We are above excited to announce that our TWON colleague, Ljubisa Bojić of the University of Belgrade, has published an extensive study addressing a compelling question: Does GPT-4 surpass human performance in linguistic pragmatics? The paper explores whether large language models (LLMs) are capable of understanding nuanced, often implied meanings in human communication that go beyond the literal and depend on context, irony, sarcasm, or subtle conversational cues.

The study examined five LLMs (GPT-2, GPT-3, GPT-3.5, GPT-4, and Google’s Bard) alongside two groups of human participants: Serbian speakers of English as a second language and U.S. native English speakers. Each model and participant was asked to interpret a series of dialogue-based tasks specifically designed to test pragmatic understanding, drawing on Gricean communication principles such as relevance, clarity, and implicature. Their responses were evaluated using a standardized five-point scale, where a score of ‘1’ indicated poor or superficial understanding, and a ‘5’ signaled a deep and accurate interpretation of implied meaning, including the detection of sarcasm, irony, and other contextual subtleties.

The results weremore than interesting. GPT-4 not only outperformed all other AI models but also exceeded the performance of human participants. GPT-4 achieved an average score of 4.80, compared to the highest human score of 4.55. On average, human participants scored significantly lower (the Serbian group averaging 2.80 and the US group 2.34) while the LLMs, overall, averaged 3.39. GPT-4 ranked first among all 155 evaluated participants.

These findings carry important real-world implications. If AI can consistently interpret pragmatic cues better than humans, it could lead to more advanced and intuitive interactions between people and machines. For instance, this could dramatically enhance the capabilities of virtual assistants, customer service bots, and social robots, making them more adept at recognizing intent, tone, and emotion. Such improvements could prove especially valuable in fields like mental health, education, and conflict resolution, where reading between the lines is often crucial.

At the same time, these advances raise important ethical considerations. As we begin to rely more on AI for interpreting nuanced communication, there is a risk of misinterpretation or misuse, particularly in sensitive contexts. It also raises questions about accountability and the potential consequences of AI misunderstanding or manipulating human intent.

In short, while the ability of GPT-4 to surpass human performance in linguistic pragmatics marks a major milestone for AI, it also underscores the need for thoughtful, responsible integration of such technologies into society. The study offers a glimpse into the future of human–AI communication; one that is more natural, perceptive, and possibly more capable than we previously imagined.

Find the open-access paper here. 

June 2nd, 2025 – Portorož, Slovenia | Part of ESWC 2025

The 1st Workshop on Semantic Generative Agents on the Web, held on June 2nd in Portorož, Slovenia, as part of the Extended Semantic Web Conference (ESWC 2025), marked a key milestone in disseminating the goals and findings of the TWON project to the academic community. The event brought together researchers from diverse disciplines to explore how Semantic Web technologies and Large Language Models (LLMs) can be combined to develop intelligent, interpretable, and communicative agents for the web.

Opening Keynote

The workshop opened with a keynote by Matthias Nickles (National University of Ireland, Galway), who presented a comprehensive overview of the history and recent advancements in generative agents within computer science, setting the stage for the diverse presentations to follow.

Paper Presentations

Jan Lorenz (Constructor University) kicked off the presentations with a talk on “Filter Bubbles in an Agent-Based Model Where Agents Update Their Worldviews with LLMs“. His work replaced abstract numerical opinion spaces with LLM-generated human-like statements to simulate opinion dynamics. The goal was to assess whether filter bubbles would still emerge in this more realistic setting and to examine the practical integration of LLMs into agent-based simulations.

Next, Martin Žust (Jožef Stefan Institute) presented a web-based negotiation agent designed to assist unskilled negotiators in real time. The agent transcribes dialogue, builds dynamic world models, and combines analytical reasoning with human-like intuition to offer context-aware negotiation support. This hybrid approach aims to enhance interpersonal outcomes through collaborative human-AI interaction.

Abdul Sittar (Jožef Stefan Institute) followed with an agent-based simulation of social media engagement during German elections. By incorporating past conversational history, motivational factors, and resource constraints, the model used fine-tuned AI to generate posts and replies, applying sentiment analysis, irony detection, and offensiveness classification. The findings highlighted how historical context shapes AI responses and how behavior shifts under different temporal constraints.

Afternoon Keynote and Talks

In the afternoon keynote, Denisa Reshef Kera (Bar-Ilan University) addressed philosophical perspectives on generative agents, focusing on bias, representation, and agency. She emphasized the role of generative agents in public policy and civic participation, highlighting their potential for enhancing digital society.

