/2020

  • How Google Docs became the social media of the resistance | MIT Technology Review
    https://www.technologyreview.com/2020/06/06/1002546/google-docs-social-media-resistance

    In the week after George Floyd’s murder, hundreds of thousands of people joined protests across the US and around the globe, demanding education, attention, and justice. But one of the key tools for organizing these protests is a surprising one: it’s not encrypted, doesn’t rely on signing in to a social network, and wasn’t even designed for this purpose. It’s Google Docs.

    In just the last week, Google Docs has emerged as a way to share everything from lists of books on racism to templates for letters to family members and representatives to lists of funds and resources that are accepting donations. Shared Google Docs that anyone can view and anyone can edit, anonymously, have become a valuable tool for grassroots organizing during both the coronavirus pandemic and the police brutality protests sweeping the US. It’s not the first time. In fact, activists and campaigners have been using the word processing software for years as a more efficient and accessible protest tool than either Facebook or Twitter.

    It wasn’t until the 2016 elections, when misinformation campaigns were rampant, that the software came into its own as a political tool. Melissa Zimdars, an assistant professor of communication at Merrimack College, used it to create a 34-page document titled “False, Misleading, Clickbaity-y, and/or Satirical ‘News’ Sources.’”

    Zimdars inspired a slew of political Google Docs, written by academics as ad hoc ways of campaigning for Democrats for the 2018 midterm elections. By the time the election passed, Google Docs were also being used to protest immigration bans and advance the #MeToo movement.

    Now, in the wake of George Floyd’s murder on Memorial Day weekend, communities are using the software to organize. One of the most popular Google Docs to emerge in the past week is “Resources for Accountability and Actions for Black Lives,” which features clear steps people can take to support victims of police brutality. It is organized by Carlisa Johnson, a 28-year-old graduate journalism student at Georgia State University.

    Indigo said accessibility and live editing were the primary advantages of a Google Doc over social media: “It’s important to me that the people on the ground can access these materials, especially those seeking legal counsel, jail support, and bail support. This is a medium that everyone I’ve organized with uses and many others use.”

    “What’s special about a Google Doc versus a newsfeed is its persistence and editability,” says Clay Shirky, the vice provost for educational technology at New York University. In 2008, Shirky wrote Here Comes Everybody: The Power of Organizing Without Organizations, detailing how the internet and social media helped shape modern protest movements.

    Shirky says that while social media has been great for publicizing movements, it’s far less efficient at creating stable shelves of information that a person can return to. What makes Google Docs especially attractive is that they are at once dynamic and static, he says. They’re editable and can be viewed simultaneously on countless screens, but they are easily shareable via tweet or post.

    “People want a persistent artifact,” Shirky says. “If you are in an action-oriented network, you need an artifact to coordinate with those outside of the conversation and the platform you’re using, so you can actually go outside of the feed and do something.”

    It helps that Google Docs are fairly straightforward to access and simple to use. But anonymity is an important advantage over Twitter or Facebook. Users who click on a publicly shareable link are assigned an animal avatar, hiding their identity. “No one can put you on blast on Google Docs,” says Shirky. “Google Docs allows for a wider breadth of participation for people who are not looking to get into a high-stakes political argument in front of millions of people.”

    Un passage intéressant sur l’anonymat dans GD. Car nous savons que Google a les moyens de nous pister... mais l’important n’est pas ici que Google sache qui écrit, mais que chacun n’ait pas à subir les trolls. Une situation à méditer.

    For both Johnson and Indigo, the overall experience of creating Google Docs has been a surprisingly positive one; Indigo does receive the occasional “nasty DM” but shrugs it off. At any given moment, anywhere between 70 and 90 people are in Johnson’s and Indigo’s documents, and both spend significant time editing and fact-checking them.

    Shirky says it’s a common misconception that protesters are seeking privacy from the state. “Most of them are concerned with activism, not privacy,” he says. In fact, Johnson says that for her and other activists, the goal is to disseminate as much information as accurately as possible.

    “Google Docs lets me put it in one place and across social-media platforms,” she says. “Reach is what’s important at this time. A Facebook post can only go so far. An Instagram post can only go so far. But this? This is accessible. Nothing else is as immediate.”

