China’s Burned Out Tech Workers are Fighting Back Against Long Hours

1The draining 996 work schedule—named for the expectation that employees work 9 a.m. to 9 p.m., six days a week—has persisted in Chinese companies for years despite ongoing public outcry. Even Alibaba co-founder Jack Ma once called it a “huge blessing.”

In early October this year, it seemed the tide might have been turning. After hopeful signs of increased government scrutiny in August, four aspiring tech workers initiated a social media project designed to expose the problem with the nation’s working culture. A publicly editable database of company practices, it soon went viral, revealing working conditions at many companies in the tech sector and helping bring 996 to the center of the public’s attention. It managed to garner 1 million views within its first week.

But the project—first dubbed Worker Lives Matter and then Working Time—was gone almost as quickly as it appeared. The database and the GitHub repository page have been deleted, and online discussions about the work have been censored by Chinese social networking platforms.  The short life of Working Time highlights how difficult it is to make progress against overtime practices that, while technically illegal in China, are still thriving.

But some suspect it won’t be the last anonymous project to take on 996. “I believe there will be more and more attempts and initiatives like this,” says programmer Suji Yan, who has worked on another anti-996 project. With better approaches to avoiding censorship, he says, they could bring even more attention to the problem.

Tracking hours

Working Time started with a spreadsheet shared on Tencent Docs, China’s version of Google Docs. Shortly after it was posted, it was populated with entries attributed to companies such as Alibaba, the Chinese-language internet search provider Baidu, and e-commerce company JD.com.  “9 a.m., 10:30 p.m.–11:00 p.m., six days a week, managers usually go home after midnight,” read one entry linked with tech giant Huawei. “10 a.m., 9 p.m. (off-work time 9 p.m., but our group stays until 9:30 p.m. or 10 p.m. because of involution,” noted another entry (“involution” is Chinese internet slang for irrational competition).

Within three days, more than 1,000 entries had been added. A few days later, it became the top trending topic on China’s Quora-like online forum Zhihu.  As the spreadsheet grew and got more public attention, one organizer, with the user name 秃头才能变强 (“Only Being Bald Can Make You Strong”), came out on Zhihu to share the story behind the burgeoning project. “Four of us are fresh college and master’s degree graduates who were born between 1996 and 2001,” the organizer said.genesis3-1-1

Initially, the spreadsheet was just for information sharing, to help job hunters like themselves, they said. But as it got popular, the organizers decided to push from information gathering to activism. “It is not simply about sharing anymore, as we bear some social responsibility,”

The spreadsheet filled a gap in China, where there is a lack of company rating sites such as Glassdoor and limited ways for people to learn about benefits, office culture, and salary information. Some job seekers depend on word of mouth, while others reach out to workers randomly on the professional networking app Maimai or piece together information from job listings.  “I have heard about 996, but I was not aware it is that common.

Now I see the tables made by others, I feel quite shocked,” Lane Sun, a university student from Nanjing, said when the project was still public. Against 996 According to China’s labor laws, a typical work schedule is eight hours a day, with a maximum of 44 hours a week. Extra hours beyond that require overtime pay, and monthly overtime totals are capped at 36 hours.125x125-1-1-1

But for a long time, China’s tech companies and startups have skirted overtime caps and become notorious for endorsing, glamorizing, and in some cases mandating long hours in the name of hard work and competitive advantage.  In a joint survey by China’s online job site Boss Zhipin and the microblogging platform Weibo in 2019, only 10.6% of workers surveyed said they rarely worked overtime, while 24.7% worked overtime every day.

 Long work hours can benefit workers, Jack Ma explained in 2019. “Since you are here, instead of making yourself miserable, you should do 996,” Ma said in a speech at an internal Alibaba meeting that was later shared online. “Your 10-year working experience will be the same as others’ 20 years.” But the tech community had already started to fight back. Earlier that year, a user created the domain 996.icu.

A repository of the same name was launched on GitHub a few days later. The name means that “by following the 996 work schedule, you are risking yourself getting into the ICU (intensive care unit),” explains the GitHub page, which includes regulations on working hours under China’s labor law and a list of more than 200 companies that practice 996.  Within three days, the repository got over 100,000 stars, or bookmarks, becoming the top trending project on GitHub at that time. It was blocked not long after by Chinese browsers including QQ and 360, ultimately disappearing entirely from the Chinese internet (it is still available through VPNs).

