软件工程的未来

原文来自 Andrew Ng（吴恩达）发表于 DeepLearning.AI The Batch 的专栏信件，原文地址为：

https://www.deeplearning.ai/the-batch/open-questions-about-the-future-of-software-engineering/

以下是原文的核心观点总结：

AI 对就业的影响：没那么悲观

Andrew Ng 持反对"AI 大裁员"论的立场。他认为，尽管部分人对 AI 导致大规模失业感到担忧，但现实数据并不支持这一悲观判断。根据 Citadel Research 的最新报告，软件工程岗位招聘需求正在快速增长。他承认应届毕业生求职困难，部分岗位（如呼叫中心）受冲击较大，但同时指出疫情期间过度招聘、高利率等因素也是就业市场放缓的重要原因，将问题简单归咎于 AI 是片面的。

已明确的五大趋势

AI 正在深刻改变软件工程的工作方式，以下几点已基本清晰：

编程门槛降低：AI 让更多非专业人士也能写代码
阅读代码变得次要：直接与 LLM 对话、在更高抽象层面操作，比手写或阅读原始代码更重要
定制化软件激增：为极小众群体开发软件在经济上变得可行
"产品管理瓶颈"出现：决定"做什么"比"怎么做"更成为制约因素
技术债成本下降：AI 可以自动重构代码，偿还技术债的代价大幅降低

仍待解决的开放问题

原文提出了若干尚无定论的关键问题：

未来高级工程师的核心能力是什么？计算机科学教育体系该如何重构？
当"人人都能写功能"时，个人和企业的竞争优势来自哪里？
软件开发的新"基础构件"（库、SDK 等）是什么？如何组织多个编码 Agent 协同工作？
未来软件团队的最优人员结构是什么（工程师、PM、设计师的比例）？
AI Agent 如何改变机器学习工程师和数据科学家的工作流程？

核心论断

全文最重要的洞察是 产品管理瓶颈（Product Management Bottleneck）：随着 AI 大幅压缩编码速度，团队的核心瓶颈已从"能否构建"转移到"该构建什么"。Andrew Ng 观察到，他的团队已开始将 PM 与工程师的配置比例从传统的 1:4 调整为接近 1:1，甚至更高，这标志着软件开发范式的根本性转变。[1][2]

来源

[1] The Product Management Bottleneck: How to Decide What to Build https://www.linkedin.com/posts/andrewyng_grok-raises-questions-meta-poaches-talent-activity-7351623555542896640-cTJT

[2] Product management is becoming the new bottleneck according to ... https://x.com/lennysan/status/1943773031459172360

[3] The Future of Software Engineering with AI: Six Predictions https://newsletter.pragmaticengineer.com/p/the-future-of-software-engineering-with-ai

[4] Are AI Agents Really About to Revolutionise Software Development ... https://www.reddit.com/r/AI_Agents/comments/1k8js1h/are_ai_agents_really_about_to_revolutionise/

[5] The Future of AI, Coding and Software Engineering with LLMs and ... https://www.youtube.com/watch?v=m-FU7wYxfjw

[6] How AI Agents Are Transforming Software Engineering and the ... https://www.computer.org/csdl/magazine/co/2025/05/10970187/260SnIeoUUM

[7] The Rise of AI Agents: How 2025 Will Transform Software Engineering https://www.linkedin.com/pulse/rise-ai-agents-how-2025-transform-software-prof-dr-daniel-russo-lopbf

[8] Bottleneck #03: Product v Engineering - Martin Fowler https://martinfowler.com/articles/bottlenecks-of-scaleups/03-product-v-engineering.html

[9] How AI Agents Are Shaping the Future of Software Development https://blog.miraclesoft.com/unlocking-the-future-of-software-development-with-ai-agents/

[10] The New Bottleneck in Product Development - Scrum.org https://www.scrum.org/resources/blog/new-bottleneck-product-development

[11] Will AI Make Software Engineers Obsolete? Here's the Reality https://bootcamps.cs.cmu.edu/blog/will-ai-replace-software-engineers-reality-check

[12] How AI will change software engineering – with Martin Fowler https://www.youtube.com/watch?v=CQmI4XKTa0U

[13] Andrew Ng: Product management is the new bottleneck in tech https://www.linkedin.com/posts/markbarbir_andrew-ng-just-confirmed-what-weve-been-activity-7349564490796912640-TSmt

[14] How AI Agents are Reshaping the Developer Experience https://devops.com/how-ai-agents-are-reshaping-the-developer-experience/

[15] When Code Writes Itself, Product Managers Become The Real ... https://www.forbes.com/councils/forbestechcouncil/2026/03/23/when-code-writes-itself-product-managers-become-the-real-bottleneck/