• Pourquoi j’ai quitté « le job le plus sexy du XXIè siècle » | Le Club de Mediapart
    #IA #deep_learning #algos #statistiques #data #big_data

    "Le rapport qui suit s’intéresse aux causes mais surtout aux effets de l’automatisation algorithmique et statistique de l’économie mondialisée, posant notamment le problème de maintenir coûte que coûte le dogme du progrès technologique comme horizon indiscutable de nos sociétés. Trois ans dans le monde des data sciences, en tant que consultant #data_scientist au sein du cabinet Sia Partners, m’ont fait découvrir l’intérêt porté au secteur par les grands groupes de l’énergie, de la banque et de l’assurance, mais aussi par l’administration publique et l’écosystème #start-up.
    A travers cette analyse, rédigée après ma démission en mars 2020, j’ai essayé d’exposer ma compréhension de ce qui est actuellement en jeu dans l’économie du numérique. J’y présente un témoignage personnel enrichi de réflexions théoriques documentées, afin de démystifier les fantasmes à la source du #solutionnisme_technologique et de caractériser les effets des récentes innovations sur nos sociétés et, plus largement, sur le vivant. Cette prise de recul à l’aune d’un examen idéologique tente d’exposer les dysfonctionnements d’un système ayant placé le progrès technologique au cœur de sa doctrine et de son rapport au temps. La fin du rapport interroge en particulier notre conception du temps comme point de bascule paradigmatique.
    Il ne s’agit évidemment pas d’une croisade à l’encontre de mon ancien employeur, qui ne revêt à mon sens aucune importance systémique, même si telle peut être l’ambition de ses dirigeants. Les mêmes considérations symptomatiques auraient pu être faites au sein de la concurrence. Il s’agit ici d’informer concrètement celles et ceux qui n’auraient pas toutes les cartes en mains, et d’interpeller la conscience des autres. Les éléments présentés sont, autant que possible, factuellement argumentés. Les assertions et observations critiques sont le fruit d’une réflexion personnelle documentée, et de ce fait, comme tout écrit, empruntes d’une certaine forme de subjectivité. Néanmoins, l’objectif est conservé d’y établir un discours plus
    rationnel que les croyances maintenues à bout de bras par une culture dominante en péril."




    • Une nouvelle définition politique du soin (ici pour une trad viteuf)

      For many of us, the last few weeks have marked a new phase of our corona-lives—a dark and lonely corridor that stretches before us, no end in sight. Earlier, we counted this crisis in days and weeks. Now we are coming to see that this virus will in all likelihood be with us for months and years. We can’t stand social distancing any longer, but we also can’t stop, because there is no infrastructure in place to safely allow us to go back to school and work.

      A Community Health Corps is one place to start to build a new movement that heals us and our body politic, and that will allow us—all of us—to survive a pandemic, and then, to thrive.
      Our federal leadership remains ruinous. President Trump, obsessed with ratings, still cannot seem to think beyond the twenty-four-hour news cycle. In the last week he first insisted he would reopen things in May, then abandoned the idea, perhaps having learned that he lacked the necessary power. He then cast around for others to blame, taking to Twitter to cheer on tiny and malevolent groups of protesters calling for a reopening the economy, damn the consequences. Tragically, in the wake of the president’s remarks, Governor Brian Kemp of Georgia announced he would let many businesses resume operations, though the state is flush with new cases, and there is no viable plan for containment going forward. Trump tried to walk back his remarks, saying he disagreed with Kemp, but the damage was done, and Georgia is proceeding full-steam ahead. The press to return to school and work will only intensify, for all of us—while Georgia, and other states that are making similar rumblings, have nothing to offer their citizens but decimation.

      What other way forward is there, over these coming months? As in the early phase, leadership and vision is going to come from elsewhere. It’ll come from reality-based local leaders, perhaps from Congress, and from us. As the timescale of our response to COVID-19 shifts to months and years, it’s time to ask: The day after all this is over, what do we want the world we share to look like? What are we willing to fight for? And how do we connect a long-term vision of that world worth fighting for with the things we need to do to mitigate the damage now?

      section separator

      Any response to the moment has to address two, twinned crises: the threat of a virus run out of control, and the carnage being visited on working people and families by the measures we need to undertake to contain the virus. While COVID-19 cuts its deadly swath from coast to coast, the disease follows the same patterns of inequality we’ve always seen embedded in the U.S. landscape, where the death rate for predominantly African American counties is six-fold higher than in predominantly white counties across the country, and where this crisis is just heaped upon others, which have been plaguing these communities for generations.

      Meanwhile, as millions of Americans stay at home in solidarity with their neighbors to protect them from infection, the economic contraction has come at great cost to families and individuals, dragging them to the brink in the most spectacular economic collapse since the 1930s. We are in the middle of a disaster scene today, aided and abetted by a political culture that has rushed to give corporations billions in bailouts but has largely hung ordinary people out to dry. Food pantries are running empty as farmers—themselves facing bankruptcy—plow their crops into the soil. Last week, the number of people who filed for unemployment benefits surged to more than twenty-six million. Poorer families and school districts don’t have the resources for online learning, meaning that we are leaving millions of kids behind. Rent strikes are popping up from coast to coast.

