The Invisibility of Viruses and Data

von berries
7 min readApr 25, 2020

--

Sociometrics and AI hinterland by Christian von Borries

Still from the film “AI is the Answer — What was the Question”. The director in front of the Palantir pop-up building, Davos, World Economic Forum 2020.

Sometimes we now hear that we are all in the same boat, which is not true.

“Equal treatment” only consists in the fact that this virus takes neither social status nor income into account, not to mention national borders. Inequality is increasing between countries that have the resources and those that do not, between those who have access to health and care services and those who do not.

Migrants of this world who suffer the most are those who are not protected by any social or health system. The collateral damage in Africa and South America caused by the shutdown of the rich countries is accepted as well as the death marches of the poor in India. The virus thus becomes a pretext for internal and external xenophobia (which can be observed particularly in the US, Hungary and Poland).

I don’t want to talk about alarmist numbers, given the 0.0003% global infection rate that a US university publishes every minutes, appearing objective while clearly ideologically biased towards cold war and big business; nor about the fact that in Berlin and Brandenburg just every thousandth person is infected while 40% of the German intensive care beds are empty and staff on holidays, and not about the ratio of these numbers to the half million deaths of the Syrian war, or the countless dead on the bottom of the Mediterranean Sea.

I could try to explain why Cuba, China and Russia are helping the Italian healthcare system, while the European north decides to extend its austerity policies, that had been implemented in Greece, to Spain and Italy, heralding the end of Europe as we know it.

I could also report on how the agro-industry and its endless drive to grow have led to the expansion of monocultures into the farthest corners of the world, where hundreds of thousands of viruses are waiting to make the zoonotic jump. “If we don’t take care of nature, it will take care of us” warns Elisabeth Mrema, Acting Executive Secretary of the Convention on Biological Diversity at the UN.

So let’s not forget that humans, by analogy with their dealings with one another, have meanwhile mangled nature in such a way that even cyberneticists of the Californian Ideology run out of arguments, but not cash. More on that in a moment.

I tried to understand the crisis more broadly by reading Elias Canetti’s analysis of fascism “Crowds und Power”. It says that “people fear nothing more than being touched by the unknown. … keeping a distance becomes the last hope. … It is strange how the hope for survival individualizes the people.”

In view of this situation, I would like to speak about the technological equalization of the individual in late capitalism, which seems to be a reason for the measures to which we are now exposed worldwide. Access to our lives takes place through a new form of data mining, in other words a surveillance of life, a previously unknown data colonialism.

Decolonizing data’s role in explaining the social world on the other hand mean rejecting the assumption that only data can explain data — that is, that only quantitative methods based on big data methodologies can accurately describe the world from now on. While it should not reject the idea of gathering data as such (for example, for clearly defined social purposes), the evolving project to think together about the consequences of data colonialism would have to insist on close attention to those “aspects of social life that remains unseen, unheard, uncounted or unacknowledged within prevailing understandings of capitalism”, as Johanna Montgomerie argues.

Around Easter, all German users of the Russian messaging app Telegram received a message from the federal government asking them to use a tracking app. As a mere reminder, Telegram has not been willing to work with states (especially not with the Russian state), such as WhatsApp with the US government.

Health apps for tracking with the choice of opting-in are one thing, and it sounds better than “location based social intelligence” or even “intent recognition”. Palantir, the mysterious data processing company from California, who, as Bloomberg reported, is now working with European countries. (Palantir Foundry: “removing the barriers between back-end data management and front-end data analytics… deep subject matter… use it to take action” ). Hiring this relatively unknown Silicon Valley company seems to be a clandestine and more effective approach. In addition to dubious methods of big data procurement, Palantir uses machine learning to generate complete population profiles. In other words, the Central European countries that can afford such service are now able to maintain a big data pool with little latency and for predictive policing in the streets and online.

