Concept: Pro-Authorship & Voice Discovery
Pro-Authorship is an operational framework and philosophy created by nate-b-jones to eliminate ai-slop and model homogenization in professional knowledge work. It asserts that knowledge workers must take full ownership, conviction, and accountability for written communication produced with AI assistance.
The Model Convergence Problem
Standard anti-slop approaches rely on negative checklists (e.g., banning phrases like “tapestry”, “delve”, or M-dashes). However, Nate B Jones points out that generic checklists fail due to model convergence:
- RLHF and preference optimization train AI models to climb the same mathematical hill toward average, broad definitions of “good” writing (confident, balanced, polite).
- Applying a uniform ban list simply forces models to climb a different hill, leading to a secondary wave of homogenized, formulaic output.
Core Principles of Pro-Authorship
- Authorship as an Iterative Wrestling Match: Authorship does not require typing every keystroke manually. It requires knowing what you want to communicate, wrestling with the draft through iterative AI passes (e.g., 50+ iterations akin to Amazon PR/FAQ drafts), and refining the work until it accurately reflects human intent.
- Accountability & Intent: “If you didn’t read it / if you don’t mean it, don’t send it.” Unchecked AI text pushes cognitive work downstream, forcing recipients to spend hours deciphering low-intent tokens.
- Earning Human Attention: As token generation costs fall to zero, genuine human attention becomes the ultimate scarce resource. Pro-authorship ensures communication stands out by reflecting distinct human perspective.
Voice Discovery Skills
To operationalize Pro-Authorship without manual style specification, Nate introduced Voice Discovery Skills:
- Dynamic Analysis: Instead of forcing users to manually code tone guidelines, a Voice Discovery Skill analyzes rough voice dictations, draft edits, and personal communication patterns.
- Forming a Theory of Voice: The skill builds an evolving, testable theory of the user’s personal writing style, enabling AI models to prune default LLM-isms while preserving the author’s dynamic rhythm and expression.