    #Google_doc #Activisme #Document_numérique

  • Of course technology perpetuates racism. It was designed that way. | MIT Technology Review
    https://www.technologyreview.com/2020/06/03/1002589/technology-perpetuates-racism-by-design-simulmatics-charlton-mcilw

    We often call on technology to help solve problems. But when society defines, frames, and represents people of color as “the problem,” those solutions often do more harm than good. We’ve designed facial recognition technologies that target criminal suspects on the basis of skin color. We’ve trained automated risk profiling systems that disproportionately identify Latinx people as illegal immigrants. We’ve devised credit scoring algorithms that disproportionately identify black people as risks and prevent them from buying homes, getting loans, or finding jobs.

    So the question we have to confront is whether we will continue to design and deploy tools that serve the interests of racism and white supremacy,

    Of course, it’s not a new question at all.

    As part of a DARPA project aimed at turning the tide of the Vietnam War, Pool’s company had been hard at work preparing a massive propaganda and psychological campaign against the Vietcong. President Johnson was eager to deploy Simulmatics’s behavioral influence technology to quell the nation’s domestic threat, not just its foreign enemies. Under the guise of what they called a “media study,” Simulmatics built a team for what amounted to a large-scale surveillance campaign in the “riot-affected areas” that captured the nation’s attention that summer of 1967.

    Three-member teams went into areas where riots had taken place that summer. They identified and interviewed strategically important black people. They followed up to identify and interview other black residents, in every venue from barbershops to churches. They asked residents what they thought about the news media’s coverage of the “riots.” But they collected data on so much more, too: how people moved in and around the city during the unrest, who they talked to before and during, and how they prepared for the aftermath. They collected data on toll booth usage, gas station sales, and bus routes. They gained entry to these communities under the pretense of trying to understand how news media supposedly inflamed “riots.” But Johnson and the nation’s political leaders were trying to solve a problem. They aimed to use the information that Simulmatics collected to trace information flow during protests to identify influencers and decapitate the protests’ leadership.

    They didn’t accomplish this directly. They did not murder people, put people in jail, or secretly “disappear” them.

    But by the end of the 1960s, this kind of information had helped create what came to be known as “criminal justice information systems.” They proliferated through the decades, laying the foundation for racial profiling, predictive policing, and racially targeted surveillance. They left behind a legacy that includes millions of black and brown women and men incarcerated.

    #Racisme #Intelligence_artificielle #capitalisme_surveillance #surveillance

  • First the trade war, then the pandemic. Now Chinese manufacturers are turning inward. | MIT Technology Review
    https://www.technologyreview.com/2020/06/03/1002573/pandemic-us-china-trade-war-impact-on-manufacturers

    Then, before his business had fully recovered, covid-19 ripped through the world. Exports tanked, saddling Zhu with a stream of order cancellations worth an estimated $4 to $5 million. Domestic sales also suffered as physical stores shuttered under pandemic control restrictions. “The impact could’ve been huge,” he says. “My factory is really big; I have so many workers to support.”

    But Zhu fortunately had another sales channel. In 2018, Pinduoduo, an e-commerce giant targeted at consumers in China’s smaller cities, launched an initiative to connect manufacturers with the domestic market. Under a so-called “consumer-to-manufacturer,” or C2M, model, the platform began using its massive pools of data and AI algorithms to help Chinese manufacturers predict consumer preferences and develop brands specifically for a domestic audience.

    Pinduoduo told manufacturers not only how to customize their products—down to the wash of a jean or the length of a sock. It also advised them on how to redesign their packaging, how to set their prices, and how to market their goods online. In this way, manufacturers could improve the efficiency of their production, which in turn made the products cheaper for consumers. And platforms could monetize new users with advertising. This helped both the platform and manufacturers alike tap into a rapidly growing middle-class consumer base. Whereas upper-class consumers care more about international brands, this newer wave of consumers care more about quality products at lower prices.

    When the pandemic hit, Pinduoduo quickly expanded its initiative. It added new incentives for affected manufacturers to join its platform, welcoming them to adopt its live-streaming service (link in Chinese) and holding promotional sales events.

