We conducted a barcamp session on motivation - because it is essential to motivate our students to learn the basics without AI tools. Only thereby they can gather an understanding of the fundamentals. Having learnt the basics, we then need to teach how to make use of AI tools efficiently.

Also, the first workshop on the Societal Effects of AI in mobile social media that we conducted yielded interesting discussions and ideas. A definitely relevant topic that we will pursue further.

Privacy improvements still proliferate only slowly. In a world where digital privacy increases in relevance, our work is important as it underlines which aspects users are especially afraid of and what needs to change from their perspective.

I am thankful that our paper "European Users' In-Depth Privacy Concerns with Smartphone Data Collection" received an honourable mention award - a great team effort with my coauthors @maximilianewindl.bsky.social , Tobias Knochbloch, and @svenmayer.bsky.social !

Check out our paper: doi.org/10.1145/3743...

By evaluating the placeholder text of a textfield, we derive the user's input intention. This allows to distinguish between private contents, public posts, or functional inputs such as search queries. Researchers who study typed language on smartphones can thereby obtain higher quality data.

Doruntina Murtezaj and I propose some HCI directions in our workshop paper ➡️ doi.org/10.48550/arX...

However, a discussion about industry experiences of our students showed that (web) developers are and will still be needed. They reported on an increased debugging demand in the code of colleagues who heavily rely on LLM-based tools, and difficulties with the integration in larger infrastructures.

LLM-based support tools, such as @cursor.com.web.brid.gy and Github Copilot, do a great job supporting web developers in their work. Development proceeds faster and feels more satisfying, according to research at our lab that was led by Thomas Weber and @svenmayer.bsky.social.