Alex Pentland, head of MIT’s Connection Science Lab and author of “Social Physics“ and “Data for a new Enlightenment” is the mastermind of this reality mining approach:

“In a recent study of a large European city, for instance, we found that social bridges were three hundred percent better at predicting people’s behaviors than demographics, including age, gender, income, and education. …

The way to think about society is in terms of these behavior groups. Who do they associate with? What are those other people doing? The idea of social bridges is a far more powerful concept than demographics, because social bridges are the most powerful way that people influence each other. …

The “secret sauce” is the learning rule, called the credit assignment function, which determines how much each connection contributed to the overall answer. Once the contributions of each connection have been determined, then the learning algorithm that builds the AI is simple: connections which have contributed positively are strengthened; those that have contributed negatively are weakened. …

This same insight can be used to create a better human society, and, in fact, such techniques are already widely used in industry and sports. …

The key is to have a credit assignment function that both makes sense for each individual and yet at the same time yields global optimal performance. This would be a real version of the “invisible hand,” one that works for real people and not just for rational individuals.”

Pentland prefers “populations” to societies, “statistics” to meaning, and “computation” to law.

In order to understand and classify this, I have to take a few detours and ask questions:

  • Can’t we see the global embracing of social distancing as an indication for the meaningfulness of digital technologies?
  • Can the use of these AI-controlled apps be compared with arcade games (where dopamine is released from the first second onward) to test what is supposedly playful (nudging) and what is not reasonable for people?
  • Isn’t the South Korean health surveillance with contact tracing, data + security tracking, personal interviews, the evaluation of individual payment behavior and GPS a sensible method to isolate only really vulnerable people instead of an entire population?
  • Hasn’t the Chinese state shown its ability to control the pandemic by evaluating user data from the social media app WeChat, which covers all areas of life, without imposing a lockdown on the whole country? Or are we only witnessing the nationwide implementation of the Social Credit Score System, this time disguised as aid from the state?

In Korea, people are aware of a dark side of technology: “It is true that public interest tends to be emphasized more than human rights”.

  • So what happens if the governments of the rich West now declare public space to be a danger zone? If they run out of arguments for leveling a general suspicion at a potentially ill total population. This contagion has unforeseeable social, psychological, that is, societal consequences.

We are already hearing the call for austerity measures (for example regarding social pensions) to defuse public budgets and compensate deficits.

“This is really a civil war. The enemy is not out there but inside of us,” as Giorgio Agamben correctly observed.

The new reality is defined by fictionalizing reality.

This fictionalization follows the Aristotelian life-or-death paradigm. The new narration again corresponds to that of the Hollywood film (or political speech, or advertising). They simplify reality and make it seem without alternative. Binge-watching of TV series will kill streaming services and all of us because it makes us depressed when we turn it off. Everyone is noticing this now.

The digital world is not the alternative to analog reality. However, now this reality must be enforced top-down, and this is new. Only it shouldn’t sound that way.

It seems to be helpful that the invisibility of the virus becomes one with the invisibility of data.

Corona becomes a journey to ourselves, or to say it with (the American rapper) Childish Gambino: “You just a barcode”. We are witnessing an exercise in homogenizing behavior, because the technologies of artificial intelligence are based on generating statistical abnormalies and thus resemble the narrative of virology. Both argue in the dark. They extrapolate the future, which then becomes the new paradigm of the present. This is and should not be comprehensible, but nevertheless without alternative.

In Russia, the state’s approach is still called “digitized concentration camp”. Algorithms on your phone and in surveillance cameras determine what you do outside of your home, and the disciplinary state is standing by outside.

This is dealt with differently in the West: Corona is already internalized obedience, hygienic humanism, the control society.

The participation of the entire population is required. And denouncing is allowed again.

“Binding recommendation” is the official German language regulation to make the people obey.

“And then they’ll have us all on a short leash”, the kiosk owner in Berlin-Wedding summarizes our situation.

Berlin, April 2020

--

--