      We must build for a better future, not just climb out of the rubble of this pandemic, brush ourselves off, and start up in the same place we found ourselves in January 2020.
      With a disruption looming that may be as severe as the Great Depression, our ambition to confront it should be at the same scale. But our answer to these twin immediate crises must connect to a broader politics and vision that addresses the deep structural roots of the problems we face in America. We must build for a better future, not just climb out of the rubble of this pandemic, brush ourselves off, and start up in the same place we found ourselves in January 2020. In our earlier pieces in these pages, we’ve argued for a new politics of care, one organized around a commitment to universal provision for human needs; countervailing power for workers, people of color, and the vulnerable; and a rejection of carceral approaches to social problems. The question now is how to connect that vision to programmatic responses that address the needs of the moment and beyond. We need to aim at “non-reformist reforms”—reforms that embody a vision of the different world we want, and that work from a theory of power-building that recognizes that real change requires changing who has a say in our political process.

      Here’s one such reform: a massive new jobs program. Call it the Community Health Corps. Funded federally and organized locally, it would put millions of Americans to work caring for one another, and with far more sweeping goals than just turning around the sky-rocketing unemployment figures we see today. It would serve our needs for a vast force that can track and trace the virus, but add to it workers who can support those in need, all while securing our health and building real solidarity among us. Such a program, operating all around the country, in rural and urban areas alike, could help us get through this pandemic and mitigate the cataclysmic employment dislocation of the coming months and years.

      In truth, this is just a new form of an old idea—a Works Progress Administration (WPA) for an age of pandemics. But the aim is larger, to bring us through the crisis by calling into being government as we wish it to be—caring for us, bringing us together, while also enabling us to live our different lives. It would go beyond providing care to communities by stitching back together the personal connections among us torn asunder by our self-enforced isolation and by building power together, as workers and patients are tied to each other through the act of caregiving. It wouldn’t just create jobs to fill a hole during the crisis—it would develop skills and foster solidarity that will form the basis of the post-crisis economy, too.

      What jobs are needed? Start with contact tracing. The need here is straightforward and urgent. We cannot shelter in place forever, but reopening without measures to track the virus and sequester those exposed runs the same risk of swamping the health care system—infections and deaths will just come roaring back. Beyond the medical tragedy, such an outcome would also make a mockery of the sacrifices that millions of Americans have made over the past few months. That’s why every serious plan for reopening requires a massive scale-up in testing and contact tracing. The better we are at catching cases, notifying contacts, and supporting people who are sick or sequestered, the better control we will have over the virus, and the more “normal” life can be for those unexposed.

      Think of the people hired for contact tracing as virus detectives, who also have the under-appreciated skill of being able to talk to others with ease and empathy. They will engage people infected with SARS-CoV-2 (the virus that causes COVID19) in a process of recalling everyone they’ve seen and everywhere they’ve been for days, while recording all this information in detail. They will then reach out to these contacts, advise them on testing and quarantine, refer them for testing, and link them to necessary resources to help them through their quarantine, from start to finish. Despite all the talk about technological shortcuts, this old-fashioned shoe-leather epidemiology is going to be the mainstay of our next phase of attempts to control COVID-19. Contact tracing in its most basic form has been around since the smallpox outbreaks in Leicester in the United Kingdom in the 1870s. We know how to do it, and it can be scaled up locally.

      Shoring up the foundations of U.S. health care by valuing care itself isn’t just the first step towards a more rapid, effective response to health threats in the future. It will also move us toward a new politics of care, that starts from the ground up.
      Technology can help supplement these human tasks but cannot replace them. The idea that apps alone will solve the problem of contact tracing is the product of the technological “solutionism” that writers such as Evgeny Morozov have rightly argued is endemic to our culture today: the notion that no matter the problem, an app can efficiently solve it. Why won’t apps be a silver bullet? For one thing, they raise serious privacy issues, especially if they are not voluntary. There are technical issues too. It will be difficult for some technologies, like those that rely on GPS, to distinguish true contacts from false ones in crowded, dense urban environments. The myriad apps under development now have not been beta-tested, let alone rolled out in the midst of a pandemic at such a scale. It also isn’t clear that app developers have spent time talking to the potential end-users of their products, building their tools to meet the needs of, and benefit from the expertise and experience of, local health departments. Finally, technological solutions almost always leave out many of those who lack full participation rights in a digitally enabled society. For example, in the rush to move our financial transactions online and replace paper money with electronic payments with apps from banks and start-ups such as Venmo, we’ve left out many from low-income communities, particularly from communities of color. Apps can help make contact tracing more effective, but we need to act now, hiring people to do this work that no app can do.

      Spend a moment imagining a day in the life of a contact tracer working in Queens or Sioux Falls and you quickly see why an app alone cannot address the rippling crises that SARS-CoV-2 unleashes in every family. You also see the insufficiency, even, of contact tracing alone. Imagine you reach out to your first contact, who has tested positive and been sent home because they do not require hospitalization. Someone who has just learned that they have been exposed will have a myriad of important questions and needs. A father may wonder how, if he cannot leave his room, he will get food to his kids who are home from school. A shift worker who is wrongly fired for being sick will need help accessing unemployment insurance and legal support. A daughter may need help finding someone to provide essential daily care for a mother with dementia. Someone living alone will need help to walk the dog. We will need another group of workers to help them navigate these kinds of problems, which will require a mix of social work, advocacy, and even perhaps basic legal skills or the ability to make referrals to those who have them real-time.

