Hum a few bars
Appearance
INT. STUDY – DAY A visiting tech entrepreneur stands near the fireplace, flanked by a calm, headphone-clad sound technician holding a boom mic. Joanna is seated, motionless. Gyles is humming to himself and straightening his posture. ENTREPRENEUR We just need a clean vocal pass, some phoneme mapping, and a baseline tone curve. Shouldn’t take more than an hour. GYLES Lovely. Will there be a screen? I don’t mind reading lyrics cold, but I do prefer a guide. [The sound tech looks up, uncertain.] ENTREPRENEUR Sorry—lyrics? GYLES I assumed this was for karaoke. [Joanna turns her head.] JOANNA Karaoke. GYLES A tech-forward karaoke concept. Very modern. ENTREPRENEUR No—Mr. Brandreth, we’re here to record your voice for use in an AI assistant. GYLES Oh! ENTREPRENEUR We’re building a voice interface modeled around conversational warmth and low-friction relatability. Your tone tested well across all demographics. JOANNA (blinking) You tested Gyles? ENTREPRENEUR Extensively. He scored high in “benign familiarity” and “safe optimism.” There’s a quality of... casual reassurance. [Joanna raises an eyebrow.] JOANNA Reassurance? ENTREPRENEUR Yes. We tested models that listened attentively—too attentively— and found users actually preferred the realism of an assistant that only partially listens to what you're saying. JOANNA (to Gyles) Looks like you have a fan base. I’ll be upstairs. [She stands and exits.] GYLES I’m not familiar with that song— “AI Assistant”—but if you hum a few bars, I might be able to get into it. [Joanna rolls her eyes on the way out.] [The sound technician shrugs, glances at the entrepreneur, and presses record. Gyles begins to sing—something vaguely showtune-like. Everyone winces. The entrepreneur quickly gestures to stop recording.] FADE OUT.