Ljubisa Bojic (University of Belgrade) presented an innovative AI-based recommender system designed to reduce echo chambers and polarization. His model incorporates emotional tone, content diversity, and political balance into the recommendation process, improving content exposure without sacrificing accuracy. The approach aligns with ethical AI principles, offering user autonomy through customizable preferences.

Denisa Reshef Kera returned with Avital Dotan to present “AI Beyond Rules, Heuristics, and Dreams“, introducing the concept of ergative-absolutive AI agents. Drawing on linguistic structures from languages like Basque, they proposed a new way of modeling agency in LLMs—treating them not just as predictors but as performative systems that enact grammar and interaction. Their two-step framework involves analyzing grammatical alignments and creating participatory simulations with diverse agent alignment patterns to encourage adaptive and accountable behavior.

Simon Münker (University of Trier) concluded the paper presentations with “twony“, a micro-simulation platform that models emotional contagion and discourse dynamics in online social networks. Using fine-tuned BERT models and LLMs to simulate politically engaged personas, twony visualizes emotional cascades in various feed algorithm scenarios—offering a powerful, open-source tool for explaining polarization and online behavior.

Closing Discussion

The workshop concluded with a fishbowl discussion featuring Achim Rettinger, Damian Trilling, Marko Grobelnik, Matthias Nickles, and Denisa Reshef Kera. The panel reflected on the interdisciplinary insights presented throughout the day and discussed future directions for generative agents in real-world applications.

Takeaways

SemGenAge 2025 fostered rich dialogue across fields including semantic web technologies, AI, computational social science, and digital media studies. Discussions emphasized the potential of generative agents in areas such as online discourse moderation, content recommendation, opinion shaping, and consumer behavior analysis.

The workshop’s insights will directly support TWON’s mission: combining empirical observations, simulation, and participatory methods to create evidence-based recommendations for improving social network regulation and enhancing digital citizenship.

For full program details, visit the official workshop page.

From May 30 – June 1st, all nine TWON partner institutions gathered in Portorož, Slovenia for the fifth TWON Consortium Meeting. This in-person event offered a key opportunity to assess our progress, align goals, and prepare for the final year of the project.

The meeting began with the general assembly led by consortium leader Damian Trilling, where we reviewed project milestones, discussed ongoing challenges, and set priorities for the months ahead. A central objective was to optimize integration across TWON’s thematic and methodological strands, reinforcing the coherence of our collective efforts. We then focused on planning our large-scale simulations — from sharpening research questions to technical implementation. The day concluded with a consortium dinner.

On Sunday, we had a workshop on design features of a democracy-preserving online social network, as a step towards developing policy recommendations. Later on, we had a session for plannung the upcoming Citien Labs, where we discuss our research with citizens. The day concluded with a plenary wrap-up, and an early-career researcher event, where we had the chance to discuss and feedback our PhD projects with each other.

The Portorož meeting not only advanced TWON’s agenda but also reinforced collaboration at a critical stage of the project. With renewed momentum, the consortium is well-prepared for the final project year.

Thank you, Jozef Stefan Institut for hosting us in beautiful Slovenia!

In TWON we do not only want to translate our scientific results into actionable recommendations for decision-makers in politics and industry – we also want to foster digital citizenship and the public debate on the role Online Social Networks should play in our society. This is why, our CitizenLabs are an essential part of TWON!

The Citizen Labs are conducted by TWON consortium member “DialoguePerspectives. Discussing Religions and Worldviews e.V.” who trains young European leaders to become experts in a new, society-oriented interreligious-worldview dialogue. The program brings together participants from diverse communities and backgrounds, encompassing individuals with 19 different religions and beliefs across 25 European countries. Through their unique perspectives and expertise, they contribute to fostering understanding, cooperation, and a pluralistic, democratic, and cohesive Europe.

Following the first Citizen Lab in Karlsruhe in September 2024, the second Citizen Lab recently took part in Vienna from May 11-14th, 2025! 35 young leaders from various communities from all over Europe came together to discuss a digital and pluralistic European future. The program was a mixture of input and discussion from our TWON researchers, impulses from external researchers and civil society organizations, reflexional workshops and a public evening event – all centered around tackling hate, misinformation and polarization in the age of AI and tech-oligarchs.

Dr. Jonas Fegert opened up the discussion, introducing the group to the TWON project, the goals and the need. Prof. Dr. Damian Trilling (University of Amsterdam) opened up a critical conversation around the limits of current research on social media dynamics, challenging our assumptions about echo chambers, filter bubbles, and the spread of disinformation. His interactive talk underscored how intuitive beliefs often outpace empirical evidence — and invited us to think more deeply about what we can actually measure.

