Articles on this blog are produced through a research-first pipeline. For each topic, we run competitive SERP analysis to understand what currently ranks, then conduct iterative deep research using web-grounded AI models to gather book examples, data points, academic sources, and practitioner perspectives. A structured outline is generated from that research before any draft is written.
Drafts are generated using Claude Opus with the research bundle as primary source material — not training data alone. Every factual claim in the draft must map to a named source in the research bundle.
Every draft goes through human editorial review before publication. The review process includes: verifying that all book examples are real and the craft analysis is accurate; checking data points against cited sources; editing for voice, accuracy, and emphasis. No article publishes without this gate.
Articles on this blog are written with AI assistance. Specifically: research synthesis, outline generation, and first drafts are AI-assisted. Human editorial review, fact-checking, and final editing are human. The voice guidelines, quality rules, and entity requirements that govern AI generation are set by Charles Antoine.
We disclose this because AI-assisted writing, done rigorously, is compatible with genuine expertise and craft journalism. The research foundation and the human review gate are what distinguish what we publish from AI slop.
If you find a factual error in any article, contact us at editorial@unsloppy.ai. Verified corrections are applied within 48 hours and noted with a dateModified update in the article schema.
Last updated: April 2026