2026 Gaokao: Is Majoring in Computer Science or Software Engineering a Trap or the Right Path in the AI Era?

AI impact on programming jobs

Industry shows a K‑shaped split: low‑end roles are being rapidly replaced while high‑end roles see a surge in demand.

Low‑end roles

Current state: mainstream AI coding tools (Codex, Claude Code, GitHub Copilot) can generate more than 80% of basic code.

Typical positions replaced: junior developers who only write CRUD operations, simple APIs, or perform page layout.

Evidence: a Harvard study reports that entry‑level developer positions fell 9%–10% within six months after AI adoption; some large tech firms cut new‑graduate hiring by over 50%.

Conclusion: pure “code‑copying” software engineering becomes a low‑end, salary‑compressed field.

High‑end roles

Irreplaceable abilities: architecture design, complex system construction, low‑level development, AI model training, security, and deep business understanding.

Talent scarcity: engineers/architects who combine AI knowledge with system design are in short supply.

Conclusion: AI eliminates low‑end labor and rewards technical decision‑makers.

Overall assessment

Computer science remains a valuable major, but students must acquire skills that AI cannot perform.

Choosing the major solely for a diploma or easy income is discouraged.

2026 Gaokao major selection

Priority ranking of majors (risk‑resistant in the AI era)

First tier (most stable, high salary, hard to replace):

Computer Science (CS): strong foundation in algorithms, systems, and networks; broad employment; high risk resistance.

Artificial Intelligence / Intelligent Science and Technology: focuses on large models, deep learning, and algorithm research – the “builders of AI”.

Cybersecurity / Information Security: demand grows with AI advancement; near‑zero replacement risk.

Second tier (practical, broad employment, avoid pure CRUD):

Software Engineering (SE): main battlefield for AI‑assisted development; must emphasize architecture, cloud‑native, and AI engineering.

Data Science and Big Data Technology: core is data governance, modeling, and business interpretation; AI is a tool.

Third tier (high risk): ordinary “Software Engineering (application)” at second‑tier colleges, focusing on basic coding and easily replaced by AI.

Score‑band recommendations (2026 reference)

630+ (985 / top 211): recommended directions – Computer Science, AI, Information Security. Strategy – deepen low‑level + AI knowledge and pursue high‑end R&D.

550–630 (regular 211 / first‑tier): recommended – Software Engineering (architecture / AI) or Data Science. Strategy – avoid coding‑only focus; emphasize system design and AI collaboration.

Below 550 (second‑tier / associate): recommended – hybrid “Computer Science + X” (e.g., medical, finance, IoT), Security, Operations, Testing. Strategy – avoid pure SE; choose AI‑hard‑to‑replace practical roles.

For scores below 550, a “Computer Science + medical/finance/IoT” hybrid is preferable to a pure software‑engineering track.

University selection criteria

Curriculum includes low‑level or AI content.

Presence of industry‑university labs.

Graduate outcomes – availability of high‑end positions.

Common pitfalls during four‑year university study

Pitfall 1: Focusing only on “how to write code”

Wrong: daily Java/Python syntax drills, CRUD projects, memorizing interview questions.

Correct: master data structures, algorithms, operating systems, networks, compilers – core abilities AI cannot replace.

Pitfall 2: Treating AI as a “cheat tool”

Wrong: rely entirely on AI to write code without understanding logic, debugging, or optimization.

Correct: use AI as an assistant while remaining the decision‑maker for requirement breakdown, architecture, code review, and performance tuning.

Pitfall 3: No projects or internships

Wrong: four years of only classes and exams, no real‑world projects or big‑tech internships.

Correct: start real projects (open‑source/GitHub or enterprise) in sophomore year; aim for a big‑tech internship in junior year – project experience outweighs a diploma in the AI era.

Final guidance

Students well‑suited for the major

Strong math/physics foundation (Math 110+, Physics 70+), logical thinking, passion for solving complex problems.

Genuine enthusiasm for technology, enjoys programming and experimenting with new tools, willing to engage in lifelong learning.

Clear career goal toward top tech firms, AI R&D, or architecture design, prepared for intensive early effort for long‑term high salary.

Students who should reconsider

Pure trend‑followers attracted by high salary but disliking math, logic, or coding.

Those seeking an easy, low‑effort path (9‑to‑5, no overtime, minimal learning) and cannot cope with rapid tech iteration.

Nature of the computer‑science profession in the AI era

Previously, a programmer was defined as “someone who writes code”.

Now, a programmer is “someone who commands AI to write code, designs systems, and solves problems”.

In 2026, computer science / software engineering is neither a dead end nor a guaranteed win; it remains a hard‑core field where strong candidates thrive and weaker ones are eliminated.