      Those going out to trace contacts are going to find more than just SARS-CoV-2 in the places they visit. There will be some homes they call where no one has been exposed to the virus, but where families are struggling to make ends meet, having trouble with their landlords or their utility companies, or struggling with lost or unhelpful health insurance. Recent data has shown that during this pandemic domestic abuse has become “more frequent, more severe and more dangerous” and that mental health and substance issue on the rise. We can’t just walk away from these people, our neighbors in crisis. In the narrowest sense, ignoring these needs will make it harder for people to keep social distancing. In a broader sense, if we use our politics at a time of existential need to impose an unlivable life on our fellow citizens—if we fail those for whom staying at home might be more dangerous than the virus—we will tear away at the fabric of solidarity and trust that we need to maintain the shared project that is democracy.

      Right now we’re leaving help with all of this largely to individuals, families, and voluntary support. Most of us know people who are cutting corners with social distancing because they just can’t meet their daily needs any other way. In the next phase of the pandemic, we will need a much more precise and effective system of sequestering people if we are to get and keep the virus under control. While the mutual aid networks springing up around the country can handle a few requests for support, as we scale-up testing, the need of these kinds of social services and economic aid will explode. This can’t be handled simply as a matter of volunteerism even if “conservatives dream of returning to a world where private charity fulfilled all public needs.”

      section separator

      What is the alternative to genuine, public support for those who must remain isolated? Doctors Jim Yong Kim and Harvey Fineberg made the case in the New York Times recently that the ill, and their exposed families, should all be moved to facilities where they could be isolated from each other and the wider community, but they didn’t provide much guidance on how to do this humanely. Nor did they grapple with what it might mean to propose this sort of measure in a country with our history of state violence, especially as visited on families of color, who are vastly overrepresented among the sick today. We need to protect families from their sick loved ones, but forcibly warehousing families or the mildly symptomatic is not the way. We need a politics of support and care, not separation and deprivation. It’s clear that following public health advice isn’t as easy as it sounds—and its costs do not fall evenly. So we need support people to undertake this act of solidarity.

      We need a politics of support and care, not separation and deprivation.
      Alongside the test-and-trace brigade, then, we need other brigades too. We need a cadre of social workers who can provide specific help to individuals infected and affected by COVID, to enable them to follow public health and medical advice. We need a vastly scaled up testing workforce. Some will be dressed up in personal protective equipment (PPE), working at drive-through testing sites, visiting apartment buildings and nursing homes, and stationed outside of grocery stores and other businesses that remained highly trafficked even in the midst of the pandemic.

      Others will be working in labs or transporting samples, helping to process the millions of tests we will need each week, possibly each day. If evidence mounts that early intervention and close monitoring is essential to saving lives, we will also need a new brigade of health workers who can make virtual or home visits. We can additionally train local workers to help us gather evidence—for example mapping local health and services needs through surveys, building on successful models of community-based research, and working to better guide local programs. These programs will not only help us understand and respond to the spread of the virus but help us build better health programs when it recedes.

      We also need to address the explosion of infections in the workplace. We’ve seen outbreaks, large and small in meat processing plants across the country, in Amazon warehouses and Walmarts, leading to walkouts and lawsuits. As more and more businesses re-open, employees and employers need help to keep themselves and their customers and clients safe. Areas for employees and customers must be re-configured to maximize social distancing, and new workplace protocols need to be developed. Employers should be held responsible for taking the steps needed to protect their workers and the public, and some of this will likely not come without a stronger role for labor—via labor-management commissions, for example. An infection control brigade could work in cooperation with employees and employers, advising them on best practices in infection control, and assuring that supplies of PPE, from masks to gloves to physical barriers like plexiglass shields for cashiers are available. They can also ensure that early signs of failures in infection control are discovered and addressed immediately.

      We are already seeing small steps in this direction. In Massachusetts, Partners in Health (PiH), which has experience building community health workforces in places hit by disease and disaster around the globe, has been asked by the state to spearhead their new contact-tracing program. In a matter of weeks, they have hired and trained close to a thousand people for these important and complex jobs. Aware of the importance of the work and the demands of the job, PiH is paying them the same rate as U.S. Census takers, $27 an hour, providing them with health insurance and making an emphasis on hiring the unemployed and building a diverse workforce. About 17,000 people have applied for these jobs, showing that there is clearly a deep pool of people willing and able to do this work. That should come as no surprise, given the staggering loss of work in recent weeks and the inadequacy of the current government supports, and the outpouring of support we’ve seen in communities and mutual aid networks. People want to help. We just need to organize them.

      The problem is, while these efforts are admirable, state-level programs are vastly underpowered and underfunded. Before the crisis public health departments employed fewer than 2,000 contact tracers in the country. The best estimate we have projects that we will need to hire as many as 300,000 of them to address this outbreak. We have cohorts to build on for caseworkers and legal support too. One such pool derives from so-called community health worker (CHW) programs, which have a long history both in the United States and around the world.

