Artificial Intelligence Opinion Monitor (MeMo:AI)

In recent years, numerous strategy papers on the development of digitization and so-called artificial intelligence (AI) have been published from the European to the state level. What they all have in common is the intention of a stronger orientation toward people and the common good. If politics takes its own claim seriously, it is dependent on the support of an active and enlightened civil society that contributes its needs and ideas to the shaping of digitization policy in a mature and self-determined manner.

Otherwise, the development and implementation of AI technologies will remain the preserve of interested industry, whose primary goal is to maximize profits and increase efficiency. Studies from the U.S. and the U.K. show that the public debate to date has been heavily dominated by industry (Brennen et al. 2018; Fast & Horvitz 2017). These actors argue from their own specific logic. So far, no systematic findings are available for the German discourse. The Meinungsmonitor Künstliche Intelligenz [MeMo:AI] aims to close this gap. In research partnership with the Center for Advanced Internet Studies (CAIS), a team of communication and social scientists led by Prof. Dr. Frank Marcinkowski is investigating the constitution and change of public and media published opinions on AI.

The project is based on the assumption that the realization of a comprehensive claim to shape society requires a politicization of the topic. This means, first of all, that the topic of AI and digitization is recognized as an object of political debate and will formation. Research from other topic areas such as EU politics (e.g., De Wilde, 2011; Schattschneider, 1957) shows that three conditions must be met: (1) the existence of alternative positions in the public debate (polarization); (2) broad media attention that ensures popularization of issues and dissemination of different positions (intensity); and (3) societal attention to an issue and electoral significance (resonance). An apt example of such a development is the recent issue and political career of climate change. To what extent there will be a comparable trajectory for AI, MeMo:KI will assess this by systematically monitoring public and published opinion.

The study uses a combination of different empirical methods. Both time-honored methods of social science research such as media content analyses and surveys as well as new computer-based methods are used:

  • Continuous monitoring of interest in AI, relevant opinions and intended behavior is carried out via a monthly survey of the German population aged 18 and older who use the Internet at least occasionally. In addition, special surveys will address current and specific issues surrounding AI.
  • In parallel, a semi-automated topic analysis of media coverage will be carried out. This will make it possible to map the course of public and published opinion on AI. In the following, the findings will be deepened with manual content analysis and thus, among other things, the tenor of the reporting or the actor structures will be examined.
    • Since April 2021, the project has been funded by Stiftung Mercator for a period of three years.
      Previously, the Ministry of Culture and Science of the State of North Rhine-Westphalia funded the project for a period from January 2020 to March 2021.