Digital Methods Summer School and Data Sprint 2025 Media Studies, University of Amsterdam 30 June - 11 July 2025 Call for Participation The Digital Methods Initiative (DMI), Amsterdam, is holding its annual Summer School on 'Social media at a crossroads, and the sensitivity of AI platforms'. The format is that of a (social media and web) data sprint, with training tutorials as well as hands-on work for telling stories with data. There is also a programme of keynote speakers. It is intended for advanced Master's students, PhD candidates and motivated scholars who would like to work on (and complete) a digital methods project in an intensive workshop setting. This year’s Digital Methods Summer School has two related themes. The first is ‘social media at a crossroads’ and the second the ‘sensitivity of AI platforms’; they are related given that mainstream social media seem to have lost their normative compass while AI platforms strive to improve theirs through enhanced content moderation. AI platforms are now occasionally called ‘oversensitive’, however. This oversensitivity charge is reminiscent of attacks on Dorsey’s Twitter and other social media platforms with active content moderation. How to consider the rising critique of the oversensitivity of AI platform moderation in light of the turn to free-for-all expression on social media? The Summer School takes up this larger question, but first breaks it down into the two individual areas of study. Social media at a crossroads Does the current crisis in social media seem novel? Fake news, data breaches, trolling, neo-reactionary takeovers and other style shifts and infrastructural setbacks have long overtaken the participatory culture, platform cooperativism, produsage, neo-pluralistic potentials and other more buoyant notions from past digital cultures. But when Meta announced in January 2025 it was re-orienting its content moderation policies to filter only the most egregious violations of its standards and also end its fact checking program, it perhaps marked a breaking point with respect to any prospective redemption of social media. Dressed up in masculine vitalism and aesthetics, Meta’s announcements follow on from the trenchant criticism directed at X/Twitter. The EU has found it in breach of its regulations by using ‘dark patterns’ in advertising and blocking researcher data access, and a recent Berkeley study found that since Musk’s takeover hate speech rose by 50%. And yet, Meta headily pointed to X’s content moderation style as its own guide! Will Instagram and other social media platforms follow X’s slide, empirical questions we would like to explore through auditing and other projects at the Summer School. The sensitivity of AI platforms The second theme is the so-called sensitivity of AI guardrails which refer to the strength level of the moderation and filtering built into AI platforms. AI chatbots, trained to be 3H’d (helpful, honest, and harmless) can be sycophantic and overly polite. At the same time these platforms, while racing for new performance levels, continue to build content moderation APIs and other in-built measures to filter out a variety of harmful and offensive content. The filtering work may be easier to perform (compared to social media platforms) given that it’s sentence completions rather than user posts that are affected. Users are not cancelled, their posts are not shadow banned. (But they do experience refusals to complete prompts.) Will the moderation backlash experienced by social media platforms come to AI platforms? How to attune AI platforms to the critique? Content moderation research continues to find porous guardrails on AI platforms. While Llama appears to be highly sensitive, Falcon seems to be more permissive, for example. Will AI platforms follow social media and position themselves on the free speech spectrum? Given the challenges of universal moderation speech norms, will they turn to personalisation as the answer? These and similar questions about the state of content moderation online motivate the Summer School. Applications: Key Dates There are rolling admissions, and applications are now being accepted. To apply please send a letter of motivation, your CV, a headshot photo, 100-word bio as well as a copy of your passport (details page only) to summerschool [at] digitalmethods.net. Notifications of acceptance are sent 1-2 weeks after application. Final deadline for applications is 19 May 2024. The full program and schedule of the Summer School are available on or about 27 May 2025. Tuition Fees The fee for the Digital Methods Winter School 2025 is EUR 895, and upon completion all participants receive certificates of completion or transcripts (worth 6 ECTS). Full call for participation and all additional information is here: https://www.digitalmethods.net/Dmi/SummerSchool2025 Prof. Richard Rogers Media Studies University of Amsterdam