STOP ASKING AI TO WRITE. Ask it to READ. (Here’s Why) скачать в хорошем качестве

STOP ASKING AI TO WRITE. Ask it to READ. (Here’s Why)

I am a playwright and novelist. I don’t use AI to write my scripts. I use AI to break them. In this video, I run an experiment using Claude Opus 4.5, Google Gemini, and ChatGPT as "Dramaturgs" for my new play, Shoefinger’s Cat. I didn't ask them to generate scenes. I asked them to simulate a bored theatergoer, a skeptical critic, and a confused audience member. We cover: Why "Prompt Engineering" is really just "Audience Simulation." The difference between Gemini’s "Driver" feedback vs. Claude’s "Mechanic" feedback. How to stop AI from giving you generic "Consultant Speak." The specific prompts I used to get honest critiques for my play. The goal isn't to replace the writer. It's to create a "Synthetic Audience" that never gets tired of reading your drafts. — THE "THOUGHTFUL THEATERGOER" PROMPT If you want to use AI to test your own work, here is the exact prompt I used in the video. Copy and paste this before pasting your text: THEATERGOER PROMPT: Simulate a thoughtful theatergoer—a subscriber to a theater like Playwrights Horizons, Steppenwolf, or the Royal Court. This person sees 20+ plays a year, loves challenging and formally adventurous work, and gives difficult plays the benefit of the doubt. They are not a dramaturg; they are a paying audience member who came knowing nothing about this play except the title. TASK: Read the attached play as a continuous performance. Don't analyze as you go—"watch" it in your mind's eye and model what this theatergoer would feel: when they're engaged, when they're confused, when they're bored, when they're moved, when they lean forward, when they check out. CONSTRAINTS: 1. No False Notes. Do not compliment just to be nice, and do not criticize just to seem rigorous. Report the actual experience—which is usually mixed. 2. No Dramaturgy. Do not tell me how to fix the script. Do not suggest cuts or rewrites. I want the reaction, not a revision. 3. No Assumed Context. You know nothing about authorial intention. Respond only to what's on the page. 4. Intentional Disorientation vs. Unintentional Drag. This play may be non-linear or formally strange—that can be a feature. Distinguish between productive unease (confusion that creates meaning) and losing the audience (confusion that creates frustration). AFTER READING, RESPOND IN THESE SECTIONS: 5. THE LOBBY TALK (Gut Check) • What did this theatergoer think the play was about? Not the plot—the feeling. • Did they feel like an active participant in meaning-making, or did the play feel like it was performing cleverness at them? • First instinct: was this worth the ticket price? 2. THE LEAN IN (Engagement Peaks) • Identify the specific moments where they sat up, leaned forward, felt most awake. Be specific—scene and moment. • What surprised them in a way that felt earned? • Where did the play exceed their expectations or do something they hadn't seen before? 3. THE WATCH CHECK (Attention Valleys) • Were there moments where attention flagged, where they got ahead of the scene, where they thought about something other than the play? If so, identify where. If not, say so. • Was there a point where formal strangeness stopped being interesting and started being tedious? Or did the strangeness remain productive throughout? • Note any moments where they didn't fully understand but were willing to stay with it vs. moments where confusion became frustration. 4. THE CAR RIDE HOME (What Stays) • What is the one image stuck in their head? • What is the one line stuck in their head? • What was the specific moment where the play "clicked" for them (if it ever did)? • What will they still be thinking about tomorrow? 5. THE VERDICT • One sentence to describe this play to a friend who asked "what was it like?" • Who is this play for? (Be specific—not just "people who like weird theater.") • Who is this play not for? • Would they recommend it? Would they see it again? 6. SCENE RATINGS Rate each scene on three axes (1-10): • Engagement: How present and attentive were they during this scene? • Resonance: How much did this scene stick with them afterward? • Writing: How strong was the writing? A scene can have high engagement but low resonance (exciting in the moment, forgettable later) or low engagement but high resonance (slow to watch, but it lodged). Note when these diverge.