Long-Form Standalone Thesis Video — Plan
Parallel-track to the GCC2026 conference talk. Shares assets but has its own timeline and audience. The talk is the trailer; this video is the feature.
Status: planned, not started. Timeline below is the draft — adjust as conference talk lands.
Audience
People who land on jmchilton.net or a paper page, who want the full reproducibility-in-age-of-agents argument and don’t have a conference seat. Expected duration: 12–18 minutes (vs. 10 for the talk).
Goals (where this differs from the talk)
- Develop the thesis fully — 4–6 minutes on the “automating the reproducibility crisis” argument with examples, counterexamples, and the specific Galaxy claim. The talk gets 2 minutes; this gets the room.
- Deeper Foundry segment — pull in the compiler framing from
vault/papers/foundry/index.md(Patterns / Molds / Pipelines / Schemas / Casts / provenance). The talk only gestures at this. - Reuse the 3-minute case-study clips verbatim for demo 1 and demo 2 where useful — or film extended cuts (5–7 min each) if the short cuts feel insufficient.
- End with the three papers as concrete next reading.
Shared assets with the talk
| Asset | Reuse strategy |
|---|---|
| Demo 1 video (3 min) | Embed verbatim, or extend with biological discussion. |
| Demo 2 video (3 min) | Embed verbatim, or extend with Foundry internals reveal. |
| Slide visuals | Reuse the title beat and synthesis diagram; rebuild thesis section. |
| MRSA application | Same application; can use the same recordings. |
| Foundry visualization (if it materializes) | Reuse as recurring B-roll. |
Independent assets
| Asset | Status |
|---|---|
| Extended thesis narration | TBD — write source script in this directory. |
| Galaxy-history-of-reproducibility recap (~2 min) | Maybe. Decide if scope creep. |
| Foundry internals walkthrough | TBD — co-develop with Foundry paper drafting. |
| Long-form intro/outro | TBD. |
Loose timeline
| Phase | Timing relative to conference |
|---|---|
| Pre-talk: storyboard outline + extended-thesis script draft | -2 weeks |
| Talk day: confirm reused clips work | day-of |
| Post-talk: record extended thesis + Foundry internals; cut and edit | +0 to +6 weeks |
| Publish: alongside paper preprints if possible | TBD |
Distribution
- YouTube (primary).
- Embedded on each paper’s index page (notebooks, gxwf, foundry).
- Embedded on jmchilton.net
/talksor/paperssection. - Linked from the GCC2026 conference deck’s final slide.
Open decisions
- Single narrator (jmchilton) or interview-style cuts with collaborators?
- Tie release to one paper preprint or all three?
- Captions / transcripts published to jmchilton.net for SEO and accessibility?
- Is the long video itself one of the artifacts cited in the papers?