Künstliche Intelligenz (KI) ist eine Schlüsseltechnologie. Sie kommt in sehr unterschiedlichen Lebensbereichen zum Einsatz – von der Medizin über Mobilität und Verkehrsplanung bis hin zu digitaler Kommunikation. AI wird moderne Gesellschaften auch weiterhin grundlegend verändern. Daher haben Politik und Wirtschaft in vielen Ländern beträchtliche Ressourcen in die Entwicklung dieser Technologie investiert.
Zugleich war und ist künstliche Intelligenz Gegenstand umfangreicher und intensiver öffentlicher Debatten. Auch wenn in diesen Debatten nicht immer klar ist, was mit KI genau gemeint ist und der Begriff vage und inkonsistent verwendet wird, werden dort doch teils weitreichende Versprechungen gemacht oder Ängste geschürt. Diese öffentlichen Beschreibungen – die auch 'sociotechnical imaginaries» genannt werden – sind Ausdruck von gesellschaftlichen Vorstellungen über die Möglichkeiten, Potenziale und Risiken, die in Technologien wie der KI stecken. Und sie prägen massgeblich, wie Gesellschaften mit derartigen Schlüsseltechnologien umgehen.
Das Projekt stellt diese Beschreibungen, ihre Entstehung und Auswirkungen daher in den Mittelpunkt. Es vergleicht die öffentlichen Vorstellungen von KI in verschiedenen Ländern und im Zeitverlauf. Es konzentriert sich auf verschiedene Stakeholder und unterschiedliche öffentliche Foren, die die gesellschaftliche Debatte über KI prägen. Das Projekt leistet damit einen konzeptionellen und methodischen Beitrag zur Erforschung öffentlicher Debatten über neue Technologien.
Methodisch kombiniert das Projekt qualitative, quantitative und computergestützte Methoden der Sozialwissenschaften. Es fokussiert die Debatten in den USA, in China und in Deutschland und untersucht dort sowohl die Berichterstattung in journalistischen Medien als auch Online-Kommunikation und soziale Medien.
BACKGROUND AND RATIONALE: Artificial intelligence (AI) is considered a key technology in contemporary societies, expected to substantially change sectors as diverse as health, warfare, mobility or communication. Consequently, political and economic stakeholders in many countries have allocated considerable resources to AI development, and the technology has been the object of extensive public debates. At the same time, observers and scholars have criticized these public debates for using the concept of AI vaguely and inconsistently, for over-hyping its promise, and for oscillating between naive hopes and dystopian fears. This is indicative of the fundamental interpretative flexibility of broad technologies such as AI: Their development and societal integration is contextual, shaped by political, economic and sociocultural interests, and subject to discursive negotiations in public arenas. These public debates are crucially important. Here, the “sociotechnical imaginaries”, i.e. the societal understandings of the possibilities, potentials and risks embedded in certain technologies, are socially constructed. These imaginaries play a large role in influencing how societies deal with emerging key technologies like AI. OBJECTIVES AND AIMS: The projects maps, explains and compares the public imaginaries of AI across countries and over time. It focuses on various stakeholders and different public fora which are indicative of societal views on AI and which play an important role in shaping these views. The analysis is guided by four research questions: 1. How can sociotechnical imaginaries - for and beyond the case of AI - be understood conceptually and analyzed empirically? 2. How do key stakeholders from politics, science, corporations and NGOs strategically try to shape prevalent imaginaries of AI, inserting their views into public debate, and which of them are successful in doing so? 3. What are the dominant imaginaries of AI, their sponsors and implications present in various public arenas?4. What are the similarities and differences between these public imaginaries of AI in different countries and over time? How do different stakeholders’ strategies to shape these imaginaries compare in different countries and over time? DATA AND METHODS: The project makes a conceptual and methodological contribution to the concept of sociotechnical imaginaries, which has gained prominence in scholarship over recent years, but remains underdetermined conceptually and hard to assess empirically. The project will propose a conceptual understanding of public imaginaries that foregrounds the public negotiation of imaginaries and proposes an empirical procedure that allows researchers to assess such imaginaries in large corpora of public communication.Empirically, the project investigates public imaginaries of AI with a longitudinal and comparative research design. In the temporal dimension, it builds on a longitudinal design that addresses the approx. 10 formative years of AI development, from the 2010s until now. In the spatial dimension, it contrasts characteristics and developments in three countries which are key to the future global development of AI: the US, China and Germany. Methodologically, the project combines qualitative, quantitative and computational methods of social science. It uses qualitative expert interviews and document analyses to assess the content, production structures and strategies of different stakeholders trying to influence public imaginaries of AI, and automated and manual content analyses to reconstruct imaginaries of AI in news media, online communication and social media.EXPECTED RESULTS AND IMPACT: The project provides a comprehensive comparative study on the construction of public imaginaries of AI in the crucial time frame of 2012 to 2021 and in key countries. It contributes to the growing scholarship on the societal uptake of AI and of key technologies more generally. It offers both lacking systematic empirical results as well as conceptual innovation, providing a conceptual understanding of the widely used yet under-defined concept of sociotechnical imaginaries. The project also contains outreach elements to interface with stakeholders of AI development, and provide them with the means to better assess and potentially regulate AI. Thus, we will not only unveil AI trajectories, but also help policy and technology-makers to adjust future trajactories. In sum, the project meets the strong demand for developing expertise both in understanding the unfolding development of AI and in shaping this development for the public benefit.
Last updated:07.06.2022