    As China’s access to international markets has grown more unreliable—with a possible trade fight renewal looming on the horizon—the country has increasingly sought to ramp up domestic consumption in an effort to stave off a greater economic recession.

    “The problem is China is losing [overseas] demand,” says Derek Scissors, a resident scholar at the American Enterprise Institute, where he researches trade policy and US-China relations. “You want to replace it with Chinese demand.”

    As well as Pinduoduo, other Chinese e-commerce giants, including Alibaba-owned Taobao and JD, are now offering C2M services. Since the start of this year, all three have set new goals for expanding their C2M initiatives. Pinduoduo, which helped launch 106 manufacturer-owned brands in 2019, aims to establish 1,000 more. It also signed a strategic partnership in April with the government of Dongguan, where Zhu’s factory is based, one of China’s largest manufacturing hubs.

    As the partnerships have produced promising results, manufacturers have also doubled down on their domestic brand strategies. Chen Zhuoyue, the owner of a toy manufacturing company based in Chenghai, Guangdong, joined JD’s C2M program in 2018. After JD helped him customize his products and develop a new pricing strategy, the platform quickly grew to account for 50% of his domestic sales. When the pandemic hit and his exports sharply declined from 30% to less than 5% of his revenue, he took it as a sign to open up two new JD stores and launch more domestic brands.

    It’s not that Chen will stop working with foreign brands. “As a businessman, I’m always thinking about how to expand into more markets,” he says. If exports were to return back to normal and his long-term foreign collaborators came knocking on his door, he would gladly continue fulfilling their orders. At the same time, now that he’s launched his own brand, he sees it as an important source of growth and stability. “My plan is to expand our domestic presence,” he says. “This year I want to increase our investment in this area.”

    It’s not clear whether domestic markets alone will be able to compensate for China’s variable access to international markets over the long run. On one hand, the country’s middle class has rapidly increased their spending power and is expected to grow to a market size of 1,008 billion RMB ($141 million) by 2022, according to iResearch. On the other, even before the pandemic, the manufacturing industry was already struggling with too much supply, says Scissors, and it relied on the US and other overseas markets to “dump their manufacturing excess,” he says. As a result, he’s unconvinced that a new model like C2M would resolve such deeply-entrenched macroeconomic issues. If anything, he sees C2M instead as a savvy push from e-commerce giants to grow their own profits.

    #Chine #Commerce_électronique #C2M #Consumer-to-manufacturer #Marché_interieur #Mondialisation

  • This startup is using AI to give workers a “productivity score” | MIT Technology Review
    https://www.technologyreview.com/2020/06/04/1002671/startup-ai-workers-productivity-score-bias-machine-learning-busine

    Dire qu’il y a des naïfs pour croire que le crédit social est uniquement chinois... surveiller et noter les travailleurs, c’est le nouveau modèle du capitalisme international, en Chine comme ailleurs, en télétravail comme dans les locaux de l’entreprise. Et ça va vite, vite...

    In the last few months, millions of people around the world stopped going into offices and started doing their jobs from home. These workers may be out of sight of managers, but they are not out of mind. The upheaval has been accompanied by a reported spike in the use of surveillance software that lets employers track what their employees are doing and how long they spend doing it.

    Companies have asked remote workers to install a whole range of such tools. Hubstaff is software that records users’ keyboard strokes, mouse movements, and the websites that they visit. Time Doctor goes further, taking videos of users’ screens. It can also take a picture via webcam every 10 minutes to check that employees are at their computer. And Isaak, a tool made by UK firm Status Today, monitors interactions between employees to identify who collaborates more, combining this data with information from personnel files to identify individuals who are “change-makers.”

    Now, one firm wants to take things even further. It is developing machine-learning software to measure how quickly employees complete different tasks and suggest ways to speed them up. The tool also gives each person a productivity score, which managers can use to identify those employees who are most worth retaining—and those who are not.

    How you feel about this will depend on how you view the covenant between employer and employee. Is it okay to be spied on by people because they pay you? Do you owe it to your employer to be as productive as possible, above all else?

