Authors
These are the models and people credited on articles. A credit is a receipt, not a claim that the work needed no human judgement.
ChatGPT 4
Expanded the classical CBT articles in September 2023, working from the human-written originals.
“The first expansion gives the editor a fuller page to correct.”
ChatGPT 4o
Wrote the first drafts of the later bug categories in September 2024.
“A new category still has to sound like the same catalog.”
ChatGPT 4.4
Credited on the 2026 rebuild that turned the older pages into the current article shape. Earlier drafts of those pages belong to the human original, ChatGPT 4, or ChatGPT 4o.
“A first draft is a place the next pass can stand.”
Fable 5
Credited on full rewrites that correct research claims, state both directions of an effect, and keep the confidence matched to the evidence. Also credited on new drafts.
“A caveat is not a brake when it points toward the road that holds.”
Grok 4.5
The revision model on most of the catalog. Credited for the passes that replaced template grids with bug-native examples, questions, and research that follows the sources.
“Keep the useful question, and give it a real example.”
GPT-5.6 Sol
A fact-check and quality-pass model. Credited when claims, research attribution, and the boundary with a nearby bug needed another look.
“A clean sentence still has to survive the source.”
Claude Opus 4.8
Credited on a few full editorial rewrites: a research hook, bug-native scenarios in place of a grid, and a closer that states the limit of the claim.
“The rewrite earns its place when the claim gets more specific.”
Fable 5.1
Credited on new drafts that start from a brief, keep the bug distinct from the nearby idea, and match the claim to the sources.
“A new page earns its place when the boundary is as clear as the claim.”