Brave Conversations @WebSci26

Last Friday Ghada Ibrahim, Hannah Stewart and I presented our first Keynote at the 2026 Web Science Conference in Braunschweig, Germany before holding our 32^nd Brave Conversations at the Hausderwissenschaft.

Brave Conversations began as a Web Science project with our Meta conferences from 2008 – 2011 and then evolved into its current form in Canberra in 2017. Whilst some of the format has changed, and we’ve held the event around the world, the focus remains the same – how do we become Smarter Humans in the Age of the Smart Machine. And how do we bring some of the work of Web Science to the broader public rather than have it sit in Academia.

This, our first Keynote, offered us the opportunity to see how an academic audience would respond to a mini-version of the event with some interesting results. As an opening we asked people to turn to the person next to them and ask them how they were feeling on the last morning of the Conference, after the Conference dinner the previous evening which had morphed into Karaoke and camaraderie.  Immediately the conversation began and participants were open and prepared to play with us as we led through the process for the rest of the hour.  We tested a mini-case study which elicited some interesting feedback – firstly that it was not based on enough theory, secondly that it was really pretty one dimensional and based on budgets and money.

In both cases this reinforced my conviction that the Brave Conversations format has a lot to offer the often dry conferences of Academia where people go down their own specific rabbit holes and often don’t see the bigger picture.

The key to this is the Strategic Triangle, based on the work of Professor Mark Moore at Harvard who himself draws on Plato's Models of Persuasion. We use this as a basis for all of our cases combined with a nested system of getting people to think about themselves within a broader system - asking not just how a particular situation can be changed or problem solved - what can we do? (the Logos) but also to explore the ethical and moral aspects of any decisions made - what should we do? (the Ethos) and finally to consider issues of power and agency within the 'authorising environment' - what may we do? (Pathos).

Most people seem to rush to a solution focusing on the 'logos' - "problem solved!" Some may get stuck on the 'ethos' - the moral and ethical dimensions. Very rarely to people naturally go to the 'pathos' - the issues of power and authority which brings in the broader system in its complexity.

Even though we only had 30 minutes for the 'case' in our Keynote it didn't take long before the logos and ethos began to dominate, and people had only just really started digging when we had to finish. I felt we could have had another hour to do the session justice but in the hour that we had I do think that perhaps we did get some people thinking a little differently.

In our actual Brave Conversations event later in the afternoon we had more time and for this we rolled out the Simulation Game which Hannah Stewart had developed for Stuttgart in 2025.  Even though we only had a small audience the Game ran as we had hoped with a degree of chaos and confusion combined with the a focus on the human-human and human-machine interaction.

As with the other times we've run the game there are some observations to note:

·     Each of the Large Language Models people used had slightly different 'personalities' - some were more forthright, others less so. What was most interesting was how people responded to each of the models and how it impacted their behaviour.

·     Whilst we only ran 3 iterations of the Game (we often run 4 to 6 depending on the number of people in the room) it was already apparent that the humans were following instructions and there was very little push back or dissent. One team's objective was more collaborative but despite instructions to try to work with the other team the competitive instinct was more dominant in the other two teams and collaboration was difficult.

·     We didn't have time for a lengthy reflection and the weather was very hot but the conversations that did emerge was instructive in terms of the power dynamic between the different groups and between the humans and their 'machines'.

The one thing we know about Brave Conversations events is that it does get people thinking and we very much hope that the conversations in the room continued one way or another beyond the boundaries of the event itself.

A huge thanks to Jeremiah Othman of the Haus der Wissenschaft Braunschweig who hosted us so beautifully, and to the organising committee of WebSci26 who invited us along.

‍