What is your professional background and what is your primary research area?
"I graduated from the Aesthetic Studies programme at the Barracks at AU with a BA in Comparative Literature and a MA and subsequently a PhD from Aesthetics & Culture. Before that I studied Practical Philosophy in Gothenburg, but I never got credit transfer for it. I belong to the part of Aesthetics & Culture that is more sociological, anthropological and social science orientated, which is also reflected in my two main research fields.

Firstly, for many years I have been interested in how and when our ideas about creativity as something specifically human have emerged historically, under what influences - or even been invented - and how they have changed.
Initially, my primary interest was related to our notions of work, for example in terms of what could be called the ‘creatisation of working life’, i.e. how the expectation of creativity in working life has been highlighted as a value, first by employees and later also by employers, and what this might reflect. More recently, my interest has turned to how these historical ideas of creativity now need to be rethought in the light of AI, not least with the newer generative forms.
My other major field of research is democratic theory, especially in light of the posthuman, more-than-human ‘turn’, i.e. how democratic theories can embrace those phenomena that do not fit into the notion of democracy as an exclusively human practice. All those phenomena that tend to remain silent. Again, AI, especially the language-based generative AI, has become a game changer, which is what my current research is about."

What is your ongoing AIAS-SHAPE research project about?
"My SHAPE project at AIAS is about what happens to democratic dialogue when we are no longer alone in language - or perhaps more accurately: when generative AI based on large-scale language models makes it clear that we have never inhabited language alone. For a long time, various technologies have been weaving themselves into our language use relatively unnoticed. 

The project, which I call ChatDemocracy, has two parts. The first is a theoretical desktop component that is based on the question: What AI, and what use of what kind of language, for what kind of democracy? In other words, I try to theorise how different types of AI fit into different forms of democracy - e.g. election-centric vs. speech-centric, formal/narrow vs. informal/broad, institutionalised vs. everyday - and what roles language can play in these contexts. Not least, I explore how the different ways of using language in these democratic practices are affected differently by the different flavours of language model-based AI.

The second part of the project has a more interventionist, action research approach. Here I collaborate with a small external company of open source democracy enthusiasts (like myself), where we develop, test and - hopefully in time - disseminate AI-based software and interfaces as well as workshop formats for digital and democratic education in the education system. This primarily applies to primary and secondary schools as well as further and higher education institutions. We call this project DemBots.’ 

What impact do you expect your project to have on society or your research field?
"The most classical academic part of the project is trying to insist on the nuances of the discussion. I see too many posts in the debate where people discuss what impact AI has on democracy. I would like to see a more informed discussion that doesn't just talk about AI and democracy in the implied singular, as if they were clearly definable, narrow phenomena. Both are dynamic and changing fields of social (and economic) struggle, as is, of course, the role of language in them. For example, it may well be the case that certain forms of AI can be useful or streamline the part of representative democracy where the authorities or the public sector interact with citizens in, for example, case processing, but that it also has costs in terms of citizen involvement, participation, etc.

Conversely, it may also be that some forms of AI can facilitate our conversations and thus potentially have positive effects on social cohesion and the like, without necessarily providing solutions to other societal problems (if one were to subscribe to the notion that problem solving is the raison d'être of democracy). These nuances are essential - but unfortunately often absent in both the broader public debate and in academia.
The practical intervention part of my project attempts to bring some of these discussions into learning situations, where students have the opportunity to experience very hands-on how language models ‘talk to’ and intervene in democratic dialogue processes and thus ultimately also in how we experience each other, ourselves and the world.

In my view, perhaps the most important discussion that we should always make an effort to revisit in democratic societies is centred around the question: How and by what means (linguistic, technological, etc.) could democracy be practised? (i.e. as opposed to how we do it now). Not that I imagine that in these workshops we can find definitive answers to this, but the discussion is not only worth having, it is also important as an entry point to life as democratic citizens."

Read more about Jan Løhmann Stephensens' research project here.