book

Beyond Vibe Coding

by Addy Osmani

August 2025

Intermediate to advanced

254 pages

7h 51m

English

O'Reilly Media, Inc.

Book available

Read now

Unlock full access

Who This Book Is ForWhat to ExpectConventions Used in This BookO’Reilly Online LearningHow to Contact Us
The AI Coding Spectrum: From Vibe Coding to AI-Assisted EngineeringThe Vibe-Coding Approach: Code by ConversationThe AI-Assisted Engineering Approach: Structure with an AI PartnerDifferent Mindsets, Different ExpectationsFinding Your Place on the SpectrumBeyond Lines of Code: Programming with IntentThe Rise of the Prompt: From Instructions to DescriptionsHow It Works: The Iterative Cycle and AI’s Role in Code GenerationProductivity, Accessibility, and the Changing Nature of ProgrammingA Glimpse of the Tools: The Emerging EcosystemVSCode + Copilot: Microsoft’s Integrated AI Development PlatformVSCode + Cline: The Open Source Autonomous Coding AgentCursor: The AI-Driven Code EditorWindsurf: An AI-Powered IDE with Full Codebase IndexingAI Models: The Landscape for Code GenerationUnderstanding Model CategoriesChoosing the Right Model for Your TaskPractical Tips for Any ModelMajor ModelsGoogle Gemini: The Multimodal Coding PowerhouseClaude: The Reasoning VirtuosoChatGPT: The Versatile Coding CompanionChoosing the Right Model for Your NeedsThe Benefits and Limitations of Vibe Coding: A Nuanced ViewIdeal Use Cases for Vibe CodingWhere AI Still StrugglesSummary and Next Steps
Prompt Engineering FundamentalsSpecificity and Clarity: Writing Prompts That DeliverIterative Refinement: The Feedback Loop with the AIComparing Two PromptsPoor promptImproved promptPrompting Techniques: A Toolbox for Effective CommunicationZero-Shot PromptingOne-Shot and Few-Shot PromptingChain-of-Thought PromptingRole PromptingContextual PromptingMetapromptingSelf-Consistency (Multiple Outputs and Majority Voting)ReAct (Reason + Act) PromptingAdvanced Prompting: Combining Techniques and Handling ComplexityKnow the Model’s LimitsStateful Conversation Versus One-Shot PromptingCommon Prompt Antipatterns and How to Avoid ThemSummary and Next Steps
How Developers Are Actually Using AICommon Failure PatternsWhat Actually Works: Practical Workflow PatternsThe Golden Rules of Vibe Coding0. Summary and Next Steps
Senior Engineers and Developers: Leverage Your Experience with AIBe the Architect and the Editor in ChiefUse AI as a Force Multiplier for Big InitiativesMentor and Set StandardsContinue to Cultivate Domain Mastery and ForesightHone Your Soft Skills and LeadershipMidlevel Engineers: Adapt and SpecializeLearn to Manage Systems Integration and BoundariesBuild Your Domain ExpertiseMaster Performance Optimization and DevOpsFocus on Code Review and Quality AssuranceLearn Systems ThinkingBe Adaptable—and Never Stop LearningGet Good at Cross-Functional CommunicationLearn System Design and ArchitectureUse AI!Venture into UI and UX DesignJunior Developers: Thrive Alongside AILearn the Fundamentals—Don’t Skip the “Why”Practice Problem Solving and Debugging Without the AI Safety NetFocus on Testing and VerificationBuild an Eye for MaintainabilityDevelop Your Prompting and Tooling Skills (Wisely)Seek Feedback and MentorshipCommunicate and CollaborateShift Your Mindset: From Consuming to CreatingSummary and Next Steps
From Intent to Implementation: Understanding the AI’s InterpretationThe “Majority” Problem: Most Common Doesn’t Mean Most AppropriateCode Readability and Structure: Patterns and Potential IssuesDebugging Strategies: Finding and Fixing ErrorsRefactoring for Maintainability: Making AI Code Your CodeThe Importance of Testing: Unit, Integration, and End to EndSummary and Next Steps
Rapid Prototyping with AI AssistantsAI Prototyping ToolsFrom Concept to Prototype: Iterative RefinementEvolving a Prototype Toward ProductionAddressing Challenges in AI PrototypingSummary and Next Steps
Setting Up the Project: Scaffolding with AIFrontend Development Patterns with AIBackend/API Development Patterns with AIDatabase Design and IntegrationUsing an ORMDatabase QueriesChecking AI-Generated QueriesFull Stack Integration: Marrying Frontend and BackendAligning Frontend and Backend ContractsReal-Time Collaboration with AITesting and Validation for AI-Generated Web ApplicationsExamples of Successful AI-Built Web ProjectsSummary and Next Steps

