Answer first, question second

For most of its history, software has been a collection of tools. You learned the word processor so that you could write the letter. You learned the content system so that you could publish the article. The tool sat between you and the thing you wanted, and mastering it was the price of entry.

Luke Wroblewski makes the case in Skip the tools, make the outcomes that AI changes this order. A tool was only ever a way to reach an outcome, so if a model can produce the outcome directly, the tool stops being necessary.

What this means for a blog

A traditional blog is a pile of tools on both sides of the page.

• Writers log in, pick categories, fill in tags, write an excerpt, resize an image and set a publish date.
• Readers scan a homepage, guess at a category, open several articles and assemble the answer themselves.

Neither group wanted any of that. The writer wanted the piece to be published. The reader wanted to understand something.

The version without the tools

A writer hands in the text. Everything around it, the topic, the summary, the cover and the formatting, is produced for them and checked by an editor before it goes out.

A reader arrives with a question, or with nothing more than curiosity about what is new, and gets a direct answer assembled from the archive, with links back to the articles it came from. Wroblewski calls this kind of writing just-in-time content: made for one person, at the moment they need it.

The archive stops being a place you browse and becomes the material answers are made from.

What stays the same

The articles still have to be good. A generated answer is only as reliable as the writing underneath it, which is why this site keeps a plain, permanent archive and labels anything written with the help of AI. The tools disappear. The responsibility for what gets published does not.