people know about it, it could generate rumors.”

But the crackdown was history. It wasn’t a rumor. How would he reconcile that?

“For certain things,” he said, “one just has to obey the rules.”

[ad_2]

Source link

قالب وردپرس

Technology

More groups join in support of women in STEM program at Carleton

Published

on

By

OTTAWA — Major companies and government partners are lending their support to Carleton University’s newly established Women in Engineering and Information Technology Program.

The list of supporters includes Mississauga-based construction company EllisDon.

The latest to announce their support for the program also include BlackBerry QNX, CIRA (Canadian Internet Registration Authority), Ericsson, Nokia, Solace, Trend Micro, the Canadian Nuclear Safety Commission, CGI, Gastops, Leonardo DRS, Lockheed Martin Canada, Amdocs and Ross.

The program is officially set to launch this September.

It is being led by Carleton’s Faculty of Engineering and Design with the goal of establishing meaningful partnerships in support of women in STEM.  

The program will host events for women students to build relationships with industry and government partners, create mentorship opportunities, as well as establish a special fund to support allies at Carleton in meeting equity, diversity and inclusion goals.

Continue Reading

Technology

VR tech to revolutionize commercial driver training

Published

on

By

Serious Labs seems to have found a way from tragedy to triumph? The Edmonton-based firm designs and manufactures virtual reality simulators to standardize training programs for operators of heavy equipment such as aerial lifts, cranes, forklifts, and commercial trucks. These simulators enable operators to acquire and practice operational skills for the job safety and efficiency in a risk-free virtual environment so they can work more safely and efficiently.

The 2018 Humboldt bus catastrophe sent shock waves across the industry. The tragedy highlighted the need for standardized commercial driver training and testing. It also contributed to the acceleration of the federal government implementing a Mandatory Entry-Level Training (MELT) program for Class 1 & 2 drivers currently being adopted across Canada. MELT is a much more rigorous standard that promotes safety and in-depth practice for new drivers.

Enter Serious Labs. By proposing to harness the power of virtual reality (VR), Serious Labs has earned considerable funding to develop a VR commercial truck driving simulator.

The Government of Alberta has awarded $1 million, and Emissions Reduction Alberta (ERA) is contributing an additional $2 million for the simulator development. Commercial deployment is estimated to begin in 2024, with the simulator to be made available across Canada and the United States, and with the Alberta Motor Transport Association (AMTA) helping to provide simulator tests to certify that driver trainees have attained the appropriate standard. West Tech Report recently took the opportunity to chat with Serious Labs CEO, Jim Colvin, about the environmental and labour benefits of VR Driver Training, as well as the unique way that Colvin went from angel investor to CEO of the company.

Continue Reading

Technology

Next-Gen Tech Company Pops on New Cover Detection Test

Published

on

By

While the world comes out of the initial stages of the pandemic, COVID-19 will be continue to be a threat for some time to come. Companies, such as Zen Graphene, are working on ways to detect the virus and its variants and are on the forefronts of technology.

Nanotechnology firm ZEN Graphene Solutions Ltd. (TSX-Venture:ZEN) (OTCPK:ZENYF), is working to develop technology to help detect the COVID-19 virus and its variants. The firm signed an exclusive agreement with McMaster University to be the global commercializing partner for a newly developed aptamer-based, SARS-CoV-2 rapid detection technology.

This patent-pending technology uses clinical samples from patients and was funded by the Canadian Institutes of Health Research. The test is considered extremely accurate, scalable, saliva-based, affordable, and provides results in under 10 minutes.

Shares were trading up over 5% to $3.07 in early afternoon trade.

Continue Reading

Chat

Trending