Time for a quick group shot.
Look at all these wonderful folks! 😍

The breadth of projects presented here, really showcases how much the DDC has grown over the past 5 years.

DDC RESEARCH RETREAT
(Spring edition)
We've returned to Nyborg to meet as a group of DDC scholars and fellows (plus invited guests) to spend 2 days with updates on our projects and deliberate on future pathways.

Thread below⤵️

I will indeed answer your email about the scoping review with the idea of sending you our database which we made for that project ;-)

And of course @Jonas vistrup' s explainable AI and the law project!

The PRORESIL project in which @andre-k-rodarte.bsky.social and I looked at susceptibility to misinformation, a large scale study about the municipal election by @rasmusschmokel.bsky.social, using social media *for* democracy by @hannahdecker.bsky.social - and finally my new #FOODFRAMES project

High-choice platform environments and the spreading of intolerance by @Anne Clausen, AI's role in creativity, imitation, and organisational adaption by @emilbakke.bsky.social , childrens' experience with misinformation and how to empower them- presented by @Patricia Lyk

In todays' @ddc-sdu.bsky.social retreat, we get an update on the projects currently running at the DDC - with insights into how Danes perceive and use AI by @Therese Waalen Sandberg, impact DAOS by @Giorgia Sampo, AI in the news ecosystem by @lyngemoller.bsky.social...

also critically: The "no" campaigners started earlier, where more active - raising the question who takes the digital space for serious (and who did still not get it...)

He then presents a fascinating study on the Australian referendum - where the polls where pretty "pro" before - but the referendum failed. What they found is that "no" campaigners (partisan news + far-right central) where very unified and coherent - the "yes" campaign less so.

Based on these similarities and attributes they then create vectorised account practices and then compare the normalized vecrors to each other (cosine similarity) for each of the vectors (e.g., the "share" or "language" vector) - and then you can aggregate them (e.g., mean, multiply, weighted)💡

Also moving how digital networks can be studied - examining shared languages, beliefs, actions (e.g., on FB via MCL - similarity in link sharing (external links), similarities in on-sharing (internal sharing), similarities in video sharing, similarities in language sharing (word embeddings based)

Based on these symptoms they mapped discoursive practices looking at pattern of engagements, language use, URL sharing, images, videos etc. (doi.org/10.1177/2056...)

He highlights 5 symptoms: (1) break down of communication (talking at/about but not with each other) (2) discrediting and dismissing of information (3) erasure of complexities (4) exacerbate attention and space for extreme voices (5) exclusion through emotions

So what is the polarisation you are interested in? issue-based, ideological, affective, perceived, interpretative, interactional? And what of that is actually a problem? We need some forms of political disagreement. So when does polarisation become destructive/ disfuncational?

Destructive polarization in digital communication contexts: a critical review and conceptual framework While the digital turn in communication research offers novel opportunities to study polarization at scale, it also adds complexity to a challenging concept. Ambiguities surrounding the conceptual ...

He starts with defining and explaining the nuances and facets of polarisation (recent work led by @kathaesa.bsky.social can be found here doi.org/10.1080/1369...)

This years guest speaker is the amazing @snurb.info from @qutdmrc.bsky.social - showing his amazing work on "exploring destructive polarisation" - mapping social media debates about the voice referendum in 🇦🇺.

You can learn more about Lukas work here lgalke.github.io - including about his work on the Danish Foundational Model for LLMs and scalable mechanisms of interpretability for safe & trustworthy LLMs Agents.

two cartoon characters are standing next to a robot hand and the words artificial intelligence Alt: two cartoon characters are standing next to a robot hand and the words artificial intelligence

In the second part of the program our new @ddc-sdu.bsky.social member @Lukas Galke Poech gives insights into his super cool work on neural ecosystems and the AI-agent future - where artificial and human agents increasingly interact with each other - looking at the risks that come with that

In the last fellow pitch, @annaneifer.bsky.social talks about her project on AI Stories - putting on a cross-boarder perspective on governance and lived experiences. She asks: How are humans affected by algorithmic decisions? How can we break down the EU AI act? What's happening in ppls life in 🇩🇰& 🇩🇪

@Ansger Nim will work on DSA enforcement - asking why so few Danish actors actually contribute to enforcing the DSA. The DSA is very much centred around on different actors engaging in DSA enforcement- but very few Danes do support DSA enforcement - indicating a gab in the enforcement of the DSA.

@zeaszebeni.bsky.social studies a pivotal moment in time and question - what do young people in Hungary expect/ perceive/ want from news - how do young hungarians conceptualize "news" after growing up under the media capture by Orban? What do they expect and want now after the political change?

a woman is driving a tractor down a dirt road . Alt: a woman is driving a tractor down a dirt road .

@jasminemcnealy.com gives insights in her fascinating work on rural artificial intelligence - arguing that AI is very much centred around "cities" - but large shares of the population (in the US but also DK) do not live in cities. Super relevant work - also for Denmark as a very agricultural space.

a picture of gollum with the words " my dataaaa " below it Alt: a picture of gollum with the words " my dataaaa " below it

@williamlallen.bsky.social works on data visualisation for democracy - taking us back to the COVID19 pandemic and the explosion of data visualisations (and powerpoints) during that time. He asks the relevant question *why* and *when* some graphs are shown - and when not - data is not neutral.

a woman says we can fix this in a black shirt Alt: a woman says we can fix this in a black shirt

@Emilie Victoria de Keulenaar shares her project on "prosocial" platform design - such as "bridging" users across social, ideological, or informational cleavages, or social provenance designs such as highlighting context behind users opinions

a penguin is standing next to a piece of paper that says " follow the rules " Alt: a penguin is standing next to a piece of paper that says " follow the rules "

@Domenika looks at AI-Act and DSAs are able to regulate deepfakes - and whether they instead conflict with free speech. Super relevant research

@leahellmueller.bsky.social presents her amazing work on gender-based violence including AI-based digital sexual abuse conducted with @unwomen.bsky.social, and shows how her fellowship project builds upon this research by looking at Denmark (soft inequality) and Switzerland (structural inequality)

a digital sovereignty secure your digital future advertisement Alt: a digital sovereignty secure your digital future advertisement

@Timme Bisgaard Munk - works on digital sovereignty monitoring, empowering the Danish public sector and those working in it to evaluate their own level of digital sovereignty - including of discussing "sovereignity washing"

a person is holding a green video game controller that says take control on it Alt: a person is holding a green video game controller that says take control on it

Søren Schultz-Hansen talks about his project with mobile ethnographical work, centering on first-time voters' experience with political content/ perceptions in their digital life world - greatly supported by DDC student assistant @Miriam Riees Diemer

a man is looking at a wall of monitors on make a gif Alt: a man is looking at a wall of monitors on make a gif

@Lucas Braun talks about his project "from legacy to platform - rethinking opinion power for social media monitoring." He compared different attempts to monitor social media for regulatory purposes.

a cartoon of mickey mouse with the words `` great news '' below him . Alt: a cartoon of a jumping mickey mouse with the words `` great news '' below him .

@Heidi Røsok-Dahl from @oslomet.bsky.social talks about her work on changes in news-oriented news: content in the age of dynamic gatekeeping. She argues that journalism has warp threats (what holds journalism together) and woof threads (representing diversity in news)