Unpublished translation of the paper « Prolègomènes à une stratégie critique de la donnée ». Prolegomena to a Critical Strategy of Data Amaël CATTARUZZA Ecoles de Saint-Cyr Coëtquidan CREC Saint-Cyr Guer, France amael.cattaruzza@st-cyr.terre-net.defense.gouv.fr “We don’t hesitate to qualify a device full of micro-chips as ‘technology’ but don’t expect to learn anything from it; we just expect a ‘technologist’ to repair it but are unwilling to share any knowledge about it. What would we do with it? There’s not much else to gain from technology. It’s just a heap of complex resources. Everyone knows that.” (Bruno Latour, 2010) Abstract - This paper considers the impact new data technologies, such as Big Data and machine learning, can have on strategy. It firstly assesses the opportunities offered by these innovative ways of knowledge with respect to recent vulnerabilities. Through the prism of digital data, it then attempts to qualify what is inhered by this new form of “thinking the world”. Following a survey on the debates and expectations currently emerging in military circles regarding such tools, it also questions the very notion of "data" in defining a strategy that goes beyond a strictly technical framework. The intent is to highlight the nature of such issues and the choices decision-makers are facing today. Keywords: data; strategy; geopolitics; Big Data; machine learning I. INTRODUCTION In recent years, parallel developments in cyberspace, data generation (the proliferation of sensors of all types with the onset of the “Internet of Things”), data storage and processing capacities (cloud computing, distributed processing) have resulted in the emergence of new tools such as Big Data or machine learning. Recent technologies have been the subject of media and industrial hype because promises linked with such technologies are viewed as revolutionary ones. The assumption is that massive real-time data processing would greatly improve our knowledge of the environment, how to interact with it and that it could even foretell events and behaviours in the short and mid-terms (Babinet, 2015). Scientific knowledge per se is about to be turned upside down: the ability to link different sets of data with each other via algorithms is replacing traditional causal factors which are the foundation of scientific thinking. In 2008, Chris ANDERSON, editor of the famous WIRED MAGAZINE, announced “the end of theory” and the invalidation of traditional scientific methods that seek to explain and interpret social phenomena (Anderson, 2008). “This is a world where massive amounts of data and applied mathematics replace every other tool that might be brought to bear. Out with every theory of human behaviour, from linguistics to sociology. Forget taxonomy, ontology, and psychology. Who knows why people do what they do? The point is they do it, and we can track and measure it with unprecedented fidelity. With enough data, the numbers speak for themselves”. Linking different data sets on a large scale and in real time would enable us to observe, detect and anticipate events and facts that were previously invisible or barely perceptible. However, with the increased speed of the data we generate - which has risen exponentially with the extensive use of sensors of all kinds and the development of the “Internet of Things” - including our steadily growing capacity to collect, store and process data at constantly lower costs, have created substantial expectations from recent technologies. Fields of research and practical applications have grown rapidly in several sectors, ranging from health and agriculture to the management of cities and/or detached houses. The military sector is no exception to this trend where strategic thinking is increasingly concerned with the potential gains from using Big Data and machine learning in an operational environment. As could be foreseen, these early approaches developed arguments recalling past debates framing the issue of the 1990s concept of Revolution in Military Affairs (RAM), where the use of modern technologies would provide a major strategic advantage. However, considering the difficulties encountered by US troops in the asymmetric wars led in Iraq and Afghanistan, the results had to be revised accordingly. For all that, what is at stake today may be more important than before because it clearly lies at the heart of military strategy following what Clausewitz referred to as "the fog of war". In all respects, the processing of massive data in real time may reduce or even remove the uncertainties normally facing decision-makers. Will the overwhelming flow of data and information freely available on the Internet and the extensive use of sensors of all kinds on soldiers, military devices and in the field thus transform a battlefield into a “transparent” battlefield? Will the accumulation and analysis of such data help decision-makers in forecasting social, political and strategic risks and thus substantially reduce future vulnerabilities? Comparable assertions generate numerous discussions and controversies, depending on the status granted to these recent instruments. The issue here is not to take part in the discussions between those wanting to suppress these technologies and those commending them, but more a matter of thoroughly analysing their likely impact on strategic thinking. Will such thinking disappear in our era of massive and ever-present data? Will the art of war limit itself to a technical operation for managing and analysing available data? Before reaching such radical conclusions, we should first appraise the opportunities offered by these new modes of knowledge in addressing current threats and vulnerabilities. Moreover, even if the above-mentioned technologies change our perception of the strategic landscape, several issues cannot simply be reduced to the analysis of digital data and, more particularly, human and psychological factors that lie at the heart of strategic knowledge. Hence, for us to grasp the significance of the