Common Security Vulnerabilities in AI-Generated CodeImproper Authentication and AuthorizationPackage Management IssuesSecurity AuditsLeverage Automated Security ScannersUse a Separate AI as a ReviewerPerform a Human Code Review with a Security ChecklistPenetration Testing and FuzzingAdd Security-Focused Unit TestsProvide Updates to Compensate for Training CutoffsOptimize Your Logging PracticesUse Updated Models or Tools with a Security FocusPay Attention to Warnings in ContextSlow DownBuilding Effective Testing Frameworks for AI-Generated SystemsPerformance OptimizationEnsuring Maintainability in AI-Accelerated CodebasesWhile PromptingWorking with Code OutputFollow-UpCode Review StrategiesBest Practices for Reliable DeploymentBefore and During DeploymentOngoing Best PracticesSummary and Next Steps
Intellectual Property ConsiderationsWhat to Do If You Get Suspicious OutputGray AreasTransparency and AttributionBias and FairnessGolden Rules for Responsible AI UseSummary and Next Steps
From Copilots to Autonomous Agents: What Are Background Coding Agents?How Do Autonomous Coding Agents Work?How Do Background Agents Compare to In-IDE AI Assistants?Combining Multiple AI Models to Maximize StrengthsDifferentiate Models by Task TypeUse an Orchestration SystemHuman-AI Hybrid TeamsMajor Players in Autonomous Coding AgentsChallenges and LimitationsBest Practices for Using AI Coding Agents EffectivelyStrategically Select the Tasks Autonomous Agents Are Going to ImplementLeverage Agent-Specific Planning and Oversight FeaturesManage Concurrent Agent OperationsEvolve Your Team Practices to Integrate AgentsBuild Feedback Loops with Autonomous SystemsSummary and Next Steps
AI in Testing, Debugging, and MaintenanceAutomated Test GenerationIntelligent DebuggingPredictive Maintenance and RefactoringAI-Driven Design and User Experience PersonalizationGenerative Design ToolsAI for UX ResearchPersonalized User ExperiencesThe Evolution of Project Management with AIHow Autonomous Agents Could Change Software EngineeringThe Future of Programming Languages: Natural-Language-Driven Development?How Vibe Coding Is Reshaping the IndustrySummary and Next Steps

Content preview from Beyond Vibe Coding

Chapter 4. Beyond the 70%: Maximizing Human Contribution

You’ve seen how AI coding assistants like Cursor, Cline, Copilot, and Windsurf have transformed how software is built, shouldering much of the grunt work and boilerplate—about 70%.¹ But what about that last “30%” of the job that separates a toy solution from a production-ready system? This gap includes the hard parts: understanding complex requirements, architecting maintainable systems, handling edge cases, and ensuring code correctness. In other words, while AI can generate code, it often struggles with engineering.

Tim O’Reilly, reflecting on decades of technology shifts, reminds us that each leap in automation has changed how we program but not why we need skilled programmers. We’re not facing the end of programming but rather “the end of programming as we know it today,” meaning developers’ roles are evolving, not evaporating.

The challenge for today’s engineers is to embrace AI for what it does best (the first 70%) while doubling down on the durable skills and insights needed for the remaining 30%. This article dives into expert insights to identify which human skills remain crucial. We’ll explore what senior and midlevel developers should continue to leverage and what junior developers must invest in to thrive alongside AI.

This chapter’s goal, then, is to offer you pragmatic guidance for maximizing the value of that irreplaceable 30%, with actionable takeaways for engineers at every level.