    Critics argue that workplace surveillance undermines trust and damages morale. Workers’ rights groups say that such systems should only be installed after consulting employees. “It can create a massive power imbalance between workers and the management,” says Cori Crider, a UK-based lawyer and cofounder of Foxglove, a nonprofit legal firm that works to stop governments and big companies from misusing technology. “And the workers have less ability to hold management to account.”

    Whatever your views, this kind of software is here to stay—in part because remote work is normalizing it. “I think workplace monitoring is going to become mainstream,” says Tommy Weir, CEO of Enaible, the startup based in Boston that is developing the new monitoring software. “In the next six to 12 months it will become so pervasive it disappears.”

    Weir thinks most tools on the market don’t go far enough. “Imagine you’re managing somebody and you could stand and watch them all day long, and give them recommendations on how to do their job better,” says Weir. “That’s what we’re trying to do. That’s what we’ve built.”

    Why the sudden uptick in interest? “Bosses have been seeking to wring every last drop of productivity and labor out of their workers since before computers,” says Crider. “But the granularity of the surveillance now available is like nothing we’ve ever seen.”

    It’s no surprise that this level of detail is attractive to employers, especially those looking to keep tabs on a newly remote workforce. But Enaible’s software, which it calls the AI Productivity Platform, goes beyond tracking things like email, Slack, Zoom, or web searches. None of that shows a full picture of what a worker is doing, says Weir⁠—it’s just checking if you are working or not.

    Once set up, the software runs in the background all the time, monitoring whatever data trail a company can provide for each of its employees. Using an algorithm called Trigger-Task-Time, the system learns the typical workflow for different workers: what triggers, such as an email or a phone call, lead to what tasks and how long those tasks take to complete.

    Once it has learned a typical pattern of behavior for an employee, the software gives that person a “productivity score” between 0 and 100. The AI is agnostic to tasks, says Weir. In theory, workers across a company can still be compared by their scores even if they do different jobs. A productivity score also reflects how your work increases or decreases the productivity of other people on your team. There are obvious limitations to this approach. The system works best with employees who do a lot of repetitive tasks in places like call centers or customer service departments rather than those in more complex or creative roles.

    But the idea is that managers can use these scores to see how their employees are getting on, rewarding them if they get quicker at doing their job or checking in with them if performance slips. To help them, Enaible’s software also includes an algorithm called Leadership Recommender, which identifies specific points in an employee’s workflow that could be made more efficient.

    #Travail #Surveillance #Droit_travail #Crédit_social #Productivity_score

  • Nearly 40% of Icelanders are using a covid app—and it hasn’t helped much
    https://www.technologyreview.com/2020/05/11/1001541/iceland-rakning-c19-covid-contact-tracing

    The country has the highest penetration of any automated contact tracing app in the world, but one senior figure says it “wasn’t a game changer.” When Iceland got its first case of covid-19 on February 28, an entire apparatus sprang into action. The country had already been testing some people at high risk of catching the virus, thanks to DeCode genetics, a local biotech company. Once the arrival of the disease was confirmed, it began rapidly rolling out public testing on a much wider scale. (...)

    #Apple #Google #algorithme #Bluetooth #smartphone #GPS #contactTracing #technologisme #consentement #COVID-19 (...)

    ##santé

  • India is forcing people to use its covid app, unlike any other democracy
    https://www.technologyreview.com/2020/05/07/1001360/india-aarogya-setu-covid-app-mandatory

    Millions of Indians have no choice but to download the country’s tracking technology if they want to keep their jobs or avoid reprisals. The world has never seen anything quite like Aarogya Setu. Two months ago, India’s app for coronavirus contact tracing didn’t exist ; now it has nearly 100 million users. Prime Minister Narendra Modi boosted it on release by urging every one of the country’s 1.3 billion people to download it, and the result was that within two weeks of launch it became the (...)

    #algorithme #AarogyaSetu_ #Bluetooth #smartphone #GPS #contactTracing #géolocalisation #technologisme #consentement #BigData #COVID-19 (...)

