Skip to content
refs
REF-2243Source onlineBlueprint rev 01

Three Browser Jobs, One Assistant

A silent three-part browser demo proves document synthesis, slide edits, and site building through visible finished work.

Maker @anthropic-aiFor AnthropicOriginal post ↗
Cut timeline · 9 shots01:35
00:0000:1500:3000:4501:0001:1501:30

Field notes / why it works

The first concrete multi-source request is sent around 0:12–0:13, the video's strongest replay peak, giving the viewer a specific job to judge. Each workflow shows an input and a finished artifact: populated document by 0:27, slide edits by 1:06, and a tested website by 1:27. Chapter cards at 0:30 and 1:10 reset attention within a 95-second demo that has only nine top-level shots.

01

Plate 01 / The format in one breath

A silent, widescreen product film shows an assistant doing three distinct browser jobs: synthesizing dashboard data into a document, applying presentation feedback, and building and checking a website. A hand-drawn browser icon introduces the idea; real interface states and finished artifacts carry the proof. There is no narrator to interpret the screens for the viewer.

Why it works

  • The question of what the assistant will do opens in the first 4.8 seconds, then a concrete multi-source request appears around 0:12–0:13. That send moment is the strongest replay peak.
  • Each job ends in a visibly changed artifact: a filled document by 0:27, resolved slide comments by 1:06, and a rendered website by 1:27.
  • Two plain title cards at 0:30 and 1:10 reset attention and make a 95-second, low-cut demo feel like three short examples.
02

Plate 02 / Format card

Reference
Platform · aspect · lengthYouTube · 16:9 · 1:35 · source measured at 640 × 360, 30 fps
Pace9 shots, 8 cuts, 5.1 cuts/min; 14 internal steps, about 13.9 total picture changes/min; mean shot 10.56 s, median 9.23 s
ScriptNo detected speech; meaning comes from brief title cards and readable interface prompts/results. WPM does not apply.
VoiceNone; the tiny “1.” at 0:04.3 in automatic transcription is not credible dialogue against the silent visual grammar.
On screenAbout 8% drawn opener/extension close-up, 87% desktop UI and artifact screens, 5% beige chapter cards (estimated from shot durations). No presenter.
CaptionsNo subtitles. Black editorial serif for chapter cards; interface text stays native to each app.
SoundQuiet music-like bed (inferred from 89 BPM estimate); no clear cut effects; −25.1 LUFS integrated, −8.6 dBFS true peak.
StructureDrawn setup → analytics-to-document → feedback-to-edits → idea-to-site → brand mark. No explicit in-video CTA visible.

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