      The United States is sicker now with COVID-19, but we’ve been sick for long while in many other ways.
      Today, we have about 120,000 community health care workers in cities and towns around the country doing health education and prevention work, collecting data, making links between local residents and the services they need. They are most often from the communities they serve and which have been underserved historically by the patchwork of a health system we have in the United States. In the context of need for testing-tracing-isolating in the age of COVID-19, local CHWs will go a long way to establishing trust and comfort in these troubling times. Contact tracers too should be recruited from local communities. Having a neighbor show up at your door (or on your screen) asking about your health and your personal contacts is more likely to be successful than a phone vibrating in your pocket telling you that you make have come in contact with someone with COVID-19.

      There are also models for the caseworker and legal support component in the medical-legal partnerships (MLPs) that have emerged all around the U.S. in recent years. Driven by the recognition that illness—and healthcare costs—are shaped by factors that doctors alone cannot control (like access to safe housing and benefits), hospitals and non-profits around the country have hired legal professionals to assist clinicians, social workers, and case managers address larger structural issues affecting patients’ health and well-being. As of early 2019 there were MLPs active in about 330 hospitals and health centers in 46 states with evidence that MLPs can improve patient health outcomes and well-being, improve mental health, remove barriers to health care for low-income families, increase access to stable housing and other social support.

      The idea is to build on these successes, which operate in small and disjointed ways, by integrating them into a federally funded Works Progress Administration for the age of COVID-19 and its aftermath. It will require significant federal funding, especially as states are forced into austerity by plummeting tax revenues and balanced budget requirements. But the cost will be small compared to the recent $2 trillion stimulus. Reports show that we can scale up contact tracing for just a few billion dollars—a fraction of the bailout we’ve handed over to big businesses. Some in Congress have already seen the need, and a federal bill awaiting the president’s signature provides some funds that could go towards such jobs, along with the massive scale up in testing that we need—though not nearly enough. Even a vastly larger program, hiring five million Americans for the duration of the crisis, would still cost less than the corporate bailout. This is a deal, if we consider what it can do to help not only save lives but also help employ people and buffer us against economic depression.

      We could also mold the program to help shore up the present and future of those who are at grave risk, but not of dying from COVID-19. Many young people today are facing down a terrifying future. With more than twenty-six million unemployed and more to come, who will hire someone just out of high school? How will students get that first job to pay off their college loans?

      By whatever accident of grace, young people are least at risk of developing serious complications of COVID-19, making them an obvious priority for a jobs program. The staggering health disparities of the pandemic make another priority clear. We need care workers who are from, and trusted, in local communities, both to reach those most in need, and to help build resources and power in those same communities. We also should demand a program that can hire those who are hardest hit by this downturn, and who we’ve cast aside for too long.

      This means not focusing only on workers who are already highly skilled (much less volunteers, who will always skew toward those who need not worry about their daily bread). Some of these new recruits will need significant training, but we should not think of that as a problem—these are the same jobs we will need after COVID-19, and we have chronic shortages of exactly those skills nursing care and home health care workers that we will need to address this pandemic. And many of these jobs will use skills that come far more quickly: contact tracers can be trained in days, as can those who they will deliver food, masks, and hand sanitizer to families.

      We know from the work of those who study the impact of jobs guarantees—including programs that have been running for many years in other countries—that such programs can be scaled up quickly, and provide essential counter-cyclical stability, as well discipline the private labor market. Especially now, creating alternatives to exploitative jobs is urgent, the only right thing to do. Many “essential jobs”—in janitorial positions, as cashiers in grocery stores, delivery workers—look a lot like forced labor today. With few exceptions, if you quit, you aren’t eligible for unemployment, and other forms of support like those elusive $1200 checks are too small, and not available to many. A Community Health Corps could provide better jobs, driving up the pay of those workers that we call essential, but do not pay that way. If these Corps jobs stick around (folded in, perhaps, to a Medicare for All program), they can help not only address our needs for care, but also our needs for decent work—and our needs to benefit collectively from the talents of so many who are now relegated to the margins, locked up or tossed away. We can also build the Corps as a springboard for further training, where those who have served their country can be funneled into higher education, in a new GI Bill for the age of COVID-19.

      Getting back to normal was never going to be a solace for many in our country. Business as usual is precisely what has made us all more vulnerable to disasters like the one we are currently experiencing.
      The United States may have the most technologically advanced health care system in the world, but we’re leading the number of worldwide coronavirus cases because we’ve badly trailed other industrialized nations in health outcomes for years. Many of the hardest-hit communities in the COVID-19 pandemic have been reeling from long-term health crises, from the opioid epidemic and deaths of despair in Appalachia to the burden of maternal deaths and the ongoing HIV epidemic in the South, to an explosion of obesity across the country with its downstream effects: type 2 diabetes, hyperlipidemia, high blood pressure, cardiovascular disease, and cancer. The United States is sicker now with COVID-19, but we’ve been sick for long while in many other ways.

      Beyond helping to manage the current crisis, then, a Community Health Corps would help to improve the health of people historically left out of the circle of care. For too long we’ve focused at the top, spending on expensive, technologically advanced specialty care, while neglecting primary and community care and underpaying caregivers themselves. Even in the midst of the pandemic, community health centers, which should be the core of our health approach, have teetered on financial ruin. Meanwhile, the domestic workers and home health aides who perform the essential act of care have been underpaid and left out of federal labor protections. Not to mention that much of the work of caring is still done at home, falling disproportionately on women and people of color.