The 996.icu project was quickly followed by the Anti-996 License. Devised by Yan and Katt Gu, who has a legal background, the software license allows developers to restrict the use of their code to those entities that comply with labor laws. In total, the Anti-996 License has been adopted by more than 2,000 projects, Yan says. Today, 996 is facing increasing public scrutiny from both Chinese authorities and the general public.

After a former employee at the agriculture-focused tech firm Pinduoduo died in December 2020, allegedly because of overwork, China’s state-run press agency Xinhua called out overtime culture and advocated for shorter hours.This company delivers packages faster than Amazon, but workers pay the priceSouth Korean e-commerce giant Coupang uses AI to promise almost-instant delivery. But speed comes with troubling labor issues—including worker deaths.

And on August 26, China’s Ministry of Human Resources and Social Security and the Supreme People’s Court jointly published guidelines and examples of court cases on overtime, sending reminders to companies and individuals to be aware of labor laws. But even though authorities and state media seem to be taking a tougher stand, it is unclear when or if the rules that make 996 illegal will be fully enforced. Some companies are making changes.quintex-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-1-2-1-1-1-1-2-2-1-1-1

Anthony Cai, a current employee of Baidu, says working six days a week is quite rare in big companies nowadays. This year, several tech companies including and ByteDance, the developer of TikTok, canceled “big/small weeks,” an emerging term in China that refers to working a six-day schedule every other week. “Working on Saturday is not that popular anymore,” Cai says. “However, staying late at the office is still very common, which is not usually counted as overtime hours.” 

 Source: https://www.technologyreview.com

.

More Contents:

“Guide to Employment law in Spain”

Clearview AI Ordered To Delete All Facial Recognition Data Belonging To Australians

Controversial facial recognition firm Clearview AI has been ordered to destroy all images and facial templates belonging to individuals living in Australia by the country’s national privacy regulator.

Clearview, which claims to have scraped 10 billion images of people from social media sites in order to identify them in other photos, sells its technology to law enforcement agencies. It was trialled by the Australian Federal Police (AFP) between October 2019 and March 2020.

Now, following an investigation, Australia privacy regulator, the Office of the Australian Information Commissioner (OAIC), has found that the company breached citizens’ privacy. “The covert collection of this kind of sensitive information is unreasonably intrusive and unfair,” said OAIC privacy commissioner Angelene Falk in a press statement. “It carries significant risk of harm to individuals, including vulnerable groups such as children and victims of crime, whose images can be searched on Clearview AI’s database.”

Said Falk: “When Australians use social media or professional networking sites, they don’t expect their facial images to be collected without their consent by a commercial entity to create biometric templates for completely unrelated identification purposes. The indiscriminate scraping of people’s facial images, only a fraction of whom would ever be connected with law enforcement investigations, may adversely impact the personal freedoms of all Australians who perceive themselves to be under surveillance.”

The investigation into Clearview’s practices by the OAIC was carried out in conjunction with the UK’s Information Commissioner’s Office (ICO). However, the ICO has yet to make a decision about the legality of Clearview’s work in the UK. The agency says it is “considering its next steps and any formal regulatory action that may be appropriate under the UK data protection laws.”

As reported by The Guardian, Clearview itself intends to appeal the decision. “Clearview AI operates legitimately according to the laws of its places of business,” Mark Love, a lawyer for the firm BAL Lawyers representing Clearview, told the publication. “Not only has the commissioner’s decision missed the mark on the manner of Clearview AI’s manner of operation, the commissioner lacks jurisdiction.”

Clearview argues that the images it collected were publicly available, so no breach of privacy occurred, and that they were published in the US, so Australian law does not apply.

Around the world, though, there is growing discontent with the spread of facial recognition systems, which threaten to eliminate anonymity in public spaces. Yesterday, Facebook parent company Meta announced it was shutting down the social platform’s facial recognition feature and deleting the facial templates it created for the system. The company cited “growing concerns about the use of this technology as a whole.” Meta also recently paid a $650 million settlement after the tech was found to have breached privacy laws in Illinois in the US.

