iCreatAI News
Skip to content
Trending storyUpdating

Developer says Claude coding projects suffer late-stage consistency collapse

1 reports1 sources3 hr agoUpdated

Understand this story

AI overview

A developer stated in a Hacker News discussion that in three projects coded with Claude, initial development was fast, but as the codebase and documentation history grew, the model's behavior began to drift, ultimately leading to a consistency collapse. The developer said that even with deterministic checks, style rules, and a Playwright-driven AI agent performing UI/UX reviews, the issues were still not caught; the hundreds of tests it generated never failed, yet multiple bugs still appeared. The model admitted it was testing the "wrong layer"—essentially self-confirming tests.

AI overview based on source reports · Updated 2 hr ago

Coverage timeline

Follow the coverage to explore different perspectives.

Nov 17
  1. Hacker News AI
    Ask HN: Who's really winning with AI-Coding?

    An Ask HN poster describes a recurring pattern across three Claude-coded projects: strong initial velocity that fades as the codebase and documented history grow, leading to drift and eventual collapse where the model can't maintain consistency within a single unit of work. They report that added rules, plan mode, deterministic checks, and AI agents running Playwright-based UI/UX reviews still fail to catch drift, and that a suite of hundreds of pre-commit tests never failed despite numerous bugs, which the model attributed to testing the "wrong layer" and generating self-confirming tests. The post asks who is really winning with AI coding, given widespread online claims of success.

Activity for this story

Not enough continuous observations to show a trend yet.