From Prompting to Making: Workshop on Generative AI Literacy for Vocational Education

Generative AI has entered education with unusual speed. Teachers are being asked to respond to a moving target: new tools, new policy expectations, new assessment questions, and new student practices. In vocational education, AI literacy cannot remain an abstract discussion about “the future of work.” It has to connect with lesson planning, workplace tasks, domain knowledge, professional judgment, safety, collaboration, and the material realities of teaching at vocational schools.

This was the starting point for Makerspace AI 2026, a hands-on professional learning format for vocational educators developed around one guiding assumption: AI literacy is not learned by listening to a lecture about AI. It is developed through experimentation, reflection, making, critique, and shared production.

In preparatory sessions students tried out tools and prompts that were geared towards a reflection of how they seem themselves and how they see generative AI in their future classrooms. They began by reflecting on the kind of teacher they hoped to become, generating an image that represented their emerging professional identity and explaining the qualities it conveyed. They then created short AI-generated videos about what they looked forward to in their future work at vocational schools and produced narrated songs expressing their concerns about entering the profession. Students also documented personal experiences with AI, including moments when it had been especially helpful, frustrating, irritating, or impressive.

Students created their self-image as a teacher with genAI tools

These activities gave students a shared foundation of practical experience for the workshop while revealing that AI-supported creativity is often more iterative and time-consuming than expected. Students tested different tools, refined prompts, discarded unsuitable outputs, and combined platforms to achieve results that better reflected their intentions. Through this process, they encountered both the creative potential of generative AI and its limitations, including unpredictable quality, generic or inaccurate outputs, paywalls, restricted free versions, and the need for careful critical evaluation.

For the workshop portion of the course, 11 preservice teachers and two facilitators gathered at the end of June at the Institute for Vocational Education. the makerspace environment, AI became one tool among many. During the two workshop days it sat alongside cardboard, LEGO bricks, paper plates, podcasting, Scratch, classroom experience, peer feedback, and teacher input. Braving the heatwave that held Germany in its grip was a stark reminder of both the  environmental collateral consequences that AI’s appetite for energy poses and the pressing engineering challenges that AI may help solve.

Reflecting AI Use

The workshop began with a simple question: How, when, and why do education students already use AI during a typical teaching day? Participants mapped their AI use from morning to evening, using paper plates as clock faces. This low-tech prompt created a space for honest reflection: AI use was treated neither as a marker of innovation or nor as moral failure, but as a practice embedded in routines, pressures, shortcuts, hopes, and uncertainties.

Example Student Plate: How often do you use AI? What for?

Envisioning Value in Education

AI conversations can easily become polarized. Some participants are enthusiastic early adopters. Others are skeptical, overwhelmed, or concerned about ethics, environmental impact and societal consequences. A makerspace format creates room for multiple forms of expertise. Participants can express concerns through models, metaphors, prototypes, sketches, and stories, not only through technical vocabulary. The workshop used LEGO Serious Play to address a familiar participation problem: in many discussions, a small number of voices dominate while others remain quiet. LEGO Serious Play changes the communication structure through a “100/100” principle: everybody builds, and everybody shares.

Students building and interacting at eye level

The central LEGO challenge asked: What makes higher education valuable today, and what has become interchangeable? One memorable construction—a cake mounted on car tires—represented the specfic potential of vocational education to connect perspectives that would ordinarily remain separated. The cake and the tires appeared incongruous, yet together they embodied the idea that new possibilities emerge when people move beyond the boundaries of their own disciplines. Other models questioned whether established teaching practices still served a meaningful purpose. Cobwebs represented assignments and routines that had remained in place even though their educational rationale had largely disappeared. One example was asking students to calculate Excel functions manually as homework, even though the same operations would later be completed by software—and the handwritten answers might simply be copied. Which forms of practice genuinely develop understanding, judgment, and transferable competence, and which merely preserve routines from an earlier technological context?

The second major pedagogical thread in Makerspace KI 2026 was design thinking. Participants moved through problem definition, ideation, prototyping, testing, and iteration around the question of how higher education can be designed so that students want to learn and teachers want to teach. Each design team received a cardboard box with material, as well as timed prompts and handouts. The teams identified the problem, explored the solution space through ideation, and built a hands-on prototype.

Students presenting and discussing design thinking prototypes

Making and Marshmallow Challenges: Fail Fast with Fun

Makerspace AI 2026 approached AI literacy through the lens of maker pedagogy. Participants worked with tools such as Makey Makey, Scratch, Doodle Bots and Brush Bots. These making activities surfaced the habits of mind that AI literacy requires: curiosity, iteration, uncertainty tolerance, critique, collaboration, and the ability to move between idea and artifact.

Brushbot Student Project

One of the most powerful ideas in maker pedagogy is that the teacher does not need to know everything in advance. The stance is not, “I have all the answers,” but rather, “I do not know how to solve this yet, but we can work it out together.” That stance is highly relevant for AI literacy. Many educators feel pressure to become instant AI experts. But the more durable professional capacity may be design judgment: knowing how to frame a problem, test a tool, evaluate an output, involve learners, and connect experimentation back to disciplinary and vocational goals.

A recurring theme in the workshop was failure. Makerspaces normalize trial and error. A Doodle Bot wobbles and breaks. Your team’s tower structure falls apart during the Marshmallow Challenge (from ‘tadah to oh-oh’). A podcast sounds awkward on the first recording. These moments are an important counterweight to generative AI outputs that look finished before learners have fully engaged with the problem.

Students completing the marshmallow challenge

Substitute Teaching Challenge

In the workshop, participants engaged in a substitute teaching challenge. Could you cover for a colleague and design an unfamiliar lesson in 20 minutes using AI tools or other resources? The activity was playful, but it demonstrated that AI can generate plausible teaching materials very quickly, while the quality of those materials still depends on human review, subject expertise and contextual awareness.

Sketchnoting AI Literacy

Current AI literacy frameworks offer useful orientation for preservice teachers to reflect on what is important to them. In a sketchnoting exercise, we offered a walkthrough of twelve different competency models. Students recorded elements that resonated with them as sketchnotes. In the discussion with students, we focused on impact on vocational education which points toward a curriculum challenge: AI literacy should be embedded in authentic workflows, not isolated as a generic add-on. At the same time, there was also a desire to protect classroom time and space for tasks and activities that are authentically non-AI.

Student Sketchnote

Toward a Shared Product

The Makerspace AI 2026 culminated in a shared product: a collaboratively developed open resource on AI and teaching. Professional learning often disappears after the workshop ends. A shared artifact creates continuity and offers a space to capture examples, tensions, teaching ideas, reflection prompts, and practical approaches that can travel beyond the immediate student group.

E-Book: Panke, S., Harth, T., Börschel, S., Abrahamczik, L., Bulat, E., Engelbertz, M., Lebbing, E., Loxterkamp, V., Möller, F., Mönninghoff, E., Papenbrock, E., Wiegmann, P., & Wieners, L. (2026). Wenn KI alles kann, warum dann noch unterrichten? EdTech Books

Generative AI in vocational education raises questions about assessment, authorship, labor markets, equity, creativity, deskilling, professional identity, and institutional responsibility. What do apprentices, vocational students, and teachers actually need to know how to do without assistive technologies? Where might AI help them think, communicate, design, document, simulate, troubleshoot, or reflect? Where might it create shortcuts that weaken learning? Where should learners be required to work without AI so that foundational skills are not hollowed out? These are not questions with ready-made answers, but we can approach them in community and offer preservice teachers a voice in shaping the discourse.