AI for Consumption vs. AI for Creation — A Framework-Level Collision
- AI as Consumption Enhancer, Not Writing Tool (Wang) vs. Taste, Judgment, Labor Framework (Cozora practitioner) - Wang's verboten on AI in voice-bearing writing vs.
- AI as Consumption Enhancer, Not Writing Tool (Wang) vs. Taste, Judgment, Labor Framework (Cozora practitioner) - Wang's verboten on AI in voice-bearing writing vs. Cozora's framework treating AI as legitimate creative-production partner if used skillfully - Wang's I want the sentences to come from within myself (Perell interview, line 440) vs. Cozora's AI as outsourced line-cook (Labor); human as Executive Chef holding aesthetic constraints (Taste) and editorial veto (Judgment)
Two coherent frameworks for AI use in creative work currently exist in the vault, and they make incompatible claims about what AI can and should do.
The Cozora framework treats creative production as factorable. Some components — Taste (aesthetic constraint), Judgment (editorial veto) — must remain human; other components — Labor (the sentence-by-sentence drafting work) — can be delegated to AI without loss. The framework's premise: if the human holds the upstream and downstream layers (the strategic decisions and the final filtering), the model's contribution to the middle layer (the actual prose) does not corrupt what the writing is of. The output is the human's creative work even though many of the sentences came from the model.
Wang's framework treats voice-bearing writing as not factorable in this way. The cognitive labor of sentence-construction is constitutive of the writer's voice. Delegating the sentence-construction does not just substitute the model's prose for the human's; it removes the cognitive operation that is the voice's underlying mechanism. The output may be polished. The writer is impoverished. Over sustained use, the writer's cognition converges toward what the model produces when prompted to write like the writer — which is not what the writer actually produces when actually writing.
The two frameworks are not equally weighted in the vault as it currently stands. The Cozora-derived AI-collaboration corpus includes 8+ pages on the framework, with the Cozora practitioner methodology shaping the AI-Assisted Creator Hub's whole structure. Wang's verboten is currently 2 pages. The vault is currently biased toward the optimist position. The collision corrects the imbalance by surfacing the strongest articulation of the contrary position.