(How) can AI help build better online spaces? ✨ Last week, 16 researchers from 8 institutions gathered in Vienna to explore exactly that question. Between the Christmas markets, the WHAT-IF team discussed how LLMs can power agent-based simulations to tackle online hostility and harmful content 👇

Antico Caffè Torinese, Trieste 🇮🇹 1st visit. Stunning classical café with a booze & cakes focus. Red River & Erdinger on tap, special cocktails and wall of Amaro. The sort of bar you dream of finding in Italy. 🍸 🍻👍

Our next speaker is @emmafraxanet.bsky.social, who will talk to us about ideological segregation and news quality in social media #CCS2025

Next, we have @maxpe.bsky.social, with a talk on the semantic distance between comments in online newspaper forums #CCS2025

It was a great week at IC2S2 in Norrköping! This community is really something. Today I got to share our findings on tracking engagement with news content on Facebook (check the preprint: arxiv.org/abs/2409.13461 )...and I was lucky enough to do so on the amazing auditorium stage✨️

Excited to present our paper with @maxpe.bsky.social, @dgarcia.eu, and @mstrohm.bsky.social next week at #ICWSM! ✨ We extend social simulation with LLMs to a longitudinal setting by fine-tuning Temporal Adapters—here's how: 🧵

WHAT-IF team members met in Konstanz last week to design the WHAT-IF machine—an agent-based simulation with LLM-driven agents. It models media ecosystems to test 'what if' policy scenarios. Curious how? Read more: what-if-horizon.eu/2025/05/16/o...

Paper is out! Together with Camelia Oprea and Markus Strohmaier, we have explored the potential impact of toxic comments on Wikipedia: academic.oup.com/pnasnexus/ar...