I'm at the AI and Law conference in Tübingen and have just heard a MARVELOUS keynote lecture by Sylvie Delacroix on preserving productive uncertainty in designing AI systems. It reminded me of (among other things) an idea brough up by @andreaskuczera at #DHd2025 : How about allowing (or incentivizing) usage of AI systems in research only in scenarios where you are not sitting alone in front of the system, but use it exclusively in group settings from the outset?
A glimpse of what Delacroix discussed can be gleaned from: https://doi.org/10.1007/s11023-025-09736-x
Here is the conference webpage: https://ailawinstitute.de/conference-for-ai-and-law/
I guess I may post some more thoughts today and tomorrow with the hashtag #AIandLaw
At half past eleven back in my hotel room I find that I did not manage to post any other news from the #AIandLaw conference. Apologies if you were waiting for something. But actually it's a good sign: I have been following everything so closely, and having so many good conversations that I just didn't manage. Still enthused, I go to bed now. Looking forward to tomorrow 's sessions...