A robot learns through staged real-world tests
A 4:30 launch explainer uses a warm stage, embedded proof screen, and escalating physical tasks to make an AI capability tangible.
Field notes / why it works
The claim appears on a nearly empty screen in the first second, then a staged product reveal arrives at 0:04 and speech waits until 0:16. The proof escalates from a small object task around 1:28 to an unfamiliar fixture at 2:34 and a changed-environment test after 2:53; a recurring wood set and inset monitor make those long demonstrations visually coherent across 24 shots.
Plate 01 / The format in one breath
A quietly staged product announcement explains a technical capability while showing it work in a warm, theatrical test room. A researcher archetype speaks directly to camera, then two escalating tasks and a changing-environment sequence turn abstract claims into visible evidence. The screen embedded in the set shows a second, close view of the action.
Why it works
- A tiny autonomy claim sits alone on a blank beige screen for the first 4.3 seconds; the first spoken product claim waits until 0:16.6. That unusual restraint makes the first live-action reveal feel deliberate.
- The proof progresses from a small object task around 1:28–2:17 to an unfamiliar toilet-seat task around 2:25–2:53, then to unpredictable settings around 2:53–3:56. Each example raises the burden of proof.
- The recurring wooden stage and embedded monitor keep the visual explanation coherent across 24 shots and 54 within-shot picture changes.
Plate 02 / Format card
| Platform · aspect · length | X · 960 × 720, 4:3 · 4:29.7 · 23.98 fps |
| Pace | 24 shots; 23 definite cuts, 5.1–6.0 cuts/min with uncertain changes; 17.1 total picture changes/min; mean shot 11.24 s, median 8.25 s |
| Script | 512 spoken words; 125 wpm across the whole film, about 191 wpm while speaking; 25 sentences averaging 20.5 words; no questions |
| Voice | Low-register, animated, precise researcher-presenter; direct-to-camera explanation with falling sentence endings |
| On screen | Roughly half researcher/interview and half staged robot/proof footage (estimated from the shot sheet); occasional filmed monitor close-ups |
| Captions | No running subtitles. Tiny white centered opening statement; sparse lab label, task label and lower third. |
| Sound | Quiet tonal music-like bed throughout; no consistent cut effects; measured −21.9 LUFS integrated, −3.8 dBFS true peak |
| Structure | Silent thesis → announcement → mechanism → simple task → unfamiliar task → environment stress test → qualified future claim → logo; no spoken CTA |
Pro / 13 plates + recipe
The first plates are open. The full dissection is Pro.
Get the timed shots, type and sound specs, and the runnable recipe.
Open the full blueprint- Plate 03 / The hook, frame by frame (first 3 seconds)Pro
- Plate 04 / Structure (beat sheet)Pro
- Plate 05 / Shot grammarPro
- Plate 06 / Visual style guidePro
- Plate 07 / Voice and deliveryPro
- Plate 08 / Sound designPro
- Plate 09 / PackagingPro
- Plate 10 / Script templatePro
- Plate 11 / Production planPro
- Plate 12 / PromptsPro
- Plate 13 / Edit recipePro
- Plate 14 / QA checklistPro
- Plate 15 / Make it yoursPro
- Recipe / runnable promptPro