What happens when a digital epidemiologist and a sociologist of new technologies meet – via screens, of course – to talk about the pandemic? During the Applied Machine Learning Days 2021, a series of online conferences organised by the École polytechnique fédérale de Lausanne (EPFL), an interesting discussion arises between Prof Marcel Salathé and Prof Zeynep Tufekci.
Why this is interesting. Having first worked as a computer scientist, the Turkish sociologist Zeynep Tufekci has quickly become one of the leading voices on techno-social issues: from the study of revolutions on social media, the biases of algorithms to the political power of tech giants. In the spring of 2020, at the heart of the pandemic, she occupied a unique position, questioning the official statements of the authorities, notably on the wearing of masks and the airborne nature of the virus, to the point of receiving praise from the New York Times.
Gaining awareness. Thanks to her opinion pieces published in The Atlantic and The New York Times, the researcher has become, in the last few months, in the words of Marcel Salathé, "one of the most famous epidemiologists who is not an epidemiologist". It is by some strange coincidence that the sociologist – a specialist in information systems and then in Twitter revolutions – found herself in a situation where she was able to see the pandemic unfold, at times, with more foresight than others.
"In 2019, was closely studying the protests in Hong Kong at the time. But I was injured and couldn't travel. So in November, I had to follow the events from abroad. To understand the surveillance in Hong Kong, I had to look to China. And I started to see, via social networks, that something was happening in Wuhan. I first prepared myself with the idea of being able to return to Hong Kong."
Soon, her experience in reading the authorities' responses to the protests suggested to her that something was wrong. "I'm not a virologist or epidemiologist," concedes the sociologist. “But, just as you can see the waters recede before a tsunami, I quickly realised that something was going on."
"In some social structures, especially where there is censorship, there are incentives to lie when you don't know what is going on. I realised that the Chinese central government itself was surely in the dark. I was starting to put things together from the shadows we could see and to wonder: why had it been so difficult for the WHO to send a mission to Wuhan in early 2020?
The other tipping point was when the authorities closed the Wuhan animal market while claiming that there was no evidence that the virus was being transmitted from human to human.
The response of institutions. "In 2019, I had already started to talk to my students about the role institutions play in the face of disasters. These are mechanisms designed for a pandemic or for the climate crisis, for example," she continues. When the crisis broke in the spring of 2020, Zeynep Tufekci dared to openly criticise the WHO or the Centers for Disease Control and Prevention (CDC) in the United States.
“There were shortages of masks, and on top of that, some institutions were letting the message circulate that masks, if not worn properly, could be even more dangerous," she recalls. Of course, a few months later, when masks finally came into use, public confidence was already damaged. I waited – and hoped – for a specialised researcher, maybe from the inside of the CDC, to speak out”n the end, she did so herself in an opinion piece published in the New York Times. She continues:
"I never thought I would ever have to do this: I am on the side of science. Then I've been accused of all sorts of things. First, of being xenophobic when I warned of the risks posed by the virus before it came out of Asia, and then of being too alarmist... But no one, at the time, thought of teaching people how to prepare themselves!"
The current situation. Despite the progress of vaccination campaigns, several parts of the world are now facing a third wave, partly linked, at least in Europe, to the dominance of the famous, more contagious variant B.1.1.7. "Paradoxically, variants that are more deadly but not more contagious may be less of a problem", because they would not have the opportunity to become the majority, as B.1.1.7 has become. And Zeynep Tufekci wonders:
"How is it possible that one year after the start of the pandemic, people still don't understand how exponentials work?"
Humans versus knowledge and technology. The pandemic is also an "infodemic", where our brains are flooded with all kinds of information, some of it false. Are social networks, which are designed to amplify the most contrasting voices, responsible? Can we attach more ethics to algorithms? She reminds us that the problem is above all social and structural:
"Diversity is important. It is a problem in institutions in general. If the WHO had more specialists in the airborne transmission of pathogens, for example, perhaps the recommendation that hand washing was sufficient would not have been made.
Marcel Salathé asks: would things have been different 20 years ago? "We would have just sent specialists on television," she says. Twenty years ago, too, vaccines would never have been developed so quickly and contact tracing apps might never have been developed in a world that didn't know Facebook or Tinder.
The discussion ends with this digital theme with no time to touch on the question of balance between epidemic control and individual freedom. Marcel Salathé concludes:
"In Europe, there is a loss of trust in institutions that promises shaky years ahead."
At the end of February, the epidemiologist left the Swiss National Covid-19 task force to embark on a new project: CH++, which aims to encourage the administration and the population to develop their digital skills. The issue of trust promises to be crucial at all levels.