Breaking Out of Code Overwork: Understanding Lu Su’s Engineer Value Triangle in the AI Era

Introduction

The AI wave has created a common career dilemma for engineers: those who have spent a decade mastering low‑level code feel anxious as AI coding tools become widespread; others who only follow business requirements are lost when AI reshapes business logic; and some who chase rapid delivery see their systems riddled with bugs and diminishing long‑term value.

Engineer Value Triangle

In a recent Ant Group CTO panel featuring Lu Su, Lao Miao, and Liang He, Lu Su introduced the “Engineer Value Triangle,” a model built on Ant’s two‑decade experience from distributed architecture to the AI era. The triangle clarifies the long‑standing debate about whether engineers will be replaced by AI.

Why Traditional "Skill" Is No Longer Core

During the golden age of mobile internet and distributed systems, the prevailing growth logic was simple: hone coding skill, master framework source code, and excel at low‑level operations. In 2006, Lu Su and Lao Miao led a company‑wide migration to self‑developed distributed middleware, proving that top‑level skill was the key competitive edge.

AI now overturns this rule. The “skill” side of the triangle—API lookup, syntax writing, basic framework development, standardized testing, and generic deployment—can be performed by AI tools with zero barrier and often higher quality and efficiency. Lu Su stated that current AI already matches the execution level of an Alibaba P8 senior engineer. Engineers who continue to spend time on repetitive coding, syntax memorization, and generic bug fixing are essentially competing with AI’s strongest capability, leading to deeper over‑competition.

Creativity as the New Core Barrier

Creativity is identified as the first core barrier for engineers in the AI era. AI can recombine existing knowledge and replicate mature business logic, but it cannot autonomously discover new problems, build solutions from zero to one, or assess long‑term customer and business value.

Lao Miao offered a four‑dimensional standard for evaluating AI project value: genuine customer need, alignment with corporate strategy, business ceiling, and product self‑evolution potential. This standard reflects the practical application of engineer creativity.

Craftsmanship Remains Essential

Craftsmanship defines how high an engineer can rise steadily. AI has inherent flaws—hallucinations, lack of business context, and disregard for long‑term system stability. While AI can generate code quickly, it does not consider three‑year scalability, financial safety, or user‑experience details. Ant’s two‑decade experience in financial payments, handling trillion‑level fund flows and hundreds of millions of daily transactions, demands unwavering craftsmanship.

Examples illustrate the difference: an AI‑generated code snippet deployed without review may introduce hidden financial risks, whereas an engineer with craftsmanship conducts thorough logic verification, anticipates future scaling, and mitigates hidden safety issues, sacrificing short‑term speed for long‑term differentiation.

Re‑orienting Growth Focus

The article proposes moving from a skill‑driven model to a dual‑wheel model of “craftsmanship + creativity.” Past hierarchy: Skill > Creativity > Craftsmanship. AI era hierarchy: Craftsmanship + Creativity > Skill. Engineers should let AI handle repetitive tasks, invest the saved time in creativity and craftsmanship, and thus build an irreplaceable core value.

Practical Steps for Front‑line Engineers

When receiving a requirement, first probe the underlying business essence instead of jumping straight into development.

Review existing system pain points and proactively propose optimization or refactoring plans.

Consider industry trends to explore whether AI can enable entirely new possibilities for the current business.

Broaden perspective beyond a single technology view; learn business and commercial knowledge to understand profit logic and user value.

Free up skill effort by mastering AI development tools and delegating repetitive coding and debugging to AI.

Continuously train creativity by stepping out of pure technical view, actively understanding business, customers, and commerce, and practicing problem definition and technical direction planning.

Polish long‑term craftsmanship by establishing code review and risk self‑check habits, emphasizing system stability, compliance, and taking responsibility for verifying AI outputs.

Conclusion

AI lowers the barrier to coding but amplifies the value of human thought. Engineers who focus on creativity and craftsmanship will remain indispensable, while those who cling to pure skill risk being eclipsed by AI. The future of engineering lies in mastering AI as an assistant while cultivating the uniquely human abilities that AI cannot replicate.