Source: Clearview AI ordered to delete all facial recognition data belonging to Australians – The Verge

.

More Contents:

The 5 Biggest Blockchain Trends In 2022

Blockchain is one of the most exciting tech trends at the moment. It is a distributed, encrypted database model that has the potential to solve many problems around online trust and security. Many people know it as the technology that underpins Bitcoin and cryptocurrencies in general. However, its potential uses are far broader, encompassing digital “smart” contracts, logistics and supply chain provenance and security, and protection against identity theft.

There are countless others – blockchain evangelists say it can potentially be used to improve security and integrity in any system that involves multiple parties sharing access to a database. During 2022, spending on blockchain solutions by businesses is forecast to hit $11.7 billion. Here are some of the trends that will be driving this and some thoughts on how this will impact more and more lives over the course of the next year.

Green blockchain initiatives

Blockchains can potentially use a lot of energy and create high levels of carbon emissions – this fact was behind Tesla CEO Elon Musk’s decision to temporarily stop accepting Bitcoin in payment for his cars earlier in 2021. For this very good reason, during 2022, we are likely to see a great deal of emphasis on attempts to “greenify” blockchain. There are a few ways this can be done, including carbon offsetting, although many people consider that this often equates to simply patching up a wound that shouldn’t have been caused in the first place.

Another is by moving to less energy-intensive models of blockchain technology – typically those that rely on “proof-of-stake” algorithms rather than “proof-of-work” to generate consensus. Ethereum – the second best-known blockchain after Bitcoin – plans to move to a POS model during 2022. Another route to a greener operating model is the one championed by Cathy Wood, CEO of tech-focused hedge fund Ark Invest. This posits the view that growing demand for energy will lead to greater investments into generating renewable energy, which will then be used for other applications as well as operating blockchains.

NFT expanding beyond online art

Non-Fungible Tokens (NFTs) were the big news in the blockchain scene during 2021. Astronomical prices achieved by artwork such as Beeple’s The First 5000 Days created plenty of headlines, placing the concept of unique digital tokens residing on blockchains firmly in the public consciousness. It’s also firmly taken hold in the music world, with artists including Kings of Leon, Shawn Mendes, and Grimes all releasing tracks in NFT format. But like blockchain in general, the idea has potential beyond it’s first publicity-grabbing use cases.

Distillers William Grant and Son recently sold bottles of 46-year-old Glenfiddich whisky alongside NFTs, which are used to prove each bottle’s provenance. NFTs in gaming are starting to take off in a big way – monster-breeding game Axie Infinity allows players to “mint” their own NFT creatures to send into battle and currently has around 300,000 concurrent players (Fortnite, for comparison, has around 3.5 million). Dolce & Gabbana and Nike have both created clothing and footwear that come with their own NFTs. And the metaverse concept – championed this year by Facebook, Microsoft, and Nvidia – brings plenty of opportunities for innovative NFT use cases.

More countries adopt Bitcoin and national cryptocurrencies

2021 saw El Salvador become among the first nations to adopt Bitcoin as legal tender, meaning it can be accepted across the country to pay for goods and services, and businesses can use it to pay their employees. According to many commentators, during 2022, we will see a number of other countries follow suit.

Alexander Hoptner, CEO of cryptocurrency exchange BitMEX, predicts that at least five developing countries will start to accept Bitcoin next year, driven by global inflation and growing remittance fees from financial “middlemen” organizations used to send money home by overseas workers.

National cryptocurrencies – where central banks create their own coins that they can control, rather than adopting existing decentralized coins – are another area where we will see growth in 2022. These projects typically involve digital currencies that will operate alongside existing traditional currencies, allowing users to conduct their own transactions and manage their custody without relying on third-party service providers, while also allowing the central banks to keep control of the circulating supply – keeping the value of the token pegged to the value of the country’s traditional currency.

While the UK government-endorsed Britcoin is unlikely to be ready for launch during 2022, others, including China, Singapore, Tunisia, and Ecuador, have already done so, with more, including Japan, Russia, Sweden, and Estonia likely to join soon.

