Every month brings another story about the abuse of data: how the Taliban gained access to a large amount of personal data on the people of Afghanistan gathered by the US military and the Afghan state; how Europol had unlawfully amassed a “vast store of personal data”; how the International Committee of the Red Cross had suffered a cyber-attack that exposed data from over half a million vulnerable people; or how the the entire Shanghai police database, with personal details of a billion Chinese citizens, was for sale for 10 Bitcoin.
Our data will not fall into the wrong hands, because it’s already in the wrong hands; whether that is influence operations facilitated by social media databases; mass surveillance by nominally liberal governments; the increasingly dysfunctional online advertising ecosystem on which much of the web is built; or simply the drip-drip-drip of stories about how social media is changing the way we think and act. We are continually assailed by such stories, yet somehow they simply slide down the sides of our non-stick attention spans.
In the past few years a new vocabulary has appeared that attempts to describe what is happening to us: algorithmic bias, surveillance capitalism, data colonialism, and so forth. These are all useful concepts, yet none seems to fully capture what is happening to us; even those of us who work in this digital space struggle to articulate the full range of our concerns.
The reason that we struggle is because what we are dealing with is a hyperobject, a term coined by the philosopher Timothy Morton to describe objects which are massively distributed in time and space relative to humans, making them almost impossible for us, as individuals, to grasp. As a result we only ever experience local manifestations of the hyperobject, not the hyperobject itself. Morton intended the concept to help us comprehend our experience of climate change – but it can be readily applied to other phenomena, and in this case our hyperobject is a data crisis.
Using the analogy of the climate crisis, these local manifestations such as those described at the start of this piece bear the same relation to the data crisis as a wildfire does to the climate crisis. Just as the climate crisis has forced its way into public consciousness through the slow but steady increase in extreme local events – the fires and floods and all of the above that have become increasingly difficult to ignore – the data crisis is also forcing its way into public consciousness.
How that crisis is experienced varies depending on the country. In China extensive surveillance raises concerns whether it is applied to control the Uighur population or the Covid-19 virus. In India the government’s digital identity framework that simultaneously claims more efficient management of public services and enables enhanced discrimination against political opponents. In the US large technology platforms promote a model in which we all become complicit in our own surveillance, while state security actors take advantage of cracks in that model to spring up like weeds.
If this sounds apocalyptic, it is. The hyperobject hovers at the edge of our perception like the monster in a horror movie waiting for the third act. When discussing the climate crisis, Morton asserted that
“The end of the world has already occurred. We can be uncannily precise about the date on which the world ended… It was April 1784, when James Watt patented the steam engine, an act that commenced the depositing of carbon in Earth’s crust – namely, the inception of humanity as a geophysical force on a planetary scale.”
The world has ended more than once, of course; and another date on which the world ended was January 1889, when Herman Hollerith patented his punch card technology. If the climate crisis is the result of burning fossil fuels in unsustainable ways, then the data crisis is the result of managing data in unsustainable ways. It should be clear that we cannot address the data crisis through tweaking the way in which we collect data, any more than we can address the climate crisis through tweaking the way we extract fossil fuels. This is a structural problem with the entire system we have constructed to manage data.
That system has already begun to transform civilisation, and bringing that process to a halt is unfeasible and probably undesirable, given the measurable benefits that technology brings us. The crisis is about much more than the power of platforms to manipulate data in ways which might help or harm us; it is about the long-term impacts of the digitalisation of every single aspect of everyday life, all the way down to our very identities, not just as consumers, not just as citizens, but as humans. We need to create new political technologies – institutions, processes and norms – that can transform this data regime in ways which accommodate the benefits of digital technologies while mitigating the costs of digital harm.
Regardless of what approaches are tried, all must be based on recognition that the crisis is already upon us. In the last few years acceptance that the climate crisis is already upon us has become mainstream, but acceptance of the data crisis is at a much earlier stage; the concept is too huge, too troubling, too alien, and even those of us working closely with these issues are unable or unwilling to recognise it. This makes it difficult to develop appropriate responses; while we must obviously take measures to deal with wildfires now, and attempt to prevent and mitigate any damage from similar local events in future, we must also deal not just with the symptoms but with the cause; otherwise we are doomed to firefighting forever.