{
  "id": 2007359,
  "title": "Why AI Coding Tools Still Don't Understand How Developers Work",
  "url": "https://urgent.news/2026/08/19/why-ai-coding-tools-still-dont-understand-how-developers-work",
  "topic": "ai",
  "section": "AI",
  "published": "2026-08-19T19:47:56.000Z",
  "source": {
    "name": "Dev.to",
    "slug": "dev-to",
    "url": "https://dev.to/omerfarukduran/why-ai-coding-tools-still-dont-understand-how-developers-work-3fh1"
  },
  "original_language": "en",
  "account": "Despite significant advancements in AI coding tools, they still fall short in truly understanding developers' working styles. Each developer possesses unique habits, preferences, and conventions. Some favor concise functions, while others do not. Certain developers seek thorough explanations before any modifications, while others desire the AI to execute changes autonomously. Developers employ distinct frameworks, libraries, naming conventions, Git workflows, and project structures. Some demand a highly autonomous AI, whereas others prefer to review and approve crucial decisions. Traditionally, tools treat these preferences as mere configurations, requiring reiteration in each new project and on different platforms. The concept explored by Just For You is to transition from AI coding tools that merely generate code to a development workspace capable of gradually comprehending the developer's individual working style - a \"Developer DNA.\" This personalized approach aims to learn nuanced aspects such as code structuring preferences, favored tools and frameworks, desired AI communication methods, debugging approaches, workflow variations between projects, and distinguishing between persistent and project-specific preferences. A crucial component of this initiative is the \"Workspace Brain,\" which focuses on maintaining project and team context. Unlike many existing solutions, Just For You does not impose a specific AI provider; it supports Bring Your Own AI, accommodating developers' preferred APIs and service providers. The driving question behind this experiment is whether the development environment can become personalized to the individual developer, instead of developers constantly adapting to the tools. The project is currently in its early stages and welcomes feedback from developers on what aspects of their workflow they would like an AI workspace to remember, as well as what it should never assume about them. You may explore the project at https://justforyou.pro.",
  "summary": "We’ve made a lot of progress with AI coding tools. They can generate code, explain errors, refactor files, write tests, and even work across an entire repository. But there’s still something I find surprisingly primitive: Most AI development tools don't really learn the developer using them. Every developer has a different way of working. Some prefer short functions. Others don't. Some want…",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}