If Europe wishes to assert itself as a hub in the digital economy – it must undertake the responsibility of spreading digital literacy and become a source of knowledge creation and innovation. This digital culture goes beyond research and education, and entails embracing a multidisciplinary approach that combines hard and social sciences.
The research community in the United States has for long understood the necessity to view digital technology as a societal phenomenon in order to aid its development. Ivy league Universities such as Stanford, Yale and Harvard have thus created research centers that fall under the category of “Internet and Society”. These are the cases of The MIT MediaLab established in 1985, and the Stanford Center for Internet and Society. Other research centers advantageously adopt more contemporary titles, such as Data&Society or AI Now both located in New York, founded respectively by a new generation of researchers danah boyd and Kate Crawford.
These institutes are at the core of the conversation on digital ethics and particularly ethics of artificial intelligence, receiving 27 million dollars in funding from the Ethics and Governance of Artificial Intelligence Fund to study the impact of Artificial Intelligence on society:
“Because of this pervasive impact, it is imperative that AI research and development be shaped by a broad range of voices—not only by engineers and corporations—but also social scientists, ethicists, philosophers, faith leaders, economists, lawyers, and policymakers.”
What sets these institutes apart is their multidisciplinary approach. Combining hard and social sciences, they are comprised of historians, engineers, philosophers, programmers, mathematicians, sociologists, biologists and neuroscientists and more - confronting ideas and viewpoints in an experimental environment, resulting in a comprehensive lens that combines theory and practice to understand artificial intelligence and digital technology and their impacts on society today and tomorrow.
What are the ethics of self driving cars? How do we avoid the propagation of fake news? Is encryption necessarily a social good in a democratic society? Should we allow individuals the right to be forgotten? If so, to what extent and to which spheres of the digital world? What are the prospects of killer robots and drones? How can we avoid the pitfalls of predictive justice, and how can we avoid the transference of human biases onto machines? Should coding be a mandatory part of the schooling curriculum? These questions, amongst many others, are theoretical questions that will directly affect the lives of individuals, society as a whole and the development of digital technologies.
Europe holds advanced institutes tackling these questions – the Oxford Internet Institute, the Humboldt Internet Institute are striking examples. However, these institutions are too far in between. In France, the brilliant Bruno Latour created the Medialab of Sciences Po in 2009. These initiatives, that we occasionally find at PSL, at CNRS or at François Taddei’s CRI remain unfortunately rare.
The nature of the ever evolving technological landscape creates an urgency that requires immediate attention. In 2017, a report created by Dell and the Institute for the Future asserts that roughly 85% of jobs that will exist in 2030 do not exist today. New disciplines and professions will see the light of day: history of technology, robotic sociology, data ethics etc.. Dubbed under the label of “digital humanities”, it showcases more poignantly the need for knowledge hybridization, in particular in the curriculum of future generations.
This urgency goes beyond preparing for the medium to long term impact of emerging technologies on employment and education – but understanding how to tackle short-term issues that can potentially have corrosive effects on democratic systems, civic engagement, hiring practices and many more.
The main factors driving these short-term concerns are the increasing opacity of machine learning algorithms, their misuse in the public and private sectors and their potential to exacerbate human biases. These concerns were on full display at the most edition of the Conference on Neural Information Processing Systems (NIPS), where Kate Crawford of AI Now explored the rising area of research around bias in ML systems. It highlighted the importance of understanding bias and the different categories of harm that may arise as a result, and multiple practical strategies that can potentially alleviate this hurdle. The implications of bias become extraordinarily clear in settings of predictive justice, automated hiring and the allocation of work in the ever growing gig economy.
On its side – France is reaching new heights in all things digital. Our country is clearly working towards its goal in becoming a global digital hub – capable of competing with Silicon Valley, Singapour, Shanghai, Bangalore, Tel Aviv, London and Berlin.
Recent events provide ample evidence for this growth. In 2016, aided by 230 incubators, French startups have raised a record two billion euros in funding. In 2017, France witnessed the inauguration of Station F – which after three years of renovation and 250 million euros in construction costs – has become the biggest startup cost in the world. And in 2020?.. It should be the time to create multidisciplinary research institutes dedicated to understand the impact of emerging technologies on society.
It is in this perspective that the report by French MP Cédric Villani, published in 2018, recommended the creation of a network of interdisciplinary research institutes on artificial intelligence, a proposal that we had defended to the MP during the consultation organized for the drafting of this report.
Following the publication of this report and on the occasion of the “AI for Humanity” Summit, President Emmanuel Macron unveiled France's artificial intelligence strategy. In particular, he announced the establishment of an "emblematic network of four or five dedicated institutes, anchored in university centers and linking the territory".
The Digital Culture project as we imagine it materializes with the launching of the interdisciplinary PR[AI]RIE Institute in early October 2019 with Yann Le Cun's inaugural speech. This institute, which has 45 research chairs, will focus its research on the fields of health, transport, environment and biology. It is supported by several large companies, mainly acting in the industrial sector.
The promise of these research institutes reinforces multiple synergies. Whether it is in creating innovative products and the bolstering of European research and development, understanding the skills needed in tomorrow’s professions, enabling the political class towards further understanding the implications of technology on society, and developing a European perspective on the ethics and governance of emerging digital technology.