Blockchain and IoT integration

Blockchain is hugely compatible with the idea of the Internet of Things (IoT) because it is great for creating records of interactions and transactions between machines. It can potentially help to solve many problems around security as well as scalability due to the automated, encrypted, and immutable nature of blockchain ledgers and databases. It could even be used for machine-to-machine transactions – enabling micropayments to be made via cryptocurrencies when one machine or network needs to procure services from another.

While this is an advanced use case that may involve us traveling a little further down the road before it impacts our day-to-day lives, it’s likely we will start to hear about more pilot projects and initial use cases in this field during 2022. Innovation in this field is likely to be driven by the ongoing rollout of 5G networks, meaning greater connectivity between all manner of smart, networked equipment and appliances – not simply in terms of speed, but also new types of data transactions including blockchain transactions.

Blockchain in vaccine manufacture and tracking

It’s now clear that tackling the Covid-19 global pandemic will continue to be a priority throughout 2022 and a key use case for many of this year’s top tech trends. Blockchain technology has several important potential use cases in vaccine tracking and distribution.

In a world where counterfeiters are known to be creating and selling fake vaccines, blockchain means the authenticity of vaccine shipments can be proven, and their distribution can be traced to ensure they are arriving at their intended locations. There’s also a need to ensure integrity at every point of the supply chain – for example, to ensure batches of vaccines are consistently stored at the correct temperature, as is needed by many of them. IBM has created a system to allow coordination between the many different and varied agencies and healthcare authorities involved with vaccine distribution, using blockchain to unify recording of vaccination rates and efficacy across the various tools and platforms they all have in use. A pilot project also showed how blockchain could potentially speed up the ability to recognize where a product recall might be needed – for example, in a case where a batch seems to be causing an unusually high occurrence of side-effects – from three days to just a few seconds. Breakthroughs that come about due to the unprecedented response to this pandemic are likely to go on to enable more use cases for blockchain technology in the manufacture, distribution, and management of vaccinations in 2022.

Follow me on Twitter or LinkedIn. Check out my website or some of my other work here.

Bernard Marr is an internationally best-selling author, popular keynote speaker, futurist, and a strategic business & technology advisor to governments and companies.He helps organizations improve their business performance, use data more intelligently, and understand the implications of new technologies such as artificial intelligence, big data, blockchains, and the Internet of Things. Why don’t you connect with Bernard on Twitter (@bernardmarr), LinkedIn (https://uk.linkedin.com/in/bernardmarr) or instagram (bernard.marr)?

Source: The 5 Biggest Blockchain Trends In 2022

.

More Contents:

How Are Developing Countries Spending Climate Financing Money?

Over the past decade, rich nations and private enterprises have raised at least $500 billion to help developing countries cope with climate change. This financing plan, hatched in 2009, was supposed to build to an annual mobilization of $100 billion by 2020, and was designed to offset the unfairness of climate change—of poor countries suffering because rich countries had already emitted their way to wealth.

Developed countries, as we’re finding out now, have missed that $100 billion target. Just as importantly, though, it turns out that no one—no individual, no government, no multilateral agency—knows precisely how all this climate funding is being spent, or even if it’s being spent at all. Even the best such database, maintained by OECD, is a broad-brush one and has many gaps—not least in the details of private financing. The very term “climate financing,” in fact, has often proved to be slippery and malleable, defined by parties according to their own need or convenience.

At the COP26 climate summit in Glasgow, scheduled to end on Nov. 12, ever-bigger climate financing numbers are being proposed: $130 trillion from a private sector consortium, an annual $1 trillion demanded by India, an annual $1.3 trillion demanded by African nations. But without a way to track how this money is used, these larger numbers “feel like greenwashing,” said Liane Schalatek, associate director of the Heinrich-Böll-Stiftung, a policy think tank headquartered in Berlin. “It isn’t just, ‘Tell me the number,’ it should also be, ‘Show me how the number is computed,’ otherwise it’s a fig leaf.”

Insofar as some funds are more important than others, these climate financing funds are among the most important quantities of money in the world; human civilization itself hinges on whether these funds are raised in full and spent well. The absence of a detailed, publicly available account of this financing, said Schalatek, risks all sorts of omissions: donors mislabelling their funding, or money being misspent, or an under-estimation of the true volume of money required. Which, in turn, risks leaving the world far less prepared for climate change than it needs to be.

How much money are countries raising to fight climate change?