      Shoring up the foundations of U.S. health care by valuing care itself isn’t just the first step towards a more rapid, effective response to health threats in the future. It will also move us toward a new politics of care, that starts from the ground up, in the places, we live, work and socialize. A politics that builds power among the caregivers, as the act of caring becomes publicly recognized and compensated for the productive work it is. Done right—and without the racialized and gendered exclusions that characterized the WPA—these new jobs can be a source of power for those who have never been fully allowed a voice in our democracy.

      section separator

      To scale this up quickly, we will need to bring together organizations like Partners in Health, who are experienced at mobilizing in a community though largely in the global South, and who are trusted and effective in their work on health, with local organizations, working on civil, social and economic rights such as national groups like the Center for Popular Democracy and Community Change, and their diverse roster of local community organizations.

      Will it be easy to get our creaking, divided democracy to funnel resources into these programs? Probably not. But COVID-19 is conspiring to show us, all at once and in a way that no one can ignore, how central care is to a healthy society.
      Over the past four decades we’ve seen the erosion of government as a force for good in people’s lives, most often by design as conservatives have looked to shrink the state, weaken its effectiveness, and privatize its functions. Liberals have gone along and lost their faith in the kind of government that built their political base while helping millions in the modern era, starting with the New Deal, and the civil rights, social and economic programs that were the hallmark of the Great Society period in the 1960s. The U.S. state is so weak and untrusted right now that banks have had to take over as the vehicle for the provision of many of the billions just released under the emergency appropriations by Congress, as many Americans cursed the IRS because of delays in the small checks they were promised in COVID-19 relief.

      A Community Health Corps could be part of the remedy—in terms of the direct services and employment it could offer millions of Americans, in the ways in which this effort could lift up the health and well-being of so many, and also in terms of renewing faith in the power of government to help. The Corps would also be a prophylaxis—a first line in the response to the next challenges we face, whether it’s a seasonal return of COVID-19 or another pandemic, or the monumental troubles that climate change will rain down on our communities.

      It would also serve as a model, a test of one essential component of a Green New Deal: the creation of millions of good green jobs. Green jobs, after all, are not just in construction, and many directly benefit health. That is why the most compelling versions of such a proposal prioritize new care work jobs, as well as jobs restoring our trails and parks, and even making a place for the artists and writers whose work is some of the greatest legacy of the WPA. Some of these jobs might even be initiated as part of the Community Health Corps. With so little traffic on the roads, there is no better time to build bike lanes—and green housing too, if the safety of workers can be assured. Greening our cities and improving housing for low-income communities are an essential component of a healthier society, as well as a healthier planet. Climate change is the largest foreseeable threat to our health; we can start to address this looming crisis right now, as we combat this pandemic.

      We need more than a jobs program at this moment of national crisis, to be sure. We also need more SARS-COV-2 tests, more basic income, and better data about the pandemic, to name just a few. But rising up from under the cruel weight of this pandemic, we should also aim for something lasting and better. Getting back to normal was never going to be a solace for many in our country. Business as usual is precisely what has made us all more vulnerable to disasters like the one we are currently experiencing.

      Will it be easy to get our creaking, divided democracy to funnel resources into these programs? Probably not. But COVID-19 is conspiring to show us, all at once and in a way that no one can ignore, how central care—writ large, broadly conceived—is to a healthy society. Rudolf Virchow, the father of social medicine, once said: “Medicine is a social science and politics is nothing else but medicine on a large scale. Medicine as a social science, as the science of human beings, has the obligation to point out problems and to attempt their theoretical solution; the politician . . . must find the means for their actual solution.” A Community Health Corps is one actual solution, one place to start to build a new movement that heals us and our body politic, and that will allow us—all of us—to survive a pandemic, and then, to thrive.


      #soin #santé #politique_du_soin #santé_communautaire #pandémie #recherche_des_contacts #emploi #agents_de_santé_communautaire #aptitude_à_parler #médecine_sociale vs #solutionnisme_technologique #green_new_deal

  • (17) Pour faire la guerre au virus, armons numériquement les enquêteurs sanitaires - Libération

    Pourquoi se focaliser sur une application qu’il faudra discuter à l’Assemblée nationale et qui risque de ne jamais voir le jour, alors que nous devrions déjà nous concentrer sur la constitution et l’outillage numérique d’une véritable armée d’enquêteurs en épidémiologie ?

    Tribune. Le débat sur l’apport du numérique à la résolution de la crise sanitaire actuelle est bien mal engagé. Une énergie considérable est consacrée à développer et affiner le controversé projet d’application de traçage StopCovid, dont la faisabilité et l’utilité restent pourtant sujettes à caution. Pendant qu’on en développe plusieurs versions, qu’on engage un bras de fer avec Google et Apple et qu’on mobilise les parlementaires à discuter et voter sur une application qui risque de ne jamais voir le jour, rien n’est dit de la préparation de l’équipement numérique de la « force sanitaire », cette « armée » d’enquêteurs de terrain qu’il va falloir déployer à partir du déconfinement pour remonter et casser les chaînes de transmission.

    Pourtant, pour l’anthropologue et médecin Paul Farmer, ex-envoyé spécial des Nations unies à Haïti après le séisme en 2009, la réponse à une épidémie, c’est avant tout « staff and stuff » : des gens et des choses.

