Anthropic 的 Boris Churnney 预测软件工程岗位将在今年年底开始消失，而微软的一项研究揭示了 AI 采用与机构奖励之间的差距。
Boris Churnney of Anthropic predicts software engineering jobs will start disappearing by end of this year, while a Microsoft study reveals a gap between AI adoption and institutional rewards.
要点 · TL;DR
AI 编码工具提升生产力，但不会消灭软件工程师岗位。 AI coding tools boost productivity but won't eliminate software engineering jobs.
工作流必须围绕 AI 重新设计，而非简单叠加。 Workflows must be redesigned around AI, not just augmented.
工程师演变为融合编码、产品和设计的构建者。 Engineers evolve into builders blending coding, product, and design.
核心观点 · Key points
像 Claude Code 这样的 AI 编码工具将极大提升工程师的生产力，但失业并非不可避免。 AI coding tools like Claude Code will make engineers vastly more productive, but job loss is not inevitable.
公司必须围绕 AI 重新设计工作流程，而非简单叠加，才能实现生产力提升。 Companies must redesign workflows around AI, not just add it on top, to realize productivity gains.
软件工程师的角色正在演变为融合编码、产品和设计的“构建者”角色。 The role of software engineer is evolving into a 'builder' role that blends coding, product, and design.
AI 采用造成鸿沟：员工担心落后，但很少因尝试 AI 而获得奖励。 AI adoption creates a divide: workers fear falling behind but are rarely rewarded for experimenting with AI.
初创公司拥有前所未有的杠杆作用；一个人借助 AI 就能打造价值数十亿美元的公司。 Startups have unprecedented leverage; one person with AI can build billion-dollar companies.
反共识 · Contrarian takes
对许多任务而言，编码已“解决”，但工程远不止编写代码。 Coding is 'solved' for many tasks, but engineering involves far more than just writing code.
随着 AI 降低门槛，编写代码的人数将增加 100 倍，而非减少。 The number of people writing code will increase 100x, not decrease, as AI lowers barriers.
电工和医生等非工程师通过使用 Claude 构建应用，在黑客马拉松中获胜。 Non-engineers like electricians and doctors are winning hackathons by building apps with Claude.
Token 最大化——浪费性地运行 AI 智能体——在亚马逊等大科技公司是真实存在的问题。 Token maxing—running AI agents wastefully—is a real problem at big tech companies like Amazon.
转换成本等商业护城河将因 AI 能轻松在供应商间迁移数据而削弱。 Business moats like switching costs will erode as AI can easily migrate data between vendors.
本期章节 · Chapters（共 21）
开场与嘉宾介绍Opening and Guest Introduction
微软 AI 工作研究Microsoft Study on AI Use at Work
亚马逊 Meta 的 Token 最大化Token maxing at Amazon and Meta
Boris Churnney 与 Claude Code 介绍Introduction of Boris Churnney and Claude Code
构建首个原型Building the first prototype
AI 工具的早期发现与传播Initial discovery and spread of AI tools
对软件工程岗位的影响Impact on software engineering jobs
历史类比：拖拉机与马Historical analogy: tractors and horses
变化速度与生产力悖论Rate of change and productivity paradox
AI 采用带来的生产力提升Productivity gains from AI adoption
AI 解决编程问题Coding is being solved by AI
生产力与工作生活平衡Productivity and Work-Life Balance
给年轻软件工程师的建议Advice for Young Software Engineers
从编程到通用协作Co-work: From Coding to General Use
就业影响与社会转型Impact on jobs and societal transition
AI 鸿沟与工具获取AI divide and access to tools
预测未来一年的颠覆Predicting disruption in the next year
商业模式与创新Business models and innovation
小型初创公司的杠杆Small startups and leverage
自动化社交媒体互动Automating social media interaction
通过用户反馈改进产品Product improvement through user feedback