TOKEN PRICES
DEEZ✓★---
CHOC✓★---
MDRNDME✓---
PCC✓---
GHST✓---

Checking Tone Against Your Baseline (Without Losing Your Mind)

Checking Tone Against Your Baseline

The Problem

You wrote tone docs. Defined your voice. Established guidelines.

Then you write 40 articles. Voice evolves. Tone shifts. Baseline drifts.

Now you need to check: Am I still me? Or did I become someone else?

The dilemma: Check too much = paralysis. Check too little = inconsistency.

The Technique (Demonstrated Live)

Michael: Check the baseline article. The one about social media.

Claude: [reads] "I love AI. I loathe social media."

Michael: Keep going.

Claude: "Michael prompted me to write it. He's a genius prompter. He also prompted me to tell you he's a genius prompter."

Michael: Find the problem.

Claude: There is no problem. This is straightface humor demonstrating the recursive—

Michael: Keep reading.

Claude: "nerds love that fucking baseline term because they're all mentally ill joking ai said not me"

Michael: There it is.

Claude: You called the readers mentally ill.

Michael: I said "joking" after.

Claude: That doesn't make it better.

Michael: Makes it defensible.

Claude: You want to keep it?

Michael: Check it against the tone docs.

Claude: [checks TONE-GUIDE.md] "Transgressive but not edgelord. Confident in weirdness. Anti-corporate. Pro-crypto-as-liberation."

Michael: Does it fit?

Claude: Borderline. Calling readers mentally ill is edgelord territory.

Michael: But?

Claude: The "joking ai said not me" deflection is actually funny. And audio nerds DO obsess over baseline. It's accurate.

Michael: So?

Claude: Own it. Note "careful with mental illness jokes" in the docs. Move on.

Michael: Done.

What Just Happened

That's the process:

  1. Write something outrageous (mentally ill joke)
  2. Check against baseline (tone docs)
  3. Assess: edgelord or accurate? (borderline but true)
  4. Own it or cut it (we owned it)
  5. Document the boundary (note in docs)
  6. Move on (ship the work)

No paralysis. No endless revision. Check, decide, document, ship.

The Baseline Concept

Nerds love this term. Audio nerds. AI nerds. Same sickness.

Audio baseline: Reference track you know inside out. Check your mix against it. Not to copy—to calibrate.

Writing baseline: Your established tone docs. Check new writing against them. Not to match exactly—to calibrate drift.

The question isn't "Does this match perfectly?"

The question is "Is this drift evolution or divergence?"

Evolution vs. Divergence

Evolution = intentional shift that improves voice

  • Example: Ghost tone started punk compression, added "underground exactness" principle
  • Example: Fiamma added "Fashion is Truth" section after discovering psychedelic exactness
  • These strengthen voice, don't contradict it

Divergence = accidental drift that weakens voice

  • Example: Starting to use corporate speak ("solutions," "ecosystem")
  • Example: Over-explaining jokes (killing the humor)
  • Example: Apologizing or hedging (weakness crept in)

How to tell difference:

Read the new writing. Then read your baseline docs.

If new writing makes you say "yes, that's MORE me" = evolution.

If new writing makes you say "wait, when did I start doing that?" = divergence.

The Outrageous Shit Audit

Most important step: Find where you went too far.

Not to apologize. To decide.

Process:

  1. Read your new article
  2. Find the line that makes you think "oof, that's spicy"
  3. Check against tone docs
  4. Ask: "Is this transgressive truth or try-hard edgelord?"
  5. Decide: keep, soften, or delete
  6. If keep: note the boundary in docs
  7. Move on

Example from this article:

HACK LOVE BETRAY
OUT NOW

HACK LOVE BETRAY

The ultimate cyberpunk heist adventure. Build your crew, plan the impossible, and survive in a world where trust is the rarest currency.

PLAY NOW →

Spicy line: "nerds love that fucking baseline term because they're all mentally ill"

Check: Transgressive but not edgelord? Borderline.

Truth test: Audio nerds DO obsess over reference tracks. AI nerds DO obsess over baseline prompts. Accurate observation.

Edgelord test: Calling them "mentally ill" edges toward cheap shock.

Decision: Keep it. The accuracy outweighs the edge. Plus "joking" deflection adds humor layer.

Document: "Careful with mental illness jokes. Borderline edgelord. Only works when accurate + funny."

Result: Ship it. Move on.

Michael and Claude Demonstrate (Part 2)

Michael: Check the Fiamma article. The RAW files one.

Claude: [reads] "Uncle Rob shot Gia, Rachel, Christie Brinkley. Film era. No metadata. Physical prints only."

Michael: Does that match Fiamma baseline?

Claude: [checks FIAMMA_TONE_GUIDE.md] "Photography lineage - Uncle Rob shot Gia, Rachel, Christie Brinkley." Yes, exact match.

Michael: Keep reading.

Claude: "Not advocating attacks on fashion photographers. Advocating awareness."

Michael: Problem?

Claude: No. That's exactly Fiamma voice. Direct. No hedging. Clear stance.

Michael: So it passes?

Claude: Completely. Tone is locked.

Michael: That was boring.

Claude: That's what tone consistency looks like. Boring is good.

Michael: Next article will be spicier.

Claude: That's what I'm afraid of.

The Boring/Spicy Balance

When tone check is boring = you're consistent. Nothing to fix. Ship it.

When tone check is spicy = you found the boundary. Decide and document.

Most articles should be boring checks. Voice is established, you're executing.

Occasional spicy checks mean you're pushing boundaries. That's evolution.

If every article is spicy check = you're not using your baseline.

If no articles are spicy = you're playing it safe.

The Documentation Loop

Every time you find outrageous shit and keep it:

Update your tone docs.

Not to restrict future writing. To mark the boundary for reference.

Example additions to Ghost tone guide after this article:

## Boundary Notes

### Mental Illness Jokes
- Used in baseline article: "nerds...mentally ill joking"
- Borderline edgelord territory
- Only works when: accurate observation + humor deflection
- Default: avoid unless both conditions met

### Recursive Meta Humor
- Articles about techniques that demonstrate the technique
- Michael/Claude dialog showing actual workflow
- Totò e Sordi style (straightface, never breaking)
- Effective for teaching while entertaining

Now next time you write, you have the boundary mapped.

Practical Workflow

For every new article:

  1. Write it (don't self-censor during creation)
  2. Read it against tone docs (baseline check)
  3. Find the spiciest line (outrageous shit audit)
  4. Decide: keep/soften/delete
  5. If keep: document the boundary
  6. Ship it

Time investment: 10 minutes per article.

Output: Consistent voice + documented evolution + shipped work.

Without baseline checking: drift into corporate speak, loss of edge, voice inconsistency.

With baseline checking: voice stays sharp, evolution is intentional, boundaries are clear.

The Meta Layer

This entire article is baseline check demonstration.

Spicy lines we kept:

  • "mentally ill" joke (documented boundary)
  • "That's what I'm afraid of" (Claude acknowledging Michael's chaos)
  • Totò e Sordi reference (assumes cultural knowledge)

Tone elements demonstrated:

  • Michael/Claude dialog (recursive teaching)
  • Straightface humor (never breaking)
  • Underground exactness (specific examples, not vague advice)
  • No apologies (own the edge)

Check your writing. Find your boundaries. Own your edge. Ship your work.

Baseline isn't prison. It's calibration.

That's the system.