about our most intimate feelings and behaviour. We seem to have embarked on an irreversible track of trusting them. Predictive analytics reigns supreme in financial markets where automated trading and fintech risk assessments were installed long ago. They are the backbone of the military’s development of autonomous weapons, the actual deployment of which would be a nightmare scenario.
However, the COVID-19 pandemic has revealed that we are far less in control than we thought. This is not due to faulty algorithms or a lack of data, although the pandemic has revealed the extent of grossly underestimating the importance of access to quality data and its interoperability. There was no need for predictive algorithms to warn of future epidemics; epidemiological models and Bayesian statistical reasoning were sufficient. But the warnings went unheard. The gap between knowing and doing persists if people do not want to know or offer many reasons to justify their inaction. Thus, predictions must also always be seen in context. They can fall on fallow ground or lure us into following them blindly. Predictive analytics, although couched in the probabilities of our ignorance, comes as a digital package that we gladly receive, but rarely see a need to unpack. They appear as refined algorithmic products, produced by a system that appears impenetrable to most of us, and often jealously guarded by the large corporations that own them.
Thus, the observations made during my patchy journey began to converge on the power of prediction and especially the power exerted by predictive algorithms. This allowed me to ask questions such as ‘how does Artificial Intelligence change our conception of the future and our experience of time?’ I could return to my long-standing involvement with the study of social time, and in particular the concept of Eigenzeit, which was the subject of a book I wrote in the late 1980s. A few years ago I followed up with ‘Eigenzeit. Revisited’, in which I analysed the changes introduced through our interaction with digital media and devices that had by then become our daily companions (Nowotny 2017). New temporal relationships have emerged with those who are physically distant but digitally close, so that absence and presence as well as physical and digital location have converged in an altered experience of time.
Neither I nor others could have imagined the meaning that terms like physical and social distancing would acquire only a few years later. In the midst of the COVID-19 pandemic, I saw my earlier diagnosis about an extended present confirmed. My argument had been that the line separating the present from the future was dissolving as the dynamics of innovation, spearheaded by science and technology, opened up the present to the many new options that were becoming available. The present was being extended as novel technologies and their social selection and appropriation had to be accommodated. Much of what had seemed possible only in a far-away future now invaded the present. This altered the experience of time. The present was becoming both compressed and densified while extending into the immediate future (Nowotny 1989).
What I observe now is that the future has arrived. We are living not only in a digital age but in a digital time machine. A machine fuelled by predictive algorithms that produce the energy to thrust us beyond the future that has arrived into an unknown future that we desperately seek to unravel. Hence, we scramble to compile forecasts and engage in manifold foresight exercises, attempting to gain a measure of control over what appears otherwise uncontrollable because of its unpredictable complexity. Predictive algorithms and analytics offer us reassurance as they lay out the trajectories for future behaviour. We attribute agency to them and feel heartened by the messages they deliver on the predictions that concern us most. Such is our craving for certainty that even in cases when the forecast is negative, we feel relieved that we at least know what will happen. In offering such assurance, algorithmic predictions can help us to cope with uncertainty and, at least partly, give us back some control of the future.
My background in science and technology studies (STS) allowed me to bridge the gap between science and society and reach a better understanding of the frictions and mutual misunderstandings that beset this tenuous and tension-ridden relationship. STS opens up the possibility of observing how research is actually carried out in practice and allows us to analyse the social structures and processes that underpin how science works. The pandemic has merely added a new twist, albeit a largely unfortunate one. While at the beginning of the pandemic science took centre-stage, combined with the expectation that a vaccine could soon be developed and therapeutic cures were in the pipeline, science soon became mired in political opportunism. A nasty ‘vaccine nationalism’ arose, while science was sidestepped by COVID-19 deniers and conspiracy theories that began to flourish together with anti-vax and extreme-right political movements. After a brief and bright interlude, the interface between science, politics and the public became troubled again.
The pandemic offered an advanced testing ground, especially for the biomedical sciences, whose recourse to Artificial Intelligence and the most recent digital technologies proved to be a great asset. It allowed them to sequence the genomes of the virus and its subsequent mutations in record time, with researchers sharing samples around the world and repurposing equipment in their labs to provide added test facilities. It enabled the COVID-19 High Performance Consortium, a public-private initiative with the big AI players and NASA on board, to aggregate the computing capability of the world’s fastest and most advanced computers. With the help of Deep Learning methods it was possible to reduce the 1 billion molecules analysed for potential therapeutic value to less than a few thousand.
The response to the pandemic also brought a vastly increased role for data. The pressure was enormous to proceed as quickly as possible with whatever data was available, in order to feed it into the simulation models that data scientists, epidemiologists and mathematicians were using to make forecasts. The aim was to predict the various trajectories the pandemic could take, plotting the rise, fall or flattening of curves and analysing the implications for different population groups, healthcare infrastructure, supply chains and the expected socio-economic collateral damage. Yet, despite the important and visible role given to data throughout the COVID-19 pandemic, no quick quantitative data-fix emerged that would provide a solid basis for the measures to be taken. If the data quality is poor or the right kind of data does not exist, a supposed asset quickly turns into garbage that contaminates simulation models and radically reduces their usefulness for society.
To some extent, the COVID-19 crisis has overshadowed the ongoing discussion about innovation and how scientific findings are transferred into society. It is therefore appropriate to recall the work of STS scholars who have extensively analysed the social shaping of technologies. Their findings show that technologies are always selectively taken up. They are gendered. They are appropriated and translated into products around which new markets emerge that give another boost to global capitalism. The benefits of technological innovation are never equally distributed, and already existing social inequalities are deepened through accelerated technological change. But it is never technology alone that acts as an external force bringing about social change. Rather, technologies and technological change are the products and the outcome of societal, cultural and economic preconditions and result from many co-productive processes.
Seen from an STS perspective, what is claimed to be entirely novel and unique calls for contextualization in historical and comparative terms. The current transformation can be compared to previous techno-economic paradigm shifts that also had profound impacts on society. In the age of modernity, progress was conceived as being linear and one-directional. Spearheaded and upheld by the techno-sciences, the belief was that continued economic growth would assure a brighter and better future. It came with the promise of being in control, manifest in the overconfidence that was projected into planning. This belief in progress has, however, been on the wane for some time, and more recently many events and developments have injected new doubts. The destruction of the natural environment on a global scale confronts all of us with an ‘inconvenient truth’, reconfirmed by the Fridays for Future movement that has galvanized the younger generation. In addition, the pandemic has demonstrated the helplessness of many governments and the cynicism of their responses, while coping with the long-term consequences will require a change in direction.
The remarkable speed of recent advances in AI and its convergence with the sustainability crisis invites the question: What is different this time?
