Carmen Heger’s Post

bring your serial storytelling to the next level:

Interesting piece by Prof. Christian Pieter Hoffmann on diversity of perspectives in #journalism: ➕ What (-) examples of journalistic errors have in common is that they fit all too comfortably into the worldview of an urban, cosmopolitan and progressive milieu, a milieu that is increasingly dominating the journalistic profession. Study after study - whether in Switzerland, Germany or the USA - show that the professional field of journalism is politically shifted somewhat to the left compared to the general population. ➕Why is this homogeneous political composition of the professional field of interest to the phenomenon of journalistic errors? The answer lies in the blind spots associated with ideological homogeneity. ➕Politics or political ideology can be a relevant marker of one's own identity and group membership. This is more pronounced among those with higher education and those with a strong interest in politics: characteristics that apply to many journalists. ➕ Ideological homogeneity is (-) an enormous threat to the functioning of epistemic institutions. The established security mechanisms are not immune to it. ➕ Anyone who wants to avoid mistakes in journalism, especially systematic ones, will not be able to ignore the value of diversity - especially a diversity of perspectives. No training and further education, no professional standards, no fact checking, no award ceremony can protect against the dangers of too much homogeneity in collective thinking. Food for much thought👇

I am - again - in Strobl at Universität Wien Summerschool - which is wonderful, https://shs.univie.ac.at/. I am going to give the first lecture in 20 minutes, Again, I have the privilege to teach Law and Information Society in Europe. As an opener, I ask students to read these 3 texts: 1. https://lnkd.in/eQTKX63t 2. https://lnkd.in/eBb9WDjb 3. https://lnkd.in/e4fWipqQ I asked ChatGPT 4 to summarize the second one. This is the outcome: https://lnkd.in/eJeF6piu It took GPT4 five seconds to write this. Where/how am I better? And what does this mean for my profession?

EmotionAI is pseudo-science & should have been banned in high risk sectors such as criminal justice. AIA is not perfect, but much better than no regulation. AIA is not set in stone, delegated acts can improve legal protection. Let's work together to make this framework the best possible version it can be. My interview Handelsblatt Josefine Fokuhl Luisa Bomke Jürgen Klöckner Philipp Hacker Miika Blinn Maximilian Funke-Kaiser https://lnkd.in/ekvT35G6 Oxford Internet Institute, University of Oxford

When it comes to responsibilities in this digital era, transparency 🔍 is non-negotiable—a point that Nina Ireen Haller is making crystal clear. ✨ In her new absatzwirtschaft column, 👩🏼💻 our Managing Director, Data & Digital Media, DACH argues that policymakers, developers, advertisers and consumers alike must prioritize transparency, fairness and accountability. ✅ Especially in the wake of AI systems becoming ever bigger and more powerful, Haller writes that we have regulate digital communication tools. 🔐 “With all this technological progress, the ethical and economic implications and the social responsibility that come with it must be taken seriously,” says Haller. “We are at a critical point. In a networked world where algorithms often work opaquely in the background, these principles should be at the forefront. We must advocate for transparency and sensible regulation.” Read more in German 👇🏽

Recently I was chatting with a colleague about making corporate communications more inclusive. This article has made me think about the tools we use. #ResponsibleBusiness

😱 Be careful with what you post online, especially photos of your children. Shocking, but could be real. 💀 Our children's data needs special protection. https://lnkd.in/diK5nja9 #AI #data

Really thrilled and grateful that Maximilian Henning of German newspaper Der Tagesspiegel devoted an entire article to my newly published paper on AI liability, summarizing it better and more vividly than I could ever have done. In the piece, I have commented on some of the main ideas of the article as well, such as the need to coordinate the AI Act and AI liability; the need for legal certainty and proper harmonization at the EU level; support for SMEs; and possible consequences for the European AI ecosystem. Link: https://lnkd.in/esfassg6 If you need a PDF, send me a PM. The original article can be found here: https://lnkd.in/eCx2eUpW cc: #ai #aiact #ailiability #ml #responsibleai #gdpr #itrecht #itlaw #eulaw

💡 Serendipity, Open Data and Recommendation Systems for Academic Discovery Platforms 📣 New publication: Serendipität und Open Data. Richtlinien für die Implementierung von Empfehlungssystemen auf Entdeckungsplattformen",  Bibliothek Forschung und Praxis, 2024.  https://lnkd.in/dGPeivsH 📖 Recommendation systems are everywhere – from the platforms we use to shop and stream to the tools researchers use to find relevant studies. While these systems are incredibly helpful, they can also limit our chances of stumbling upon unexpected but valuable discoveries, als know as #serendipity. With the ever-growing number of scientific publications & research data, researchers and students often spend a lot of time searching for the right resources. Discovery platforms help by aggregating data and making it easier to find through recommender systems. However, these systems can introduce biases that affect research choices and, in the long run, innovation. At the end of the CAS in #DigitalEthics tha I attended at the HWZ University of Applied Sciences in Business Administration Zurich, I spent some time exploring the implications of recommendation systems for the discovery process in digital environments for academic research. In particular, I focused on the implementation of a recommender system for the #OpenDataNavigator (https://lnkd.in/dnARPcJc), the discovery platform developed by Switch as part of the #Connectome Project. This project aims to facilitate the sceintific exploitation of different types of #Opendata from non-academic domains by linking them to academic outputs. The paper argues for the importance of maintaining serendipity in virtual environments to foster innovation, and provides guidelines for the use of recommender systems in scientific discovery platforms. By ensuring that we can still make unexpected discoveries, we can keep research fresh and innovative! I would like to thank Cornelia Diethelm Andrea Scheller Norman Backhaus Christoph Bräunlich Kurt Baumann Sebastian Sigloch Tobias Schweizer for their insightful feedback. #DigitalEthics #RecommenderSystems #Serendipity #OpenData #ScientificResearch

Deutsche Telekom's campaign, #ShareWithCare, leverages AI to shed light on the often-overlooked perils of online sharing. Their aim is to advocate for the responsible handling of personal data, with a special focus on safeguarding the images and videos of children. Here they explore the concept of "sharenting" - the common practice of parents sharing their children's photos and videos on social media. The campaign's video might be a little unsettling for some comes with a discretion advisory. However, it's commendable how Deutsche Telekom has explored a topic that few brands dare to tackle. Parents sharing their children's milestones and achievements is a natural expression of pride and something most of us have done at various times, however, it’s crucial to be aware of the potential long-term implications. In an era increasingly dominated by AI, the images we share today could have unforeseen consequences for our children's future privacy and security. Hats off to Deutsche Telekom and creative agency adam&eveBerlin on a very clever and thought-provoking campaign. Let's all strive for a more responsible and informed approach to sharing online. #DigitalResponsibility #AI #DataPrivacy #ParentingInDigitalAge https://lnkd.in/esMedQ6V