From the ground the gladiators fought on to ground that breathes: a 30-second looping spot for the Colosseum's Belvedere, made for a company in sustainable flooring and green building and shown on a live conference screen.
The spot runs on a loop at a conference stand, with no sound, in front of a technical audience. Every shot was planned before anything was generated: what it shows, how long it lasts, how the camera moves, what text sits on screen, and whether it comes from AI or from the real site.
Shots 01–05 are AI-generated: the hourglass, ancient Rome being built, the gladiators, time speeding up, and modern Rome. Each one is built from reference images chosen in advance, so the look stays consistent from shot to shot.
Shots 06–07 use real photos from the Belvedere site, not AI. A technical audience needs to see the actual intervention (aggregates, binder, laying, curing) rather than a rendering of it. A static brand frame closes the spot and loops back to Shot 01.
The production moodboard: direction, camera, on-screen text and the reference images behind every shot.
Extreme macro of a bronze Roman hourglass, golden sand flowing slowly. Dust particles float in golden light against a dark, blurred background.
Camera very slow dolly-in, minimal depth of field.
On-screen text None: pure visual impact.
The hourglass sand turns into construction dust. A zoom-out reveals an aerial view of ancient Rome being built: the Colosseum rising, wooden scaffolding, Roman workers seen from above.
Camera fast pull-back from above, drone angle.
On-screen text “80 AD”
Bird's-eye view of the Colosseum arena. Gladiators in shadow and silhouette, moving more like choreography than combat, kicking up dust and sand. No blood, no deaths.
Camera slow top-down orbit.
On-screen text “Battles”
Creative note: The “Duel” (two figures) fits the intended tone best: less crowded, more cinematic. The other two show denser options for a full arena.
Match-cut from the hourglass to an antique pocket watch. The hands spin faster and faster, then it becomes a modern Apple Watch on a wrist.
Camera tight close-ups, quick match-cuts.
On-screen text “500 · 1000 · 1500 · 1900 · 2000”
Creative note: The reference already shows the diagonal split from antique watch to Apple Watch, exactly the conceptual transition intended. It doubles as a guide frame for the animation in post.
A fast pull-back from the smartwatch reveals today's Rome: the Colosseum still standing, now surrounded by traffic, cars and urban chaos. Cool, almost documentary palette.
Camera dramatic aerial pull-out.
On-screen text “2026”
Perfect match: A modern aerial view with streets, traffic and the urban context around the monument: exactly the concept.
A technical section built on real photos from the Colosseum Belvedere site, not generated by AI. This is where a technical audience recognises that the process is real.
Sand plus medium and large aggregates extracted from the quarry.
On-screen text “Sand. Aggregates. Lab-tested before every job.”
Big bags of Nature Premix (1,200 / 1,500 kg) and a Stabilcure canister.
On-screen text “Eco-friendly binder. Water. Aggregates. One single mix.”
Compactor roller, paver and operators at work with the Colosseum in the background.
On-screen text “Laid. Compacted. Under the Colosseum's gaze.”
Stabilcure is sprayed on and the surface turns white for a few hours.
On-screen text “White for a few hours. Then back to the colour of the earth.”
A calm view of the finished Belvedere. Visitors stroll along the earth-coloured path in warm golden light, returning to the palette of the first shot.
On-screen text “The Colosseum, today. Ready in a few hours.”
Real photo: Trees in the foreground, the Colosseum behind: the mood already matches the final shot. The red survey stakes from the site are masked in post-production.
Full-screen texture of stabilised earth with the brand logo at the centre. A dissolve into golden sand links back to Shot 01, so the loop is invisible.
On-screen text “Natural ground as ecological infrastructure”
| Shot | Timecode | Content | Source |
|---|---|---|---|
| 01 | 0:00–0:03 | Opening: Hourglass | AI-generated |
| 02 | 0:03–0:07 | Rome is born: 80 AD | AI-generated |
| 03 | 0:07–0:12 | The battles: Gladiators | AI-generated |
| 04 | 0:12–0:15 | Time flows: Hourglass → pocket watch → smartwatch | AI-generated |
| 05 | 0:15–0:19 | Urban jungle: Rome, 2026 | AI-generated |
| 06 | 0:19–0:26.5 | The process: The intervention, in real footage | Real site photo |
| 07 | 0:26.5–0:28 | The result: The Belvedere, today | Real photo / AI match |
| 08–09 | 0:28–0:30 | Closing: Logo, manifesto & loop | Static / graphics |
The AI VideoGen course teaches the same pipeline on your own project: story first, reference sheets, a shot list, the right model for each shot, and an edit that holds together.
