Category: Uncategorized

Somewhere in your company there’s a double-approval that exists because of an incident nobody remembers, a Friday report built for an executive who left in 2019, and a threshold a COBOL programmer encoded from a policy meeting nobody minuted.

Every buyer asks the same question: “will it actually learn our business?” They were told it would, and it didn’t — so they assumed they’d been naive to expect it. They weren’t. They specified intelligence correctly.

She wrote down every question her staff asked, and every answer she gave, for ten years. Hundreds of pages. Her staff still asked. What failed wasn’t her discipline — it was a step that didn’t exist yet.

I’ve spent a year arguing that agents want a wiki, not a retrieval index — and I stand by every word of it.

“Voice AI” is one label stretched over two different businesses. One sells a conversation that sounds human; the other sells an action a human would otherwise have to authorize.

A flattened, denoised DOM is to a rendered image what a tree-sitter skeleton is to code — structure without the bytes — and a language model reasons better over that skeleton than over the picture itself.

Every time an agent queries your knowledge base it produces two things worth keeping, and most systems keep neither. Karpathy files the answer back into the wiki as a new page, so explorations compound instead of evaporating.

Your coding agent writes a dated, first-person record of how you think — your intent, the alternatives you rejected, the plans you never shipped. Then it deletes it on a thirty-day timer. The repository can reconstruct none of it.

Your agent explores a wiki cleanly and thrashes doing RAG searches — repeating itself, missing the obvious, unsure when to stop. It isn’t the prompt and it isn’t the model.

Everyone replying to that tweet is deriving their take right now — hot, thin, sourceless. Yours was compiled months ago, and the agent just looks it up.