Explore. Create. Go viral.

GLOBAL LOCK: Create a vertical UGC-style tutorial video featuring a young South Asian woman with long dark hair wearing a black sleeveless top, speaking directly to camera in a simple room lit by red and magenta ambient light. She holds a small handheld microphone and explains an AI relighting workflow with energetic but clear creator-style delivery. The video should mix direct-to-camera talking-head shots with mobile screen recordings of an AI image relighting interface, plus quick example inserts of portrait images before and after lighting changes. Keep the aesthetic social-first, practical, and creator-education focused: clean vertical framing, legible app UI, punchy captions, and a fast but understandable flow. No heavy cinematic polish; this should feel like a smart social media growth/tutorial reel for creators discovering an AI lighting hack.

[00:00-00:04] Open on the creator speaking to camera in a red-lit room, centered vertically with a handheld microphone. She introduces the topic with excited tutorial energy. Immediately cut to a portrait example of a braided-haired male celebrity or model on a neutral background to establish the type of image being edited.

[00:04-00:08] Stay on the portrait example while subtle lighting changes appear across the face from one version to another. Show the subject leaning forward on folded arms, then reveal a more dramatic split-light or colored relight treatment. The point is to demonstrate that the same portrait can be re-lit in multiple moods.

[00:08-00:12] Return briefly to the creator speaking in the red room, then overlay or cut to bold tutorial text and an on-screen page showing a Hugging Face or similar AI tool listing. The creator’s tone should be “this is how you can do it too,” with fast, encouraging pacing.

[00:12-00:16] Move into a smartphone screen recording of an AI relight interface. Show a portrait being uploaded or selected as the reference, then the tool labeled around “relight” or “image relight.” The UI should feel like a real mobile app or web interface with dark background panels and bright action buttons.

[00:16-00:20] Demonstrate dragging or adjusting the virtual light source around the portrait. Use toggles for top, left, right, back, or similar lighting positions, and show how the face illumination changes interactively. Keep the screen readable and the handoff from talking-head to screen-demo clear.

[00:20-00:24] Continue the tutorial with more parameter adjustments: brightness, color hue, or preset controls. Show a second portrait example of a young man in a white shirt so viewers understand the method works across different faces and lighting conditions.

[00:24-00:27] Highlight before-and-after or multi-style results in a small comparison grid. The creator should remain the implied narrator even if only the screen recording is visible. The tone should stay practical and creator-helpful rather than technical-documentation dry.

[00:27-00:30] End back on the creator in the red-lit room with large bold caption text emphasizing the AI hack and her creator identity. Finish with a confident on-camera expression, microphone in hand, as if inviting viewers to try the tool themselves.

NEGATIVE PROMPT: unreadable phone UI, distorted app buttons, broken hands, inconsistent creator face, overproduced studio look, cinematic bokeh that hides the tutorial, cluttered room background, illegible captions, random unrelated screen content, generic stock footage, low-detail portrait examples, robotic speech, or chaotic transitions that make the workflow impossible to follow.

SPEECH PACK:
[00:00-00:04] Speaker A on camera: energetic hook introducing an AI relighting trick for portraits. Fast creator cadence, crisp mic sound, high lip-sync.
[00:04-00:08] Speaker A voice continues over example portraits, explaining that the same image can be re-lit in different styles. No visible lips required.
[00:08-00:12] Speaker A on camera: “This is how you can do it too” style transition into the tool demo. Medium lip-sync.
[00:12-00:24] Speaker A voiceover over screen recording: explains opening the tool, uploading a reference, dragging light position, and trying presets or sliders. No lip-sync required.
[00:24-00:30] Speaker A returns on camera to wrap up the tutorial and emphasize the AI hack. Confident, creator-educator tone, high lip-sync.