A First Toolkit for Exploring AI with Young People
Dernière mise à jour : 8 oct.
After several months of work, this initial phase has already produced a substantial body of shared resources. Since the launch in La Rochelle, we have been building a toolkit designed to be reused and adapted in accordance with the licence specified for each resource. Every element addresses a practical project need: How can we explore generative AI with young people? How can we assess what participants learn? And how can other teams build on what we have started?
A shared teaching methodology
We formalised our methodology around a six-phase scientific process designed to help young people adopt the mindset of researchers investigating AI. It begins with observing and documenting AI in the world around them (P1), followed by developing hypotheses through unplugged activities (P2), experimenting with language models (P3), validating ideas through robotics activities (P4), reflecting on societal and ethical issues (P5), and finally communicating and showcasing the projects developed by young people during the first five phases (P6). Each stage is linked to specific competences from our skills tree, examples of adaptable activities and key considerations drawn from field experience.
The document is now available in English and in all four project languages on our website: youth-ai-lab.eu.
The Youth AI Lab approach is structured around six complementary phases.
Activity idea sheets for designing workshops
The activity idea sheets present activities that are easy to understand, adapt and test with different audiences. They will gradually be supplemented by more detailed protocol sheets enriched with feedback from the field.
The activity idea sheets are available in English. On the wiki, automatic translation will make them accessible in all the project languages. Protocol sheets, meanwhile, will be provided directly in the local language in which each activity is implemented. Activities for the later phases will be added as they are published.
The activity idea sheets currently available are listed below according to the different phases of the learning journey.
Observe AI in the world around you
Keep an Eye on AI: An easy-to-adapt field investigation that helps a group identify AI in its surroundings and distinguish artificial intelligence from automation.
AI Lab - Inside the Machine: Three screen-free games that use hands-on experience to show how a machine identifies patterns and makes predictions.
Looks Right… Is It Right?: A fact-checking activity that develops critical thinking about convincing but sometimes false AI-generated answers.
Illusion Is All That Matters: A guided debate questioning what it means to “think” and opening a discussion about the uses of AI and robotics.
Develop hypotheses about how it works
AI Detectives: How Do Machines Learn?: A simple hands-on exercise showing how the choice and quality of examples influence what a machine learns.
Bio-inspired Algorithms: Reinforce Your Learnings: A playful, physical simulation for understanding reinforcement learning before programming a robot.
Bias: Genre Matters: A practical activity that reveals a model’s gender stereotypes and teaches participants how to observe, measure and discuss them.
The Bias Bounty Hunt Challenge: An engaging prompting challenge that turns the search for bias into a group investigation and encourages participants to refine their questions.
Who Does NOT Want to Be My Business Partner?: A creation and presentation game exploring AI’s persuasive power and the importance of maintaining a critical perspective.
Make Cities Great Again: A project combining robotics, sensors and the sustainable city, ideal for connecting AI with tangible local issues.
Experiment with models and data
From Data to the Machine’s Brain: A progressive learning journey in which participants train a model, observe its errors and gain a tangible understanding of how an artificial neuron works.
WikiYouthBot: A collaborative sequence guiding a group from researching sources to creating its own chatbot with expertise in a local topic.
Validate through design and robotics
Designing an AI with a Robot: A design framework that guides young people from a sustainable-development challenge to a robotic prototype capable of recognising, deciding and acting.
The collection will continue to grow over the coming cycles, particularly with activities dedicated to the reflection and communication phases.
Protocols shaped by field experience
When a workshop idea is tested in the field, it is then documented in a more comprehensive format: the protocol sheet. This expands on the original idea sheet with a detailed step-by-step plan, feedback from the partners who tested it, adjustments for different audiences and contexts, and related teaching materials such as printable templates, games, exercises and tutorials. Together, these sheets will form the compilation of exploration protocols scheduled for completion at the end of the project.
Assessment at multiple levels
The assessment framework covers the entire programme and every project audience: pre- and post-workshop questionnaires for young researchers, mentors and educators, supplemented by self-assessment tools used during the labs, implementation logs for facilitators and frameworks for analysing individual learning pathways. This multi-audience, multi-phase approach brings together different perspectives and makes it possible to document the programme’s impact in detail.
A multilingual wiki in development
A larger initiative will begin in September 2026: the creation of a comprehensive wiki for educators and facilitators who want to adopt the Youth AI Lab approach in their own setting. Alongside the activity idea sheets, the wiki will provide methodological guidance to help users structure their own lab, recruit and train mentors, design a learning journey and assess its impact. It will be available in all four project languages and will become our main tool for sharing the approach beyond the consortium.
