Bartninkai is a small place in southern Lithuania, the kind of village where days can pass without much interrupting the quiet. For one week in November it held young people from six countries, all there to take the measure of artificial intelligence. Our participants were among them, sent north to join groups from Lithuania, Portugal, Slovakia, Czechia and Serbia.
Erasmus+, the European Union’s youth-mobility programme, co-funded the week.
Setting the Ground Rules
The first day did none of the AI work. It built the group that would do it. A Name Story Circle and a long round of Mingling Bingo got the whole room talking, and a session called Me and the EU walked everyone through what Erasmus+ actually offers a young person before the week leaned on it. The point of the opening was less the content than the room: by the evening, the separate national groups were starting to behave like one.
Two sessions set the terms for everything after. In the Social Contract, the participants wrote their own Code of Conduct, a shared set of expectations they agreed on rather than received. In Setting the Stage, they pinned their hopes, contributions and worries for the week onto the wall and sorted them into the open. Naming the fears early is what made the harder material later feel safe to handle.
Learning to Question the Machine
The middle of the week was built around what AI gets wrong. In Curated Digital Feeds, teams took on invented personas and watched how a social media feed quietly reshapes itself around a stereotype, feeding back whatever it decides the persona wants to see. From there the group moved to the week’s sharpest exercise: Breaking News, where six teams each took a documented case of AI causing real harm and rebuilt it as a news broadcast, scripts, anchors and all. One team’s report on a drone that turned on its own operator gives the flavour.
AI-Enabled Drone Opted to Attack Its Operator
After the broadcasts came The Human Code, a session on how AI can be turned to social engineering, assembling a convincing approach out of scraps of public information. Watching the reports back and running through the exercise, the group reached a blunt shared conclusion: a system with no sense of right and wrong has no business being handed decisions that carry weight.
From Critique to Creation
Having taken AI apart, the group started building with it. In The AI Brain, teams researched how the best-known tools are trained and what that training does to their answers, then put the same odd prompts to each system to compare how differently they replied. Utopia.AI flipped the exercise: each team designed its own ideal AI, choosing its training data, its abilities and its limits, then produced a one-minute promotional video to sell it. The Impostor Game kept everyone honest, hiding one AI-generated argument inside each team’s debate and daring the others to spot it.
The thread running through all of it was judgement. The tools were never treated as magic or as menace, but as something powerful enough to deserve a critical eye. By the closing days, participants who had opened the week criticising AI were commissioning it for their own ends, and finding that the line between the two is exactly where the skill lives.
The Hours Between the Sessions
Plenty of the week lived outside the workshop room. Each country took an evening to present itself through its food, its music and the stories it brought, and those intercultural nights did the steady work of folding the national groups into one. A Secret Friend game ran quietly underneath the whole exchange, each participant looking out for someone in secret until the reveal on the final evening.
Even the learning had a social shape to it. Early on, teams built an “AI Competence Village” out of the eight key competences for lifelong learning, turning a dry framework into something they made together. The week held because the serious thinking and the easy company were never really separate.
What Travelled Home
The participants left Bartninkai able to do something they could not do on the way in: reach for an AI tool and know, in the same moment, whether to lean on it or push back. Spread back across the continent at once, that judgement is exactly the kind of result this network sets out to create. The friendships made along the way travel with it.
Photo Gallery
A few more moments from this activity :)
Check these and more photos in the activity’s photo album.
