A free, open-source curriculum that takes you from a single array element to building, training, and reasoning about modern generative models — across fifteen progressive modules, each producing a visual artifact.
The curriculum is composed in three paths. Each path holds a small set of modules; each module holds three to seven sub-topics; each sub-topic holds a handful of lessons. It's linear by design, but experienced programmers can start at a later cycle. Open a module to see its anatomy.
No lessons match that search.
Pixels, geometry, recursion, simulation. NumPy + PIL only — no networks. Build visual intuition first.
Procedural generation, classical ML, animation, intro neural networks, real-time. The bridge from arrays to inference.
GANs, VAEs, diffusion, language models, integration, capstone. Generative AI in practice.
A recurring one-day workshop that condenses the curriculum into a single sitting. The archive begins with the first cohort below; the next workshop is in planning.
Six modules, one sitting, in Berlin. The first cohort produced the data reported in /research.
The next workshop in the recurring program — one day with Academis, same shape as 01. Registration of interest is open now; date and venue follow.
A group exhibition co-organised with Kristian Rother and Academis: six artists working across printed and code-driven pieces, several built directly from curriculum modules. The archive begins below; the next exhibition is not yet scheduled.
Over two days, 6 artists turned the IT studio Academis office in Schöneberg into a gallery: printed works on the walls, code-driven installations on screens and projection. Several of the pieces grew directly out of curriculum modules — equations, algorithms, and generative models, arranged as objects you could walk between.
The curriculum is the empirical artifact of a design-based research thesis. Each cycle iterates on the same five questions; this is the third.
What pedagogical principles and design patterns effectively scaffold learning progressions from basic array manipulation to generative AI in creative contexts?
How can complex technical concepts be decomposed and sequenced to maintain optimal cognitive load while building toward advanced applications?
What strategies effectively integrate real-time systems with progressive AI learning, and at what points in the curriculum should these integrations occur?
How can learning outcomes in creative AI education be assessed across technical proficiency, creative expression, and conceptual understanding?
To what extent do learners successfully transfer foundational computational concepts to novel creative AI contexts, and what factors facilitate this transfer?
