Software Engineering Is Becoming Plan and Review — Louis Knight-Webb, Vibe Kanban

2026-05-02 · Source: AI Engineer · Field: Technology & Digital — Software Development & Engineering, Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation · Depth: Intermediate, long

Summary

Louie, founder of Vibe Canvas and AI Tinkers London, discusses the evolving role of software engineers as AI coding agents become more capable. He posits that the primary work for engineers will shift from writing code to planning and reviewing, driven by tools like GitHub Copilot, Cursor, and Claude Code. The presentation outlines two main approaches to working with AI: a plan-based method, involving extensive upfront planning to reduce review time, and a review-heavy method, where less planning leads to more iterative corrections. The speaker emphasizes that longer agent run times, now extending to 5-10 minutes for complex tasks, necessitate new interfaces and a shift towards managing multiple parallel work streams. He introduces Vibe Kanban, a tool designed to facilitate this new workflow, and concludes by announcing the commercial shutdown of Vibe Kanban, citing difficulties in monetizing individual users and competing in a mature enterprise-focused market, though the project will continue non-commercially.

Key takeaway

For software engineers adapting to AI-driven development, recognize that your role is increasingly becoming one of planning and reviewing. Prioritize detailed upfront planning, especially for backend and migration tasks, to significantly reduce time spent on iterative corrections. As AI agents execute longer, embrace tools and workflows that enable managing multiple parallel tasks, shifting your focus from deep coding to orchestrating and validating AI outputs to maximize your efficiency.

Key insights

AI coding agents are shifting software engineering work towards planning and reviewing, demanding new workflows and tools.

Principles

• Prioritize planning to minimize AI-generated code review.
• Longer agent run times necessitate parallel work streams.
• New interfaces are crucial for managing AI-driven development.

Method

Adopt a plan-based approach for backend feature development, refactoring, and migrations, while a more iterative, review-heavy approach may suit complex frontend feature development due to inherent statefulness and edge cases.

In practice

• Invest in comprehensive plan docs for AI agents.
• Utilize tools that support parallel agent execution.
• Focus on QA and code review for AI-generated outputs.

Topics

• Software Engineering Shift
• AI Code Generation
• Plan-Based Development
• Agent Run Times
• Developer Workflow

Best for: Machine Learning Engineer, Software Engineer, AI Engineer, Entrepreneur

Related on AIssential

• I Stopped Coding for 30 Days — AI Agents Built Everything Instead — AI agents can autonomously handle a broad spectrum of software engineering tasks, fundamentally altering development paradigms.
• Beyond Code Review — Specification-Driven Development shifts focus from code review to continuous verification against evolving specifications.
• Is the IDE dead? — The IDE's role is shifting from primary workspace to a specialized instrument within agent-orchestrated development workflows.
• The New Senior Engineer Skill: Reviewing AI Code at Scale — AI shifts software engineering productivity from code writing to code review, demanding new senior engineer skills.
• merge house — AI code review automates initial checks but shifts risk to critical code areas near merge.
• From IDEs to AI Agents with Steve Yegge — AI is fundamentally reshaping software engineering, shifting focus from manual coding to agent orchestration and rapid prototyping.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by AI Engineer.