Human-Written Final Mile
Use AI for leverage, then restore a recognizably human voice
- Difficulty
- Easy
- Time to result
- ~days to results
- Steps
- 5
- Confidence
- 94%
The Human-Written Final Mile separates content production into assisted and authored stages. AI supports ideation and drafting, where speed and breadth are valuable, but a person takes responsibility for the published result. The human rewrite removes repeated model conventions, restores personal phrasing, and permits small imperfections that signal genuine authorship. This is not an instruction to manufacture errors indiscriminately; it is a safeguard against allowing automated polish to erase identity. The mechanism becomes more valuable as AI drives the cost of copying successful formats toward zero. When audiences repeatedly encounter the same structures, punctuation habits, and rhetorical patterns, those conventions lose effectiveness. A recognizable human voice therefore becomes a differentiator rather than an inefficiency.
Origin
Extracted from Marketing Against The Grain during a discussion with Darius Lam about why Nex uses AI for ideation while retaining human authorship and authentic communication.
Core principles
- 01Use AI to accelerate thinking rather than replace authorship
- 02Treat polished AI patterns as signals audiences may tune out
- 03Preserve natural quirks that make the author recognizable
- 04Make originality more important as imitation becomes cheaper
How to run it
- 1
Generate Possibilities
Use AI to expand the range of topics, angles, hooks, or supporting ideas available for the piece.
Pro tip Ask for contrasting angles rather than one supposedly finished answer.
Watch out Do not mistake the first plausible output for an original idea.
- 2
Build a Working Draft
Let AI organize raw ideas into a draft that gives the author something concrete to react to.
Pro tip Treat the draft as disposable scaffolding.
Watch out Publishing this draft unchanged preserves the model's generic patterns.
- 3
Rewrite in Your Voice
Rewrite the argument using vocabulary, sentence rhythms, opinions, and examples that are natural to the actual author.
Pro tip Read the piece aloud to detect language you would never say.
Watch out Do not add artificial mistakes solely to perform authenticity.
- 4
Remove Pattern Residue
Look for punctuation, transitions, formatting, and stock phrases that make the AI's contribution conspicuous.
Pro tip Maintain a short list of patterns your audience increasingly associates with generated content.
Watch out Blindly banning a punctuation mark can also remove legitimate elements of your voice.
- 5
Run an Authenticity Check
Confirm that the final piece expresses a real judgment and would still be recognizable without a byline.
Pro tip Compare it with several pieces you wrote before using AI.
Watch out Fluency alone is not evidence of authenticity or originality.
In the wild
A marketer uses AI to produce possible angles and a first draft for a LinkedIn post. Instead of publishing it directly, the marketer rewrites the final version, replaces conspicuous model-like punctuation, adds a genuine opinion, and preserves the informal phrasing used in ordinary conversation.
→ The author retains AI's speed while producing content that sounds personal rather than mass-generated.
The hosts imagine a deliberately imperfect billboard with misspellings and an explicit claim that it was not produced by AI. The concept works because it visibly rejects the polished sameness of generated advertising rather than producing another optimized variation.
→ Imperfection becomes an attention mechanism in a market saturated with automated polish.
Common mistakes
Publishing the Draft Unchanged
An AI draft may be grammatical and coherent while still carrying the same patterns audiences see everywhere else.
Manufacturing Random Errors
Authenticity comes from genuine authorship, not from inserting arbitrary typos that damage clarity or credibility.
Polishing Away the Author
Overusing writing tools can flatten unusual but recognizable traits that make a person's communication distinctive.
Is it for you?
Best for
It is best for marketers and creators who want AI efficiency without sacrificing trust or personal voice.
Not ideal for
It is not ideal for low-stakes, purely functional copy where speed matters more than a distinctive voice.
From the transcript
“We might use them to help us write drafts, but the actual end result has to be written by us.”
“AI causes uh originality to be more important than ever.”
“the best ways to reach our customers has been to go more direct, be more authentic, and to use a lot of writing tools less.”
From the episode
This AI Startup Is Beating Apple and NVIDIA at Their Own Game