PS-DSI Meet Up: Brainstorming on how AI Tools can Facilitate your Research BYO-Lunch Meeting

The initiative was developed and will be hosted by Ozlem Durmaz-Incel (Pervasive Systems Research Group)  in collaboration with the Digital Society Institute (DSI). 

AI Research tools

Two noteworthy developments in AI for scientific research were announced over the past few weeks—developments that could become highly relevant for many of us across the natural sciences and engineering. Both OpenAI and Google introduced tools that move beyond information retrieval toward genuine synthesis, reasoning, and problem-solving.

1. OpenAI – Accelerating Science with GPT-5
OpenAI has published early experiments in which GPT-5 is used as a “research partner.” Unlike earlier generations, this model was tasked with contributing to open scientific problems: assisting with mathematical proofs, identifying biological mechanisms in minutes rather than months, and rediscovering symmetries in physics. A striking capability is its conceptual literature search, where it connects insights across fields that traditional keyword searches tend to miss.

• Paper: https://cdn.openai.com/pdf/4a25f921-e4e0-479a-9b38-5367b47e8fd0/early-science-acceleration-experiments-with-gpt-5.pdf
• Summary: https://openai.com/index/accelerating-science-gpt-5/

2. Google – Scholar Labs
Google has launched Scholar Labs, a new generative-AI feature within Google Scholar. Rather than merely listing papers, it analyses complex research questions and synthesises an answer by pulling evidence from multiple studies. It effectively automates the early stages of a literature review by explaining how different papers address different aspects of a research query.

Together, these developments suggest a shift from AI as a passive search tool to AI that actively supports synthesis, exploration, and scientific reasoning.

Invitation

We would like to explore these developments with a broader group of researchers across the natural sciences, engineering, computer science and related fields. You are warmly invited to join an open discussion session where we will consider questions such as:

• How might these tools fit into our own research workflows?
• Where could they offer genuine scientific value—or raise concerns about reliability, reproducibility, or authorship?
• What opportunities or risks do you see?

We also welcome your questions, suggestions, experiences, and—if you are interested—short presentations or demos of how you have already used AI in your research practice. Even a 5–10 minute contribution would be very welcome.

If this sounds interesting, please let us know; we would be delighted to include you.

Registration & contact 

Please register via the registration button below. For more information and questions regarding this open brainstorming session open to all interested UT-colleagues, limited amount of seats though, please, please contact Stephanie Hessing