STARTING IMAGE: this exact frame — dock with the Claude icon in focus, cursor positioned on it, ...

STARTING IMAGE: this exact frame — dock with the Claude icon in focus, cursor positioned on it, "Claude" tooltip visible, dark abstract wallpaper, warm room light bleeding in from the sides. SCREEN REFERENCE: attached screenshot of a real Claude chat interface, dark theme, showing a text response — use this exact visual style and text content on the laptop screen in shot 2. If the actual conversation topic differs from car script ideas, that's fine — the goal is to replicate the authentic interface look and legible text rendering, not necessarily this exact conversation. ACTION: the cursor completes a click on the Claude icon (a brief click animation/bounce on the icon, as apps do when opening). Then a HARD CUT (not a smooth transition, not a camera move — an abrupt jump cut like in editing) to a new shot: a side-angle view of the same laptop screen, now showing the dark-mode AI chat interface from SCREEN REFERENCE, with text actively appearing on screen. A hand rests near the keyboard at the bottom of frame. CAMERA: shot 1 (starting image) is a brief static hold with the click animation; shot 2 is a side angle from the left, screen occupying most of the frame, keyboard and hand visible at the bottom. The cut between them is hard and instant, no cross-dissolve. LIGHTING & GRADE: maintain the exact same low-key warm amber-brown color grade, dim room lighting, and film grain as the starting image throughout both shots. SOUND: no dialogue. Ambient sound only: a soft click sound when the icon is pressed, then soft typing/UI sounds as text appears. ASPECT RATIO: 1:1, matching the starting image. CONSTRAINTS: same room, lamp, and lighting mood as the starting image; the dock icon click must look natural (slight bounce, standard macOS app-launch animation); the second shot's screen content must match SCREEN REFERENCE exactly, fully legible; no face visible in either shot; no other people in frame.