AI as an Enabler for Science - Report on the 3rd NFDI Berlin-Brandenburg Network Meeting
On 21 May, NFDI4DataScience (NFDI4DS) and MaRDI welcomed about 40 participants from 16 NFDI consortia to the Weizenbaum Institute, Berlin, to explore how AI can accelerate research.
Keynotes
- Alan Ang (Wikimedia Deutschland) showed how chatbot-style Retrieval-Augmented Generation opens Wikidata’s knowledge graph to new users.
- Georg Rehm (DFKI) presented “shared tasks” as community competitions that drive cutting-edge AI development.
Tool demos
Two demo sessions showcased diverse NFDI tools in action—from AI-ready materials data and self-driving labs to conversational knowledge graphs and smart help-desk services.
Take-aways
- AI already speeds up data acquisition, analysis, modeling, and experiment design.
- Chatbots and RAG interfaces are becoming key gateways to research data and KGs.
- FAIR, high-quality data remains the foundation for sustainable AI services.
Next steps
Future meetings will focus on training, transparency, reproducibility, fairness, and more interactive formats.
Stay informed via the NFDI_BB mailing list or share ideas at nfdi_bb@mardi4nfdi.de.
The full report is available here.