Guiding Software Engineering 3.0 with Daoist Principles: Insights from the Dao De Jing (Part 1)

1. Clarifying the "Dao" and "De" of Software Engineering 3.0

The Dao De Jing’s core is “respect the Dao, value the De”. "Dao" is the invisible, fundamental law that governs all things; "De" is the concrete behavior that follows that law. Applied to AI‑augmented Software Engineering 3.0, the "Dao" is not the technology itself but the essence of "serving people, following natural laws, and evolving dynamically". This essence is broken down into three fundamental rules:

Originality of Dao: The primary goal (the "one") is to solve business problems and create core value, not merely to adopt advanced technology. AI models, code, and architecture are the "two" and "three"—means to achieve the "one". All engineering actions must revolve around the "value origin".

Evolutionary nature of Dao: The Dao emphasizes that everything moves and changes. In SE 3.0, AI models evolve with data, business requirements shift, and human‑machine collaboration adapts. Therefore the goal is not absolute stability but "adapting to evolution".

Holistic nature of Dao: Everything is interdependent. SE 3.0 is not a simple overlay of code and AI; it is an organic whole of people, AI, business, and environment. Ignoring this integration leads to imbalance, which the article calls the "harmonious Dao".

In short, the "Dao" of SE 3.0 is the underlying rule "do not stray from value, do not resist evolution, do not fragment the whole".

2. Defining the "De" – Behavioral Guidelines and Boundaries

The "De" translates the Dao into concrete conduct. Three core guidelines are presented:

Uphold righteousness: Prioritize "technology for good" and business value over flashy AI features. Avoid blind reliance on AI that may produce redundant code, biased outcomes, or security risks.

Adaptability: Respect the evolutionary nature of AI and complex systems. Do not force static processes onto dynamic changes, but also retain oversight—e.g., verify AI‑generated code against business needs.

Collaboration: Embrace human‑AI complementarity. Humans retain core decision‑making, while AI assists. Define clear boundaries so AI does not overstep into strategic decisions, and AI remains a supportive tool.

The article likens the Dao to a "root" that sets direction, and the De to a "stem" that supports implementation.

3. Empowering SE 3.0 with Three Core Daoist Thoughts

3.1 Dao follows nature – addressing evolution and adaptation

"Dao follows nature" means aligning with inherent laws rather than imposing artificial constraints. SE 3.0 faces two pains: static processes cannot keep up with AI‑driven evolution, and technology drifts from business. The proposed approach is to let engineering practices follow three natural laws: AI model evolution, business operation patterns, and system growth patterns.

Principle 1 – Do not oppose the law: Avoid demanding perfect AI accuracy or fixing business requirements into rigid waterfall steps.

Principle 2 – Do not fabricate: Do not create fake requirements or force AI to solve problems beyond its capability.

Principle 3 – Emphasize adaptability: Design architectures with AI‑iteration hooks and processes that can be adjusted on the fly.

The execution flow consists of three concise steps:

Identify laws (research): List the three natural laws—business pain points, AI capability and iteration cycle, system evolution direction.

Follow laws (implementation): Align requirement analysis, architecture design, development collaboration, and operations with the identified laws.

Validate laws (verification): At each engineering checkpoint, check whether outcomes conform to the respective law (e.g., AI‑generated code matches business logic, system updates respect evolution patterns).

Example: In a new‑energy BMS project, the team aligns battery charge‑discharge physics (business law), lets AI predict state‑of‑charge (AI law), and designs the architecture to accept retraining interfaces (system law), avoiding over‑design.

3.2 Non‑action governance – clarifying boundaries for human‑AI collaboration

"Non‑action" does not mean inactivity; it means avoiding unnecessary interference. The main pain point is imbalance: humans either drown in AI‑generated details or over‑control AI. The solution is a three‑step rule set:

Do not overstep: Humans stay out of routine AI tasks (code generation, test case creation); AI stays out of strategic decisions (priority setting, architecture).

Define rules: Create a minimal set of top‑level rules that delineate responsibilities, quality standards, and boundaries.

Guide, not control: Humans provide direction—prompt AI appropriately, steer the team toward value, and let AI operate autonomously within the rules.

Implementation steps:

Define rules (core step): Draft 3‑5 concise rules, e.g., AI only performs basic coding, all actions must align with business value, AI results require "law verification" before adoption.

Allocate responsibilities: Humans handle requirement prioritization, core architecture, safety thresholds, and AI result validation; AI handles boilerplate code, test generation, data statistics, and iterative optimization.

Guide execution: Provide clear prompts to AI, set value‑driven goals for the team, and steer system evolution without micromanagement.

Example: In an intelligent cockpit project, developers set core interaction design rules, AI generates UI code and simple tests, and developers only verify safety and scenario compliance, achieving efficient yet controlled development.

Conclusion

The article argues that integrating Daoist first‑principles—recognizing the underlying "Dao" (root) and practicing the "De" (stem)—offers a philosophical yet practical framework for navigating the complexities of AI‑enabled Software Engineering 3.0, fostering harmonious collaboration among people, AI, business, and systems.