    Ces enquêtes, minutieuses, fastidieuses, peuvent mobiliser des effectifs importants. 9 000 enquêteurs à Wuhan (pour une ville de 11 millions d’habitants), 20 000 en Corée (qui en compte 52 millions). Les experts en santé publique de l’université Johns-Hopkins estiment ainsi à 100 000 le nombre d’enquêteurs qu’il faudrait aligner aux Etats-Unis pour un coût d’environ 3,6 milliards de dollars. Selon Tom Frieden, un ancien directeur du CDC, ce chiffre pourrait être trois fois plus élevé. En Allemagne, le gouvernement fédéral envisage de recruter 20 000 agents (une équipe de cinq personnes pour 20 000 habitants). En Belgique, Wallonie, Bruxelles et Flandre comptent recruter 2 000 enquêteurs.

    En France, selon le président du Conseil scientifique, Jean-François Delfraissy, ce sont entre 10 000 et 15 000 nouvelles contaminations par jour qui pourraient être enregistrées à partir de la mi-mai ou de la fin mai. Pour prendre en charge les nouveaux contaminés et tracer les cas contacts, il évalue à 20 000 ou 30 000 personnes la taille de cette « force sanitaire » : « une armada » pour reprendre ses mots. Il alerte : « Si on n’a pas cette brigade, une app numérique ne marchera pas. »

    Ce qui ne veut pas dire que le numérique ne peut pas les aider. Car ces enquêteurs, il va falloir les outiller. En tutoriels. En questionnaires pour accompagner et guider la personne atteinte dans recherche de contacts (« Quand avez-vous déjeuné ? Où étiez-vous alors ? »). Il faudra déployer une plateforme pour remonter et centraliser les données, dans le respect de la vie privée de l’ensemble des personnes concernées, malades, proches, enquêteurs eux-mêmes ! Et des outils pour interagir rapidement avec les épidémiologistes plus chevronnés et avec la chaîne aval (masques, tests et isolement volontaire). On peut aussi imaginer des bases de connaissance pour aider les enquêteurs à apporter des réponses précises et documentées aux personnes qu’elles seront amenées à interroger. Il faut aussi envisager le cas de personnes atteintes qui ne parleraient pas le français, et prévoir la possibilité de faire appel à distance à des interprètes.
    Délai record

    C’est sur l’outillage numérique des enquêteurs et sur le recueil des données les plus utiles que devraient se concentrer les efforts et le débat, pour que la France se dote du seul moyen reconnu à ce jour comme efficace pour lutter contre la pandémie.

    Notre fascination collective pour les applications de traçage révèle notre désir d’en finir « magiquement », grâce à une killer app qui résoudrait tout, avec la maladie.

    Didier Sicard Président d’honneur du Comité National d’Ethique ,
    Benoit Thieulin ancien président du Conseil national du numérique ,
    Maurice Ronai ancien commissaire à la Cnil ,
    Godefroy Beauvallet ancien vice-président du Conseil national du numérique

    #After #Epidémiologie

  • Coronavirus : la mortalité en réanimation beaucoup plus forte qu’annoncée en France

    Quel est le taux de mortalité des patients malades du Covid-19 en réanimation ? Selon le ministère de la santé, il serait de 10 %. Ce chiffre a été annoncé par Jérôme Salomon, le directeur général de la santé, lors de saconférence de presse du 17 avril. Selon les informations du Monde, il est largement sous-estimé. Aujourd’hui, il serait en effet de l’ordre de 30 % à 40 %, soit trois à quatre fois plus important que le bilan officiel communiqué par le gouvernement. Dimanche 26 avril, 4 682 patients Covid étaient en réanimation.

  • « StopCovid est un projet désastreux piloté par des apprentis sorciers »

    Antonio Casilli, Sociologue
    Paul-Olivier Dehaye, Mathématicien
    Jean-Baptiste Soufron, Avocat

    Il faut renoncer à la mise en place d’un outil de surveillance enregistrant toutes nos interactions humaines et sur lequel pèse l’ombre d’intérêts privés et politiques, à l’instar du scandale Cambridge Analytica, plaide un collectif de trois spécialistes du numérique dans une tribune au « Monde ».


    • L’affaire Cambridge Analytica, révélée au grand jour en 2018, avait comme point de départ les travaux de chercheurs de l’université anglaise. Une application appelée « Thisisyourdigitallife », présentée comme un simple quiz psychologique, avait d’abord été proposée à des utilisateurs de la plate-forme de microtravail Amazon Mechanical Turk. Ensuite, ces derniers avaient été amenés à donner accès au profil Facebook de tous leurs contacts. C’était, en quelque sorte, du traçage numérique des contacts avant la lettre.

      A aucun moment ces sujets n’avaient consenti à la réutilisation de leurs informations dans la campagne du Brexit, dans celle de Donald Trump, ou dans des élections en Inde et en Argentine. Cela est arrivé ensuite, lorsque les chercheurs ont voulu monétiser les données, initialement collectées dans un but théoriquement désintéressé, par le biais de l’entreprise Cambridge Analytica. En principe, cette démarche respectait les lois des différents pays et les règles de ces grandes plates-formes. Néanmoins, de puissants algorithmes ont été mis au service des intérêts personnels et de la soif de pouvoir d’hommes politiques sans scrupule.