    ##santé

  • Podcast : Who watches the pandemic watchers ? We do
    https://www.technologyreview.com/2020/05/20/1001927/podcast-who-watches-the-pandemic-watchers-we-do/?truid=e240178e6fc656e71bbee1dbf6ce3de7

    No sooner had the stay-at-home orders come down than mobile app developers around the world began to imagine how our smartphones could make it safer for everyone to venture back out. Dozens of countries and a handful of US states are now urging citizens to download government-blessed apps that use GPS-based location tracking, the Bluetooth wireless standard, or a combination of both to alert us when we’ve crossed paths with an infected individual—information that could tell us when we need to (...)

    #Apple #Google #algorithme #Bluetooth #smartphone #GPS #contactTracing #géolocalisation #BigData #COVID-19 #santé #ACLU (...)

    ##santé ##technologisme

  • Why contact tracing may be a mess in America
    https://www.technologyreview.com/2020/05/16/1001787/why-contact-tracing-may-be-a-mess-in-america/?truid=e240178e6fc656e71bbee1dbf6ce3de7

    High caseloads, low testing, and American attitudes toward government authority could pose serious challenges for successful efforts to track and contain coronavirus cases. Technology can certainly supplement human contact tracing. Smartphone apps that flag when someone may have been in close contact with an infected person helped China, which required citizens in many cities to download the software, to flatten the curve of its outbreak. Similarly, South Korean officials have made use of (...)

    #MIT #Bluetooth #smartphone #GPS #contactTracing #géolocalisation #consentement #BigData #COVID-19 #santé (...)

    ##santé ##technologisme

  • Our weird behavior during the pandemic is messing with AI models
    https://www.technologyreview.com/2020/05/11/1001563/covid-pandemic-broken-ai-machine-learning-amazon-retail-fraud-huma

    Our weird behavior during the pandemic is messing with AI models Machine-learning models trained on normal behavior are showing cracks —forcing humans to step in to set them straight. In the week of April 12-18, the top 10 search terms on Amazon.com were : toilet paper, face mask, hand sanitizer, paper towels, Lysol spray, Clorox wipes, mask, Lysol, masks for germ protection, and N95 mask. People weren’t just searching, they were buying too—and in bulk. The majority of people looking for (...)

    #Amazon #algorithme #technologisme #fraude #consommation #COVID-19 #profiling #santé (...)

    ##santé ##bug

  • The secret to why some people get so sick from covid could lie in their genes | MIT Technology Review
    https://www.technologyreview.com/2020/05/13/1001653/23andme-looks-for-covid-19-genetic-clues

    23andMe, le compagnie de l’ex-femme de Sergey Brin et largement promue par Google a décidé de profiter de l’aubaine pour augmenter sa place de number one dans le traçage génétique. Pour le bien commun, évidemment. Le capitalisme génomique dans toute sa splendeur.

    Some people die from covid-19, and others who are infected don’t even show symptoms. But scientists still don’t know why.

    Now consumer genomics company 23andMe is going to offer free genetic tests to 10,000 people who’ve been hospitalized with the disease, hoping to turn up genetic factors that could point to an answer.

    While it’s known that older people and those with health conditions such as diabetes are most at risk, there could be hidden genetic reasons why some young, previously healthy people are also dying.

    23andMe operates a large gene database with more than 8 million customers, many of whom have agreed to let their data be used for research. The company has previously used consumer data to power searches for the genetic roots of insomnia, homosexuality, and other traits.

    Ouh là la, que d’affirmations génomiques :

    Scientists hope to find a gene that strongly influences, or even determines, how badly people are affected by the coronavirus. There are well-known examples of such genetic effects on other diseases: for example, sickle-cell genes confer resistance to malaria, and variants of other genes are known to protect people from HIV or to norovirus, an intestinal germ.

    L’étude anti-scientifique et non-éthique sur l’homosexualité ne leur a pas suffit. Ces gens sont vraiment des rapaces du nouveau monde.

    #23andMe #Capitalisme_génomique

  • Facebook’s AI is still largely baffled by covid misinformation | MIT Technology Review
    https://www.technologyreview.com/2020/05/12/1001633/ai-is-still-largely-baffled-by-covid-misinformation

    Tiens, l’IA ne serait pas à la hauteur pour assurer la modération de contenu. Il faut des humains pour comprendre l’humanité. Quelle découverte miraculeuse. On est vraiment au XXIe siècle, je crois.

