Successful 5th ACM Digital Humanism Summer School

Vienna 10 July – Today is the last day of this amazing Summer School on the topic of Democracy in the Digital Age. Building on the success of previous editions, the aim of the Summer School is to provide a learning environment and platform to explore, investigate and discuss key issues and latest developments related to the impact of Digital Technologies to democratic societies. Internationally renowned and leading academics from computer science, social sciences, law and humanities present and discuss important recent topics and themes, including: Foundational issues: Foundational and philosophical issues of digital technologies and AI. Democratic governance of AI and social media: Regulation, multistakeholder governance, collective control of systems and platforms, policy-making, technical implementation of values and constraints. Civil society organizing in the digital age: Political communication, organizing of protest movements, the digital public sphere. AI for democratic innovation: Algorithms for participatory democracy (incl. AI supported deliberation, participatory budgeting, citizen assemblies and sortition…) Digital sovereignty: Privacy and security aspects, infrastructure and cloud services, data sovereignty.

During this 5 day summer school an international and diverse group of 63 women and men, PhD candidates, lecturers, researchers and professionals participated in interesting projects on democracy and the digital age. A visit to the Austrian Parliament was made, joint lectures with the ESSAI conference were assisted, a panel and keynote by Prof Sennay Ghebreab was followed by a panel discusssion. The gender balance was 52-48% for M-F.

The programme started on Monday 6 July with a talk by Prof George Metakides: Introduction to Digital Humanism ( see his slides), in which he explained the threats by the expanding influence of social media and AI to democracy worldwide, quoting Tim Berners-Lee: “The system is failing”.

This was followed by a talk by Prof. Edith Elkind on the Overton Window, and preference aggregation in Reinforcement Learning from Human Feedback. She works with Meta on LLMs, especially on fine-tuning the user-interface between the LLM and the user, to stay within the margins of the Overton window. The assumptions and decisions behind the algorithms used to develop so-called representative digital twins was not discussed.

In the afternoon a visit to the Austrian Parliament was made, in which issues related to digital technologies and democracy were discussed.