subject & visual appearance

subject & visual appearance subject: a woman in her late 20s talking directly into a front-facing phone camera, mid-sentence, gesturing with one hand. south asian, warm brown skin, dark wavy shoulder-length hair tucked behind one ear, minimal makeup, and small gold hoop earrings. centered in frame with clear headroom at the top. attire: a washed-black crewneck sweatshirt, slightly oversized, with the collar sitting loose. no visible logos. setting a small home office late at night. behind her is a dim room: the corner of a desk, a dark monitor bezel with a faint blue standby glow, a half-open laptop throwing cool light up onto the wall, one drooping plant, and closed blinds. the room is dark; only her face is lit. video quality & aesthetic resolution / aspect ratio: 1080 × 1920, 9:16 vertical. composed for mobile, with the subject centered and clear headroom above her head. lighting: ring-light key directly on her face, with the room falling off to near-black behind her. color grade: neutral iphone-native grade, slightly cool shadows, natural skin tones, and no filtered look. camera style: handheld, front-facing smartphone selfie. chest-up close-up, held at eye level with tiny, natural micro-drift. shot on an iphone front camera — not cinematic, with no gimbal smoothness. audio & voice profile voice tone: conversational and low-key confident, slightly conspiratorial — like telling one friend something she just figured out. casual tiktok pacing, with no announcer energy. voice quality: clean close-mic audio, quiet room tone, no echo, and no background noise. background music: none for the first 13 seconds. at 00:13, a soft low synth pad fades in underneath and stays quiet behind the dialogue. script & scene breakdown 00:00–00:04 visual: she is already talking as the shot opens, leaning slightly toward the lens with direct eye contact. one eyebrow raises on “your name.” dialogue: “every question you've ever typed into chatgpt is sitting on a server with your name attached to it.” 00:04–00:10 visual: she counts the examples off on her fingers, with one flick of the hand per item. she glances down briefly on the third example, then looks back into the lens. dialogue: “the lab result you panicked about at 2am. the salary question. the legal thing you didn't want to ask a real lawyer. all of it, logged.” 00:10–00:13 visual: small shrug, half-smile, and one shake of the head. same woman, same black crewneck, same ring-lit face. dialogue: “i didn't stop using ai. i just stopped attaching my name to it.” 00:13–00:20 visual: she reaches off-frame and pulls the open laptop into the lower corner of the frame. the screen is angled away from the lens and blown out by glare so no interface is readable. she keeps her eyes on the camera, not the screen. dialogue: “it's called opengradient chat. same chatgpt, same claude, gemini, grok, all in one tab, except your identity gets stripped off the message before it ever reaches the model.” 00:20–00:26 visual: she taps the laptop lid twice with two fingers on the word “laptop,” then settles back. relaxed posture, steady eye contact. dialogue: “it does video and images too, and the agent runs on my own laptop, so my files never leave my machine.” 00:26–00:30 visual: she leans in close for the final line and drops her voice slightly. after the final word, she holds a beat of direct eye contact before the clip ends. dialogue: “free credits to start. go ask it the thing you'd never put in your search history.” consistency & constraints keep her face, hair length, earrings, and black crewneck identical across all six blocks. keep the ring-light key and dark room identical throughout. one continuous selfie take. no jump cuts. no excessive camera shake. no text overlays. no subtitles. no captions burned into frame. no watermarks. no readable text or logos on any screen.