In 2019, the last year for which the OECD has published full data, rich countries raised nearly $80 billion as part of their climate financing pledge. Most of this was either in the form of bilateral funds—loans or grants from one government to another—or multilateral public funds, whereby developed countries channeled their money through international banks or climate funds. The money is intended to help developing countries both mitigate the effects of climate change—including investing in cleaner energy sources—as well as adapt to a harsher climate.

                                           

The data available on these funds, Schalatek said, resembles an onion. At the core are projects for which the most details are available: those financed through multilateral agencies. Schalatek’s team tracks a portion of these, “covering the most important multilateral climate funds,” she said, “although we don’t know what fraction it is of the overall multilateral funding.” In the Heinrich-Böll-Stiftung database, for instance, grants can be found:

  • from the Global Environment Facility, in 2014, of $2.78 million to Argentina to introduce biogas technologies into the national waste management program
  • from the Green Climate Fund, in 2019, of $9.68 million to Bangladesh to build plinths that raise the land of high-risk villages above any potential climate-related floods
  • from the Adaptation for Smallholder Agricultural Programme, in 2015, of $4.56 million to Liberia to improve the climate resilience of its cocoa and coffee crops

Moving outward through the other layers of the onion, though, the details get less granular, Schalatek said. “Lots of bilateral initiatives, for instance, don’t give you a project-by-project breakdown.” A report by the non-profit Climate Policy Initiative (CPI), analyzing climate financing initiatives in 2019 and 2020, found, for instance, that “finance for buildings with high energy and thermal insulation  performances—green buildings—is growing fast but lacks transparency.” Even many multilateral funding efforts can only be slotted into broad categories: agricultural development, for example, or disaster risk reduction. As a result, Schalatek said, “there’s less data accuracy, more guesswork.”

Are rich countries greenwashing their climate funding pledges?

Another source of doubt has to do with how rich countries label the money they’re disbursing. Under the OECD system, countries attach a so-called “Rio marker” to any funds they claim to be pledging under the climate assistance rubric. The marker can either be a “2,” to signal that the money’s main purpose is climate-related, or a “1,” to signal that a significant part of the money’s purpose is climate-related.

But Schalatek pointed out that no standards exist for what the term “significant” might mean. Each country merely decides this for itself. “That makes the ‘significant’ tag very, very fishy,” she said. “There’s a risk that countries will overuse the ‘significant’ marker and inflate the overall amount of money they’re giving. There has been a lot of mislabelling—or uplabelling, you could call it—going on.”

Indeed, a study of the OECD database reveals projects that sound like traditional forms of assistance but that have been described as climate-related. Finland, for instance, gave Ethiopia a grant of $4.47 million to develop its water supply and sanitation systems, and labeled the money with a “1.” But water aid projects have existed for decades, and it is arguable whether the purpose of this grant—at least as described in the OECD database—has “significant” relation to climate change.

Some examples are even more egregious. Baysa Naran, a senior analyst who co-authored the CPI report, pointed out that her team had to regularly leave out projects that were tagged as climate financing but that were nothing of the sort: “energy-efficient coal, for instance.”

At Glasgow, developing countries called for tighter definitions of climate financing, and the United Nations Framework Convention on Climate Change is attempting to bring more transparency to the funding initiatives. There are so many undefined areas, Naran said. “Which sectors do you count for mitigation, and which for adaptation? Do you count loans, or should only the equivalents of grants be considered climate finance?

Some countries channel a lot of money through multilateral institutions, which then channel money to climate finance. In those situations, how do you account for each country’s contribution?”

Measuring what countries are ponying up thus becomes very tricky, Naran noted. “There’s no third party that reviews and audits and makes independent conclusions about this financing,” she said. “It’s going to be a very difficult topic to reach a united agreement on.”

Samanth Subramanian

By Samanth Subramanian

Source: How are developing countries spending climate financing money? — Quartz

.

AI Can Write Code Like Humans Bugs and All

Some software developers are now letting artificial intelligence help write their code. They’re finding that AI is just as flawed as humans.

Last June, GitHub, a subsidiary of Microsoft that provides tools for hosting and collaborating on code, released a beta version of a program that uses AI to assist programmers. Start typing a command, a database query, or a request to an API, and the program, called Copilot, will guess your intent and write the rest.

