Isolation on Programmer.ie: Modern AI programming

Context Is an Input, Not a Transcript

Mon, 14 Sep 2026 05:00:15 +0100

Part 3 — Give Intelligence a Runtime

The question the acceptance could not answer

Chapter 14 ended with a task that was provably complete. The acceptance cited the task’s criteria, the call, its final attempt, the interpretation, the exact artifact bytes and the checks run on them.

It did not cite what the model was shown.

Ask most systems that question and you get one of two answers. The first is “the conversation so far”: everything that happened, in the order it happened. The second is “whatever was in the prompt”: a string someone assembled and nobody recorded. Neither answer can be checked afterwards, and neither was decided on purpose.

Independent Calls

Mon, 14 Sep 2026 05:00:23 +0100

Part 5 — More Intelligence Is Not Automatically Better

Three reviewers, one paragraph

Constructed scene. Three reviewers receive the same versioned paragraph and the same source pack. Each writes a proposal. Every proposal is collected before any is revealed.

Contrast that with a second procedure: show the second reviewer the first answer and ask for agreement. The second process costs less and runs faster. It also sounds more confident, while answering a different experimental question. Agreement after exposure cannot distinguish shared insight from shared influence — which is why the blind collection exists first, and why the next four chapters need it before they ask whether extra candidates help.