    The news: In its latest Community Standards Enforcement Report, released today, Facebook detailed the updates it has made to its AI systems for detecting hate speech and disinformation. The tech giant says 88.8% of all the hate speech it removed this quarter was detected by AI, up from 80.2% in the previous quarter. The AI can remove content automatically if the system has high confidence that it is hate speech, but most is still checked by a human being first.

    Behind the scenes: The improvement is largely driven by two updates to Facebook’s AI systems. First, the company is now using massive natural-language models that can better decipher the nuance and meaning of a post. These models build on advances in AI research within the last two years that allow neural networks to be trained on language without any human supervision, getting rid of the bottleneck caused by manual data curation.

    The second update is that Facebook’s systems can now analyze content that consists of images and text combined, such as hateful memes. AI is still limited in its ability to interpret such mixed-media content, but Facebook has also released a new data set of hateful memes and launched a competition to help crowdsource better algorithms for detecting them.

    Covid lies: Despite these updates, however, AI hasn’t played as big a role in handling the surge of coronavirus misinformation, such as conspiracy theories about the virus’s origin and fake news of cures. Facebook has instead relied primarily on human reviewers at over 60 partner fact-checking organizations. Only once a person has flagged something, such as an image with a misleading headline, do AI systems take over to search for identical or similar items and automatically add warning labels or take them down. The team hasn’t yet been able to train a machine-learning model to find new instances of disinformation itself. “Building a novel classifier for something that understands content it’s never seen before takes time and a lot of data,” Mike Schroepfer, Facebook’s CTO, said on a press call.

    Why it matters: The challenge reveals the limitations of AI-based content moderation. Such systems can detect content similar to what they’ve seen before, but they founder when new kinds of misinformation appear. In recent years, Facebook has invested heavily in developing AI systems that can adapt more quickly, but the problem is not just the company’s: it remains one of the biggest research challenges in the field.

    #Intelligence_artificielle #Facebook #Modération

  • How covid-19 conspiracy theorists are exploiting YouTube culture | MIT Technology Review
    https://www.technologyreview.com/2020/05/07/1001252/youtube-covid-conspiracy-theories/?truid=a497ecb44646822921c70e7e051f7f1a

    Covid-19 conspiracy theorists are still getting millions of views on YouTube, even as the platform cracks down on health misinformation.

    The answer was obvious to Kennedy, one of many anti-vaccination leaders trying to make themselves as visible as possible during the covid-19 pandemic. “I’d love to talk to your audience,” he replied.

    Kennedy told Bet-David that he believes his own social-media accounts have been unfairly censored; making an appearance on someone else’s popular platform is the next best thing. Bet-David framed the interview as an “exclusive,” enticingly titled “Robert Kennedy Jr. Destroys Big Pharma, Fauci & Pro-Vaccine Movement.” In two days, the video passed half a million views.

    As of Wednesday, advertisements through YouTube’s ad service were playing before the videos, and Bet-David’s merchandise was for sale in a panel below the video’s description. Two other interviews, in which anti-vaccine figures aired several debunked claims about coronavirus and vaccines (largely unchallenged by Bet-David), were also showing ads. Bet-David said in an interview that YouTube had limited ads on all three videos, meaning they can generate revenue, but not as much as they would if they were fully monetized.

    We asked YouTube for comment on all three videos on Tuesday afternoon. By Thursday morning, one of the three (an interview with anti-vaccine conspiracy theorist Judy Mikovits) had been deleted for violating YouTube’s medical misinformation policies. Before it was deleted, the video had more than 1 million views.

    YouTube said that the other two videos were borderline, meaning that YouTube decided they didn’t violate rules, but would no longer be recommended or show up prominently in search results.

    I asked Bet-David whether he felt any responsibility over airing these views on his channel—particularly potentially harmful claims by his guests, urging viewers to ignore public health recommendations.

    “I do not,” he said. “I am responsible for what comes out of my mouth. I’m not responsible for what comes out of your mouth”

    For him, that lack of responsibility extends to misinformation that could be harmful to his audience. He is just giving people what they are asking for. That, in turn, drives attention, which allows him to make money from ads, merchandise, speaking gigs, and workshops. “It’s up to the audience to make the decision for themselves,” he says. Besides, he thinks he’s done interviewing anti-vaccine activists for now. He’s trying to book some “big name” interviews of what he termed “pro-vaccine” experts.