Alex Naka, a data scientist at a biotech firm who signed up to test Copilot, says the program can be very helpful, and it has changed the way he works. “It lets me spend less time jumping to the browser to look up API docs or examples on Stack Overflow,” he says. “It does feel a little like my work has shifted from being a generator of code to being a discriminator of it.”

But Naka has found that errors can creep into his code in different ways. “There have been times where I’ve missed some kind of subtle error when I accept one of its proposals,” he says. “And it can be really hard to track this down, perhaps because it seems like it makes errors that have a different flavor than the kind I would make.”

The risks of AI generating faulty code may be surprisingly high. Researchers at NYU recently analyzed code generated by Copilot and found that, for certain tasks where security is crucial, the code contains security flaws around 40 percent of the time.

The figure “is a little bit higher than I would have expected,” says Brendan Dolan-Gavitt, a professor at NYU involved with the analysis. “But the way Copilot was trained wasn’t actually to write good code—it was just to produce the kind of text that would follow a given prompt.”

Despite such flaws, Copilot and similar AI-powered tools may herald a sea change in the way software developers write code. There’s growing interest in using AI to help automate more mundane work. But Copilot also highlights some of the pitfalls of today’s AI techniques.

While analyzing the code made available for a Copilot plugin, Dolan-Gavitt found that it included a list of restricted phrases. These were apparently introduced to prevent the system from blurting out offensive messages or copying well-known code written by someone else.

Oege de Moor, vice president of research at GitHub and one of the developers of Copilot, says security has been a concern from the start. He says the percentage of flawed code cited by the NYU researchers is only relevant for a subset of code where security flaws are more likely.

De Moor invented CodeQL, a tool used by the NYU researchers that automatically identifies bugs in code. He says GitHub recommends that developers use Copilot together with CodeQL to ensure their work is safe.

The GitHub program is built on top of an AI model developed by OpenAI, a prominent AI company doing cutting-edge work in machine learning. That model, called Codex, consists of a large artificial neural network trained to predict the next characters in both text and computer code. The algorithm ingested billions of lines of code stored on GitHub—not all of it perfect—in order to learn how to write code.

OpenAI has built its own AI coding tool on top of Codex that can perform some stunning coding tricks. It can turn a typed instruction, such as “Create an array of random variables between 1 and 100 and then return the largest of them,” into working code in several programming languages.

Another version of the same OpenAI program, called GPT-3, can generate coherent text on a given subject, but it can also regurgitate offensive or biased language learned from the darker corners of the web.

Copilot and Codex have led some developers to wonder if AI might automate them out of work. In fact, as Naka’s experience shows, developers need considerable skill to use the program, as they often must vet or tweak its suggestions.

Hammond Pearce, a postdoctoral researcher at NYU involved with the analysis of Copilot code, says the program sometimes produces problematic code because it doesn’t fully understand what a piece of code is trying to do. “Vulnerabilities are often caused by a lack of context that a developer needs to know,” he says.

Some developers worry that AI is already picking up bad habits. “We have worked hard as an industry to get away from copy-pasting solutions, and now Copilot has created a supercharged version of that,” says Maxim Khailo, a software developer who has experimented with using AI to generate code but has not tried Copilot.

Khailo says it might be possible for hackers to mess with a program like Copilot. “If I was a bad actor, what I would do would be to create vulnerable code projects on GitHub, artificially boost their popularity by buying GitHub stars on the black market, and hope that it will become part of the corpus for the next training round.”

Both GitHub and OpenAI say that, on the contrary, their AI coding tools are only likely to become less error prone. OpenAI says it vets projects and code both manually and using automated tools.

De Moor at GitHub says recent updates to Copilot should have reduced the frequency of security vulnerabilities. But he adds that his team is exploring other ways of improving the output of Copilot. One is to remove bad examples that the underlying AI model learns from. Another may be to use reinforcement learning, an AI technique that has produced some impressive results in games and other areas, to automatically spot bad output, including previously unseen examples. “Enormous improvements are happening,” he says. “It’s almost unimaginable what it will look like in a year.”

Source: AI Can Write Code Like Humans—Bugs and All | WIRED

.

Related Contents:

%d bloggers like this: