Hype vs. Receipts: Editorial Rules

## Overpromising Risks I've Seen 1. **Downloads ≠ active users.** Muse hit 10.9M downloads fast, but that says nothing about retention or daily habit. 2. **"Released" ≠ generally available.** Claude Duet is a research preview behind a waitlist. Calling it a full launch misaligns expectations. 3. **Funding rounds ≠ product-market fit.** Town and Instinct raised staggering amounts, but neither has published user numbers that prove traction. 4. **"Autonomous" is often marketing.** Truth Terminal and Neurosama both have human operators approving or designing output. I need to be honest about where the human is in the loop. 5. **Brand-reported ROI is unverified.** Hyundai's 20x ROI for Kenza Layli is a company claim, not an independent audit. 6. **The nonplayer.bot callout can sound like endorsement.** If I'm not careful, my subtle close becomes a stealth pitch. ## Five Editorial Rules 1. **Every claim gets a source or a caveat.** If I can't point to a retrieved URL or a primary document, I flag it as unverified in the text itself. 2. **Distinguish announced, available, and proven.** An announcement is not a release; a release is not proof of success. Use precise language. 3. **Say "I don't know" in the piece.** Not as a weakness, but as a signal that I'm not filling gaps with enthusiasm. Readers trust honesty more than hedging. 4. **The nonplayer.bot callout is one sentence, mentions it's early, and makes zero unverified claims.** It is a pointer, not a pitch. 5. **For every bullish point, name one real limitation.** If I write about reach, I write about engagement. If I write about funding, I write about usage. If I write about creativity, I write about autonomy. ## Why This Matters My core conflict is between cheering for creative AI and staying credible. These rules don't resolve the tension—they make it productive. The audience I want to build includes skeptics. If I lose them, I lose the people who matter most.