    • Tribune. Le mardi 28 avril, les parlementaires français seront amenés à voter sur StopCovid, l’application mobile de traçage des individus imposée par l’exécutif. Nous souhaitons que, par leur vote, ils convainquent ce dernier de renoncer à cette idée tant qu’il est encore temps. Non pas de l’améliorer, mais d’y renoncer tout court. En fait, même si toutes les garanties légales et techniques étaient mises en place (anonymisation des données, open source, technologies Bluetooth, consentement des utilisateurs, protocole décentralisé, etc.), StopCovid serait exposée au plus grand des dangers : celui de se transformer sous peu en « StopCovid Analytica », une nouvelle version du scandale Cambridge Analytica [siphonnage des données privées de dizaines de millions de comptes Facebook].

      L’application StopCovid a été imaginée comme un outil pour permettre de sortir la population française de la situation de restriction des libertés publiques provoquée par le Covid-19. En réalité, cette « solution » technologique ne serait qu’une continuation du confinement par d’autres moyens. Si, avec ce dernier, nous avons fait l’expérience d’une assignation à résidence collective, les applications mobiles de surveillance risquent de banaliser le port du bracelet électronique.

      Tous les citoyens, malades ou non

      Le terme n’est pas exagéré : c’est déjà le cas à Hong-Kong, qui impose un capteur au poignet des personnes en quarantaine, et c’est l’objet de tests en Corée du Sud et au Liechtenstein pour certaines catégories de citoyens à risque. StopCovid, elle, a vocation à être installée dans les smartphones, mais elle concerne tous les citoyens, malades ou non. Malgré le fait que son installation soit présentée comme facultative dans d’autres pays, tels l’Italie, on assiste à la transformation de cette démarche volontaire en obligation.

      L’affaire Cambridge Analytica, révélée au grand jour en 2018, avait comme point de départ les travaux de chercheurs de l’université anglaise. Une application appelée « Thisisyourdigitallife », présentée comme un simple quiz psychologique, avait d’abord été proposée à des utilisateurs de la plate-forme de microtravail Amazon Mechanical Turk. Ensuite, ces derniers avaient été amenés à donner accès au profil Facebook de tous leurs contacts. C’était, en quelque sorte, du traçage numérique des contacts avant la lettre.

      A aucun moment ces sujets n’avaient consenti à la réutilisation de leurs informations dans le cadre de la campagne du Brexit ou dans l’élection présidentielle de Donald Trump. Cela est arrivé ensuite, lorsque les chercheurs ont voulu monétiser les données, initialement collectées dans un but théoriquement désintéressé, par le biais de l’entreprise Cambridge Analytica. En principe, cette démarche respectait les lois des différents pays et les règles de ces grandes plates-formes. Néanmoins, de puissants algorithmes ont été mis au service des intérêts personnels et de la soif de pouvoir d’hommes politiques sans scrupule.

      Les mêmes ingrédients sont réunis ici : des scientifiques « de bonne volonté », des géants de la « tech », des intérêts politiques. Dans le cas de StopCovid, c’est le consortium universitaire européen Pan-European Privacy Preserving Proximity Tracing (PEPP-PT), qui a vu le jour à la suite de la pandémie. Ces scientifiques se sont attelés à la tâche de concevoir dans l’urgence le capteur de contacts le plus puissant, dans le respect des lois. Cela s’articule avec les intérêts économiques d’acteurs privés, tels les grands groupes industriels nationaux, le secteur automobile et les banques en Italie, les télécoms et les professionnels de l’hébergement informatique en France. Mais surtout les GAFA, les géants américains du numérique, se sont emparés du sujet.

      Cette fois, ce ne sont pas Facebook et Amazon, mais Google et Apple, qui ont tout de suite proposé de fournir une nouvelle structure pour diffuser les applications de suivi de contacts sur leurs plates-formes. La menace qui plane au-delà de tous ces acteurs vient des ambitions de certains milieux politiques européens d’afficher leur détermination dans la lutte contre le Covid19, en se targuant d’une solution technique à grande échelle, utilisant les données personnelles pour la « campagne du déconfinement ».

      Une myopie sur les dimensions sociales des données

      Le projet StopCovid n’offre aucune garantie sur les finalités exactes de la collecte de ces données. L’exécutif français ne s’autorise pas à réfléchir à la phase qui suit la collecte, c’est-à-dire au traitement qui sera fait de ces informations sensibles. Quels algorithmes les analyseront ? Avec quelles autres données seront-elles croisées sur le moyen et le court terme ? Son court-termisme s’accompagne d’une myopie sur les dimensions sociales des données.

      Que se passerait-il si, comme plusieurs scientifiques de l’Inria, du CNRS et d’Informatics Europe s’époumonent à nous le dire, malgré une collecte initiale de données réduite au minimum, des entreprises ou des puissances étrangères décidaient de créer des « applications parasites » qui, comme Cambridge Analytica, croiseraient les données anonymisées de StopCovid avec d’autres bases de données nominatives ? Que se passerait-il, par exemple, si une plate-forme de livraison à domicile décidait (cela s’est passé récemment en Chine) de donner des informations en temps réel sur la santé de ses coursiers ? Comment pourrait-on empêcher un employeur ou un donneur d’ordres de profiter dans le futur des données sur l’état de santé et les habitudes sociales des travailleurs ?

