{"slug": "i-think-we-re-measuring-software-engineers-wrong", "title": "I Think We're Measuring Software Engineers Wrong.", "summary": "A developer argues that the software industry is measuring engineers with outdated metrics like lines of code or number of commits, while AI tools have made code generation cheap. The real value now lies in understanding business domains, reducing complexity, and making architectural decisions that prevent future problems. The developer advocates shifting focus from output to impact, such as modeling institutional knowledge and designing maintainable systems.", "body_md": "A few years ago, one question appeared in almost every engineering interview.\n\n\"How many programming languages do you know?\"\n\nThen it became:\n\n\"How many years of experience do you have?\"\n\nToday it's slowly becoming:\n\n\"Which AI coding tool do you use?\"\n\nCursor.\n\nClaude Code.\n\nGitHub Copilot.\n\nCodex.\n\nWindsurf.\n\nLovable.\n\nBolt.\n\nThe tools have changed.\n\nBut I think we're still measuring engineers using the wrong metrics.\n\nFor decades, software engineering rewarded output.\n\nMore commits.\n\nMore pull requests.\n\nMore features.\n\nMore lines of code.\n\nIt made sense.\n\nWriting software was expensive.\n\nEvery line represented time.\n\nEffort.\n\nKnowledge.\n\nToday...\n\nAI can generate hundreds of lines before you finish your coffee.\n\nSuddenly, writing code is no longer the bottleneck.\n\nSo why are we still acting like it is?\n\nThis isn't a bad thing.\n\nIt's progress.\n\nBoilerplate.\n\nCRUD endpoints.\n\nConfiguration files.\n\nDocumentation.\n\nTests.\n\nMuch of this can now be generated in seconds.\n\nThe cost of producing code has dropped dramatically.\n\nWhenever something becomes cheaper...\n\nSomething else becomes more valuable.\n\nNot syntax.\n\nUnderstanding.\n\nGood engineers don't just write software.\n\nThey answer questions like:\n\nHow should this system evolve over the next three years?\n\nWhich service should own this data?\n\nWhat happens when the database is unavailable?\n\nHow do we avoid duplicating business rules?\n\nHow do we keep different teams from stepping on each other?\n\nNone of these questions disappear because AI exists.\n\nIn many ways, they become even more important.\n\nOne thing I've noticed while building enterprise software is that AI understands programming surprisingly well.\n\nIt understands Python.\n\nGo.\n\nTypeScript.\n\nReact.\n\nFastAPI.\n\nSQL.\n\nWhat it doesn't understand is **your business**.\n\nAsk it to generate an authentication service.\n\nYou'll probably get a good result.\n\nAsk it to explain why Customer A can partially pay Invoice B under Contract C while Customer D cannot.\n\nNow the conversation changes completely.\n\nThat's no longer a programming problem.\n\nThat's institutional knowledge.\n\nEvery business has rules that never appear in tutorials.\n\nHealthcare.\n\nManufacturing.\n\nBanking.\n\nInsurance.\n\nRetail.\n\nLogistics.\n\nGovernment.\n\nEventually you discover sentences like:\n\n\"This customer follows the legacy billing process.\"\n\nOr:\n\n\"That contract uses a completely different approval workflow.\"\n\nOr:\n\n\"Invoices created before 2022 are handled differently.\"\n\nNone of this exists inside a language model.\n\nSomeone has to model it.\n\nSomeone has to protect it.\n\nSomeone has to maintain it.\n\nThat's engineering.\n\nThe engineers I admire rarely impress me because they code faster.\n\nThey impress me because they reduce complexity.\n\nThey ask better questions.\n\nThey see failure modes before they happen.\n\nThey simplify architectures.\n\nThey create systems that other engineers enjoy working on.\n\nIronically...\n\nMany of them probably write fewer lines of code than junior developers.\n\nBut every line carries much more value.\n\nSome people worry AI will replace software engineers.\n\nI think something different is happening.\n\nAI is raising the minimum expectation.\n\nIf everyone can generate CRUD applications...\n\nCRUD applications stop being impressive.\n\nThe differentiator becomes everything around the code.\n\nArchitecture.\n\nCommunication.\n\nDomain modeling.\n\nReliability.\n\nObservability.\n\nBusiness understanding.\n\nThese aren't disappearing.\n\nThey're becoming the job.\n\nI still use AI every day.\n\nProbably more than ever.\n\nBut I ask different questions now.\n\nInstead of:\n\n\"Write this API.\"\n\nI ask:\n\n\"What architecture would make this easy to maintain?\"\n\nInstead of:\n\n\"Generate a database schema.\"\n\nI ask:\n\n\"What's the domain model behind this business?\"\n\nInstead of measuring how quickly I can generate code...\n\nI measure how many future problems I can avoid.\n\nThat single shift has probably saved me more time than any coding assistant ever has.\n\nSometimes we describe senior engineers as people who know more technologies.\n\nI'm no longer convinced that's true.\n\nMaybe seniority is simply the ability to recognize patterns.\n\nTo understand trade-offs.\n\nTo make good decisions with incomplete information.\n\nAI can help us write implementations.\n\nExperience still helps us choose the right implementation.\n\nThose are very different skills.\n\nSoftware engineering isn't disappearing.\n\nIt's evolving.\n\nAs AI continues to automate implementation, the profession moves closer to what it has always been underneath:\n\nDesigning reliable systems that solve meaningful problems.\n\nThe tools will continue to improve.\n\nModels will become faster.\n\nFrameworks will come and go.\n\nBut one thing will remain surprisingly constant.\n\nTechnology changes.\n\nEngineering judgment compounds.\n\nAnd I suspect that's what will separate great engineers from everyone else over the next decade.\n\nOver the past several months, I've been documenting how these ideas apply in real enterprise software.\n\nInstead of focusing on AI demos, I built a complete **Enterprise AI Transaction Intelligence System** that covers the entire engineering lifecycle:\n\nThe complete implementation—including three technical handbooks, production-ready Python source code, synthetic datasets, and architecture documentation—is available here:\n\n📘 **Enterprise AI Automation Blueprint**\n\n👉 [https://uigerhana.gumroad.com/l/enterprise-ai-automation-blueprint](https://uigerhana.gumroad.com/l/enterprise-ai-automation-blueprint)\n\nI'm also publishing a long-form Dev.to series on Enterprise AI Engineering, Software Architecture, and Production AI Systems.\n\nIf you're interested in building systems that last—not just demos—I hope you'll follow along.\n\nHappy building. 🚀", "url": "https://wpnews.pro/news/i-think-we-re-measuring-software-engineers-wrong", "canonical_source": "https://dev.to/uigerhana/i-think-were-measuring-software-engineers-wrong-5e6b", "published_at": "2026-06-25 01:38:28+00:00", "updated_at": "2026-06-25 02:14:04.384302+00:00", "lang": "en", "topics": ["artificial-intelligence", "developer-tools", "ai-tools", "ai-agents"], "entities": ["Cursor", "Claude Code", "GitHub Copilot", "Codex", "Windsurf", "Lovable", "Bolt", "FastAPI"], "alternates": {"html": "https://wpnews.pro/news/i-think-we-re-measuring-software-engineers-wrong", "markdown": "https://wpnews.pro/news/i-think-we-re-measuring-software-engineers-wrong.md", "text": "https://wpnews.pro/news/i-think-we-re-measuring-software-engineers-wrong.txt", "jsonld": "https://wpnews.pro/news/i-think-we-re-measuring-software-engineers-wrong.jsonld"}}