-
Friend or Foe? Exploring the Implications of Large Language Models on the Science System
Authors:
Benedikt Fecher,
Marcel Hebing,
Melissa Laufer,
Jörg Pohle,
Fabian Sofsky
Abstract:
The advent of ChatGPT by OpenAI has prompted extensive discourse on its potential implications for science and higher education. While the impact on education has been a primary focus, there is limited empirical research on the effects of large language models (LLMs) and LLM-based chatbots on science and scientific practice. To investigate this further, we conducted a Delphi study involving 72 exp…
▽ More
The advent of ChatGPT by OpenAI has prompted extensive discourse on its potential implications for science and higher education. While the impact on education has been a primary focus, there is limited empirical research on the effects of large language models (LLMs) and LLM-based chatbots on science and scientific practice. To investigate this further, we conducted a Delphi study involving 72 experts specialising in research and AI. The study focused on applications and limitations of LLMs, their effects on the science system, ethical and legal considerations, and the required competencies for their effective use. Our findings highlight the transformative potential of LLMs in science, particularly in administrative, creative, and analytical tasks. However, risks related to bias, misinformation, and quality assurance need to be addressed through proactive regulation and science education. This research contributes to informed discussions on the impact of generative AI in science and helps identify areas for future action.
△ Less
Submitted 16 June, 2023;
originally announced June 2023.
-
Data Protection Impact Assessment for the Corona App
Authors:
Kirsten Bock,
Christian R. Kühne,
Rainer Mühlhoff,
Měto R. Ost,
Jörg Pohle,
Rainer Rehak
Abstract:
Since SARS-CoV-2 started spreading in Europe in early 2020, there has been a strong call for technical solutions to combat or contain the pandemic, with contact tracing apps at the heart of the debates. The EU's General Daten Protection Regulation (GDPR) requires controllers to carry out a data protection impact assessment (DPIA) where their data processing is likely to result in a high risk to th…
▽ More
Since SARS-CoV-2 started spreading in Europe in early 2020, there has been a strong call for technical solutions to combat or contain the pandemic, with contact tracing apps at the heart of the debates. The EU's General Daten Protection Regulation (GDPR) requires controllers to carry out a data protection impact assessment (DPIA) where their data processing is likely to result in a high risk to the rights and freedoms (Art. 35 GDPR). A DPIA is a structured risk analysis that identifies and evaluates possible consequences of data processing relevant to fundamental rights and describes the measures envisaged to address these risks or expresses the inability to do so. Based on the Standard Data Protection Model (SDM), we present a scientific DPIA which thoroughly examines three published contact tracing app designs that are considered to be the most "privacy-friendly": PEPP-PT, DP-3T and a concept summarized by Chaos Computer Club member Linus Neumann, all of which process personal health data. The DPIA starts with an analysis of the processing context and some expected use cases. Then, the processing activities are described by defining a realistic processing purpose. This is followed by the legal assessment and threshold analysis. Finally, we analyse the weak points, the risks and determine appropriate protective measures. We show that even decentralized implementations involve numerous serious weaknesses and risks. Legally, consent is unfit as legal ground hence data must be processed based on a law. We also found that measures to realize the rights of data subjects and affected people are not sufficient. Last but not least, we show that anonymization must be understood as a continuous process, which aims at separating the personal reference and is based on a mix of legal, organizational and technical measures. All currently available proposals lack such an explicit separation process.
△ Less
Submitted 18 January, 2021;
originally announced January 2021.
-
Social Networks, Functional Differentiation of Society, and Data Protection
Authors:
Jörg Pohle
Abstract:
Most scholars, politicians, and activists are following individualistic theories of privacy and data protection. In contrast, some of the pioneers of the data protection legislation in Germany like Adalbert Podlech, Paul J. Müller, and Ulrich Dammann used a systems theory approach. Following Niklas Luhmann, the aim of data protection is (1) maintaining the functional differentiation of society aga…
▽ More
Most scholars, politicians, and activists are following individualistic theories of privacy and data protection. In contrast, some of the pioneers of the data protection legislation in Germany like Adalbert Podlech, Paul J. Müller, and Ulrich Dammann used a systems theory approach. Following Niklas Luhmann, the aim of data protection is (1) maintaining the functional differentiation of society against the threats posed by the possibilities of modern information processing, and (2) countering undue information power by organized social players. It could be, therefore, no surprise that the first data protection law in the German state of Hesse contained rules to protect the individual as well as the balance of power between the legislative and the executive body of the state. Social networks like Facebook or Google+ do not only endanger their users by exposing them to other users or the public. They constitute, first and foremost, a threat to society as a whole by collecting information about individuals, groups, and organizations from different social systems and combining them in a centralized data bank. They transgress the boundaries between social systems that act as a shield against total visibility and transparency of the individual and protect the freedom and the autonomy of the people. Without enforcing structural limitations on the organizational use of collected data by the social network itself or the company behind it, social networks pose the worst totalitarian peril for western societies since the fall of the Soviet Union.
△ Less
Submitted 14 June, 2012;
originally announced June 2012.
