Built for practice, not promises.
Independent, open practical AI learning. Read a worked example, make an artifact, test it and keep a limitation you can explain.
Published 21 October 2026 · current independent pilot
What is available
Six foundation build paths contain 36 guided activities. Fifty additional illustrated activities cover prompting, data, verification, design and creative workflows. The library includes original templates, fictional inputs, handbooks and downloadable practice packs.
What this is not
Open Practice is non-accredited. Self-reviews are learner reports, not proctored assessment, a qualification, employment evidence or an independent evaluation of mastery. Time estimates and teaching materials have not been independently learner-validated.
Open materials, honest boundaries
You can read and download public materials without an account. Hosted AI services may require their own accounts, keys, credits or payment. Manual fixtures and design-only routes are labelled; they are not presented as live model runs. Illustrations are AI-generated editorial artwork with human-authored instructional labels, not provider screenshots.
Who maintains it
This project is maintained through its public source repository. No institutional endorsement, registered-company status, university affiliation or partner relationship is implied. See the contact page for the current public feedback route.
Follow the work
Who this is for
People across regions underserved, excluded or poorly supported by mainstream AI and payment providers—not one country. We teach legitimate available cloud routes, downloadable original materials, honest manual alternatives and explicit access limitations. Country, language, identity, cost and connectivity constraints vary; we do not promise universal provider access or help bypass restrictions.