Prof. Dr. Achim Rettinger (Trier University) tackled the complex intersection of AI agents and online discourse. Can AI replace us in some communicative functions — and should it? His workshop addressed both the dangers and opportunities of algorithmic content curation, especially in shaping public opinion and emotional response.

We also had the chance, to present our demonstrators micro & macro TWONy to the public! Led by Simon Münker (Trier University) and Fabio Sartori (KIT), participants explored our tools with which we try to make our simulations with generative agents tangible. The hands-on experience allowed for nuanced discussions about how different ranking logics affect emotional dynamics.

FZI researcher Cosima Pfannschmidt  led a workshop on envisioning a democratic online social network of the future. What would such a platform look like? Who would own and govern such a platform? How should content moderation be organized? Which criteria would the ranking algorithm prioritize? While it is crucial to research negative effects of online social networks, it is equally important to develop actionable, democratic alternatives.

At our public evening event in the Vienna Co-Innovation Factory, we dove deeper into the topic of digital democracy. Moderated by Dr. Jonas Fegert (FZI), Prof. Dr. Achim Rettinger (Uni Trier & FZI), Benjamin Fischer (CeMAS), Judith Peterka (TWON Advisory Board), Natascha Strobl (Expert on Right-wing Extremism and the New Right) and Dr. Sebastian Heidebrecht (EIF – Centre for European Integration Research, Vienna University) discussed on “Digital Democracy and the Power of Platforms: Policy, AI, and Accountability”. In a second panel session, Alina Bricman (Director of EU Affairs at B’nai B’rith International), Rosa Jellinek (Activist, Social Media Expert, Keshet Deutschland e.V.), Selin Aydın (Programme Manager CLAIM – Alliance Against Islamo-phobia and Anti-Muslim Hate) and Stefania Manca (Institute of Educational Technology, Italian National Research Council) discussed on „Countering Hate and Information Manipulation – Strategies for a Safer Digital Sphere“.

Beyond this the Citizen Lab included fantastic inputs from external researchers and civil society organizations:

• Algorithmic Amplification of Hate Speech and Misinformation with Dr. Ing. Even Kapros (Strategic Designer and Researcher on HCI, UX, and Ethics, CEO and founder of Endurae, Strategic Advisor with Project Arc)
• Foreign Influence: Digital Manipulation by Authoritarian States and Paths to Platform Accountability with Julia Smirnova (Senior Researcher, CeMAS)
• What’s Working, What’s Not: Recommender Systems and Platform Accountability (Dr. Julia Neidhardt, Head of the CD Lab for Recommender Systems, UNESCO Co-Chair for Digital Humanism, TU Vienna
• Voices That Echo: Holocaust Memory, Digital Activism and Peacebuilding in the Social Media Age with Dr. Stefania Manca | Institute of Educational Technology, Italian National Research Council)
• Social Media, AI, Disinformation, and Freedom of Speech with Nuriyatul Lailiyah | Assistant Professor Communication Department Faculty of Social and Political Sciences Diponegoro University Semarang Central Java
• World Café Discussions on Local Realities in the Digital Sphere with IGGÖ – Islamische Glaubensgemeinschaft in Österreich, JöH – Jüdische österreichische Hochschüler:innen and SEEDS – Security Education by Empowering Democratic Strength

The Citizen Lab concluded with a call to action: How can we translate these insights into political change? To wrap up the rich discussions, we developed policy proposals aimed at regulating algorithmic systems, increasing transparency, and creating digital environments where diverse voices can thrive. With DialoguePerspectives’ participatory approach we ensure that our policy recommendations are relevant, comprehensible, linked to the ongoing public debate and reflect the lived experiences of the diverse European communities. Participants worked on an existing TWON policy brief, which was previously developed in an iterative process with input from TWON researchers and the previous CitizenLab. The Vienna Citizen Lab 2025 reminded us that digital spaces are not neutral — they are designed, and as such, can be redesigned. To build an inclusive, democratic future, we need to shape not only the rules of online discourse, but the very structures that host them.

We are deeply grateful for the Vienna Citizen Lab 2025, hosted by TWON-partner DialoguePerspectives. What an unforgettable gathering that brought together European leaders committed to shaping democratic digital spaces! A heartfelt thank you to DialoguePerspectives and all those who made this experience so meaningful. It was an honor to be part of a space where listening, questioning, and reimagining Europe is not only possible—but already happening.