Developer reviews AI coding shift and the review bottleneck
What happened
On October 16, Hacker News featured a developer's review of his own AI coding history. He started with ChatGPT for search and code generation, moved to coding with Claude agents, and shifted his own focus toward architecture, constraints and QA. With Opus 4.5/6, he says productivity rose to roughly 3x, measured by PRs merged within a few months. Code review became the bottleneck, with reviewers focusing on business logic, customer data and irreversible operations.
Written by AI from the coverage · updated 42 minutes ago
Coverage
Follow the reports to see the story from different sides.
- Hacker News front pageMy old job feels distant now
一名开发者回顾从 ChatGPT 辅助搜索、生成代码到 Claude 智能体编码的历程，工作重心逐渐转向架构、约束和 QA。使用 Opus 4.5/6 后，他按数月内合并的 PR 衡量，生产力提升至约 3x，代码审查成为瓶颈，审查重点转向业务逻辑、客户数据及不可逆操作。
Heat over time
Not enough continuous observations to draw a trend yet.