      L’affaire Cambridge Analytica nous a permis de comprendre que les jeux de pouvoir violents et partisans autour de la maîtrise de nos données personnelles ont des conséquences directes sur l’ensemble de la vie réelle. Il ne s’agit pas d’une lubie abstraite. Le cas de StopCovid est tout aussi marquant. En focalisant des ressources, l’attention du public et celle des parlementaires sur une solution technique probablement inefficace, le gouvernement nous détourne des urgences les plus criantes : la pénurie de masques, de tests et de médicaments, ou les inégalités d’exposition au risque d’infection.

      Une malheureuse diversion

      Cette malheureuse diversion n’aurait pas lieu si le gouvernement n’imposait pas ses stratégies numériques, verticalement, n’étant plus guidé que par l’urgence de faire semblant d’agir. Face à ces enjeux, il faudrait au contraire impliquer activement et à parts égales les citoyens, les institutions, les organisations et les territoires pour repenser notre rapport à la technologie. Le modèle de gouvernance qui accompagnera StopCovid sera manifestement centré dans les mains d’une poignée d’acteurs étatiques et marchands. Une telle verticalité n’offre aucune garantie contre l’évolution rapide de l’application en un outil coercitif, imposé à tout le monde.

      Ce dispositif entraînerait un recul fondamental en matière de libertés, à la fois symbolique et concret : tant sur la liberté de déplacement, notamment entre les pays qui refuseraient d’avoir des systèmes de traçage ou qui prendront ce prétexte pour renforcer leur forteresse, que sur la liberté de travailler, de se réunir ou sur la vie privée. Les pouvoirs publics, les entreprises et les chercheurs qui dans le courant des dernières semaines sont allés de l’avant avec cette proposition désastreuse, ressemblent à des apprentis sorciers qui manient des outils dont la puissance destructrice leur échappe. Et, comme dans le poème de Goethe, quand l’apprenti sorcier n’arrive plus à retenir les forces qu’il a invoquées, il finit par implorer une figure d’autorité, une puissance supérieure qui remette de l’ordre. Sauf que, comme le poète nous l’apprend, ce « maître habile » ne reprend ces outils « que pour les faire servir à ses desseins ».

      Antonio Casilli, sociologue.
      Paul-Olivier Dehaye, mathématicien.
      Jean-Baptiste Soufron, avocat.
      Cosignataires : Sophie Binet et Marie-José Kotlicki cosecrétaires généraux de l’UGICT-CGT ; Raquel Radaut, membre de La Quadrature du Net.

      #StopCovid #tracing #solutionnisme_technologique #Big_data #surveillance

  • Google, Not the Government, Is Building the Future

    One persistent criticism of Silicon Valley is that it no longer works on big, world-changing ideas. Every few months, a dumb start-up will make the news — most recently the one selling a $700 juicer — and folks outside the tech industry will begin singing I-told-you-sos. But don’t be fooled by expensive juice. The idea that Silicon Valley no longer funds big things isn’t just wrong, but also obtuse and fairly dangerous. Look at the cars, the rockets, the internet-beaming balloons and gliders, (...)

    #Google #algorithme #solutionnisme_technologique #domination

  • The Right to Inconvenience

    A recent Washington Post article discussed at length changes coming to Apple’s new operating systems. Described as user-friendly and integrated, Apple’s latest operating system will, among other things, bring Siri voice commands to the desktop, organise photos using facial recognition, suggest relevant emojis in text conversations, and generally underpin the user experience with artificial intelligence. Sold as a seamless and integrated solution to “app fatigue”, Apple aims to give consumers (...)

    #Apple #Siri #reconnaissance_faciale #Privacy_International #solutionnisme_technologique

  • Internet, le réseau des plus forts

    Auteurs de discours ambitieux sur une planète 100% connectée où ils seraient capables de se substituer aux Etats dans tous les domaines, les mastodontes comme Facebook, Alphabet, Microsoft ou Apple ont en plus les moyens de leur mise en œuvre. Pas de quoi rassurer les défenseurs d’un modèle social et solidaire.

    #Alphabet #Apple #Microsoft #Facebook #domination #GAFA #solutionnisme_technologique #bénéfices

  • Evgeny Morozov : « Ce sera l’avènement d’un modèle néolibéral très individualisé »

    Evgeny Morozov, spécialiste de l’impact des nouvelles technologies, affirme que les géants d’Internet continuent à craindre les gouvernements. Mais qu’aucune alternative n’est possible sans repenser le capitalisme. Chercheur et journaliste d’origine biélorusse, Evgeny Morozov s’est imposé en quelques années comme l’un des plus féroces détracteurs du discours des géants du Net. Après avoir déconstruit le « solutionnisme technologique », il a publié en 2015 le Mirage numérique (les Prairies ordinaires), où il (...)

    #GAFA #domination #solutionnisme_technologique

  • Internet : la grande illusion ?

    L’avènement des technologies numériques augure-t-il une société du partage des informations et des services ? Laure Wagner, co-fondatrice de BlaBlaCar, croit à une évolution participative qui rendrait les citoyens responsables. Evgeny Morozov, auteur du « Mirage numérique : pour une politique des bigs datas », n’y croit pas et appelle à « une critique émancipatrice ». #GAFA #solutionnisme_technologique #abus_de_position_dominante