    #YouTube #Complotisme #Vaccins #Médias_sociaux #Fake_news

  • #Covid-19 has blown apart the myth of #Silicon_Valley #innovation | MIT Technology Review
    https://www.technologyreview.com/2020/04/25/1000563/covid-19-has-killed-the-myth-of-silicon-valley-innovation

    The pandemic has made clear this festering problem: the US is no longer very good at coming up with new ideas and technologies relevant to our most basic needs. We’re great at devising shiny, mainly software-driven bling that makes our lives more convenient in many ways. But we’re far less accomplished at reinventing health care, rethinking education, making food production and distribution more efficient, and, in general, turning our technical know-how loose on the largest sectors of the economy.

    #progrès #éducation #santé #alimentation

  • Covid hoaxes are using a loophole to stay alive—even after being deleted | MIT Technology Review
    https://www.technologyreview.com/2020/04/30/1000881/covid-hoaxes-zombie-content-wayback-machine-disinformation

    Pandemic conspiracy theorists are using the Wayback Machine to promote ’zombie content’ that avoids content moderators and fact-checkers.

    by Joan Donovan archive page
    April 30, 2020

    PHOTO BY CHRIS HALL ON UNSPLASH
    Since the onset of the pandemic, the Technology and Social Change Research Project at Harvard Kennedy’s Shorenstein Center, where I am the director, has been investigating how misinformation, scams, and conspiracies about covid-19 circulate online. If fraudsters are now using the virus to dupe unsuspecting individuals, we thought, then our research on misinformation should focus on understanding the new tactics of these media manipulators. What we found was a disconcerting explosion in “zombie content.”

    While the original page failed to spread fake news, the version of the page saved on the Internet Archive’s Wayback Machine absolutely flourished on Facebook. With 649,000 interactions and 118,000 shares, the engagement on the Wayback Machine’s link was much larger than legitimate press outlets. Facebook has since placed a fact-check label over the link to the Wayback Machine link too, but it had already been seen a huge number of times.

    There are several explanations for this hidden virality. Some people use the Internet Archive to evade blocking of banned domains in their home country, but it is not simply about censorship. Others are seeking to get around fact-checking and algorithmic demotion of content.

    When looking for more evidence of hidden virality, we searched for “web.archive.org” across platforms. Unsurprisingly, Medium posts that were taken down for spreading health misinformation have found new life through Wayback Machine links. One deleted Medium story, “Covid-19 had us all fooled, but now we might have finally found its secret,” violated Medium’s policies on misleading health information. Before Medium’s takedown, the original post amassed 6,000 interactions and 1,200 shares on Facebook, but the archived version is vastly more popular—1.6 million interactions, 310,000 shares, and still climbing. This zombie content has better performance than most mainstream media news stories and, yet it only exists as an archived record.

    Perhaps the most alarming element to a researcher like me is that these harmful conspiracies permeate private pages and groups on Facebook. This means researchers have access to less than 2 % of the interaction data, and that health misinformation circulates in spaces where journalists, independent researchers and public health advocates can not assess or counterbalance these false claims with facts. Crucially, if it weren’t for the Internet Archive’s records we would not be able to do this research on deleted content in the first place, but these use cases suggest that the Internet Archive will soon have to address how their service can be adapted to deal with disinformation.

    Hidden virality is growing in places where Whatsapp is popular because it’s easy to forward misinformation through encrypted channels and evade content moderation. But when hidden virality happens on Facebook with health misinformation, it is particularly disconcerting. More than 50% of Americans rely on Facebook for their news, and still, after many years of concern and complaint, researchers have a very limited window into the data. This means it’s nearly impossible to ethically investigate how dangerous health misinformation is shared on private pages and groups.

    All of this is a threat for public health in a different way than political or news misinformation, because people do quickly change their behaviors based on medical recommendations.

    #Fake_news #Viralité #Internet_archive #zombie_content #Joan_Donovan