Skip to content
refs
REF-1542Source onlineBlueprint rev 01

A Farmer Automates His Greenhouse

A cinematic farmer case study grounds an AI workflow in a real greenhouse mechanism, then ends with a candid laugh.

Maker @OpenAIFor OpenAIOriginal post ↗
Cut timeline · 42 shots02:23
00:0000:3001:0001:3002:00

Field notes / why it works

The 11.9-second uncut field opening lets a real operator state a concrete tracking problem before the product enters. Motor, wiring, and phone shots from 0:51 to 1:15 prove the automation physically, while the 1:23 workflow shift adds a second payoff. Replay peaks near 0:13, 2:02, and 2:18 align with the field premise and the candid chicken exchange.

01

Plate 01 / The format in one breath

A working professional explains a frustrating, specific job while already standing in the real environment. The film follows the person from inexperience and manual work through a tangible build, then shows the machine working and lets the subject describe how the new tool changed the process. A brief human joke after the main payoff keeps the brand story from ending as a sales claim.

Why it works

  • The first 11.9 seconds stay on a single shot of the subject speaking into a phone outdoors; the promise is grounded in a real field problem before any title or product claim.
  • At 0:51–1:15, motor, control box, phone, and greenhouse footage give physical proof of the claimed automation. The story shows a working mechanism rather than relying on a screen alone.
  • The result broadens at 1:23–1:55 from the device to the subject's ability to build it with less back and forth; the closing chicken exchange at 2:01–2:19 adds personality. Replay peaks occur near 0:13, 2:02, and 2:18.
02

Plate 02 / Format card

Platform · aspect · lengthYouTube · 16:9 · 2:23 (143 s); measured source 640×360 at 23.98 fps
Pace42 shots; 41 definite cuts; 17.2–20.1 possible cuts/min; 25.6 picture changes/min including 20 within-shot changes; mean shot 3.4 s, median 2.62 s
ScriptJapanese first-person interview with English translated subtitles. Machine transcript counts 515 tokens and 216 tokens/min overall, but Japanese segmentation and the corrupted final 27 s make this an unreliable spoken-word rate. Write for natural interview delivery, not this number.
VoiceThoughtful working professional, medium-low register, conversational and quietly animated; an assistant voice appears in the final exchange.
On screenRoughly 40% person at work/interview, 45% location and mechanism, 10% computer/phone proof, 5% brand card (estimated from the sheets).
CaptionsSmall white sentence-case translated subtitles, one or two lines in a tight black lower-center box. Larger white title text appears only at the chapter-like title and cover.
SoundContinuous, subdued tonal/cinematic bed is indicated by measurements (inferred); clear field dialogue and natural location sound; no consistent cut SFX. Integrated -15.1 LUFS, true peak -1.1 dBFS.
StructureCold problem → context → practical build → remote-control proof → improved tool workflow → emotional payoff → human coda → 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
  1. Plate 03 / The hook, frame by frame (first 3 seconds)Pro
  2. Plate 04 / Structure (beat sheet)Pro
  3. Plate 05 / Shot grammarPro
  4. Plate 06 / Visual style guidePro
  5. Plate 07 / Voice and deliveryPro
  6. Plate 08 / Sound designPro
  7. Plate 09 / PackagingPro
  8. Plate 10 / Script templatePro
  9. Plate 11 / Production planPro
  10. Plate 12 / PromptsPro
  11. Plate 13 / Edit recipePro
  12. Plate 14 / QA checklistPro
  13. Plate 15 / Make it yoursPro
  14. Recipe / runnable promptPro