Blog

Editorial guides and essays about AI tools, workflows, judgment, and practical adoption.

What the guides are for

dothinker AI nav articles turn fast-moving AI tool news into slower, reviewable guidance. The editorial focus is practical: how a workflow changes, what source material a reader should bring, where AI output needs human review, and which claims should be checked before publishing or buying.

We prefer durable topics over generic launch summaries. A useful guide should help a reader make a decision about code assistants, app-store metadata, Chinese model routing, visual production, workflow agents, or content review even after a vendor updates its interface.

Short notes and early drafts are kept out of the sitemap until they have enough context. Published articles should include clear scope, conservative claims, and a path back to primary sources when pricing, platform policy, API compatibility, or commercial rights are involved.

This blog index is intentionally small. It should point to reviewed essays and guides rather than every draft in the repository. When a topic is too short, too vendor-driven, or too dependent on unverified facts, it can remain unpublished or noindexed until it includes enough analysis, examples, and review notes to help readers make a better tool decision.

Before a new article is promoted here, it should answer a concrete reader question: what changed, what should be tried, what should be avoided, and which facts require a fresh source check. That keeps the index useful for repeat readers instead of turning it into a chronological list of thin posts. Older articles can stay useful when their scope, assumptions, review dates, correction paths, and source expectations remain visible for future readers too.

guide

How to Connect Codex CLI to Other Models: A Custom Provider Setup

A step-by-step guide to pointing OpenAI's Codex CLI at non-default models, from OpenAI-compatible gateways like OpenRouter to local runtimes, using the model and model_providers settings in config.toml, with the exact fields, a worked example, and the gotchas to check.

guide

AI Search and Research Tools Compared

How AI answer engines and research tools differ from traditional search, where they help, and how to use them without trusting them blindly.