Text-first dissemination strategy
This curriculum is designed to be discoverable and useful without relying on lecture recordings or a personality-led video channel. Its public teaching record will be built through concise written explanations, reproducible demonstrations, citable papers, GitHub releases, and the GitHub Pages site.
Public outputs
| Output | Primary purpose | Evidence standard |
|---|---|---|
| GitHub module | Teach one engineering-AI capability | Runnable route, baseline, evaluation, limitations, and reuse boundaries |
| GitHub Pages guide | Help beginners and adopters navigate the material | Short path, explicit prerequisites, accessible language, stable internal links |
| LinkedIn article/post | Share a timely teaching insight or case | Link to a durable repository page; label proposed versus tested claims |
| arXiv or peer-reviewed paper | Report a curriculum design, implementation, or study | Methods, evidence boundary, limitations, and reproducible public artifacts |
| Case-study release | Translate a discipline problem into shared ML learning | Contributor-reviewed data rights, variables/units, baseline, and failure mode |
Writing pattern for public posts
- Begin with a real engineering learning problem.
- Show one compact artifact: a plot, error case, checklist, short code fragment, or decision diagram.
- State what evidence the artifact does and does not establish.
- Link to the maintained module, reference, or case page.
- Invite a bounded contribution: a data card, a failure case, a discipline adaptation, or a correction.
What not to optimize for
- Do not publish recordings merely to imitate established ML educators.
- Do not use attractive AI-generated output as a substitute for a reproducible lesson.
- Do not represent exploratory modules as validated instructional interventions.
- Do not put instructor-held assessment items, student records, private data, or sponsor-restricted material in public posts.
Sustainable cadence
Publish when a maintained artifact changes: a new module, a validated case study, a dataset card, a physical demonstration protocol, a paper/preprint, or a substantial correction. The repository remains the canonical record; social posts are discovery pathways.