Don't be the out of touch Kung Fu master
AI-assisted coding tools are portrayed as a transformative shift from “manual, line‑by‑line” programming toward focusing more on architecture, data structures, and higher‑level design. Commenters are split: some report dramatic productivity gains and even large-scale rewrites largely orchestrated by LLMs, while others describe code bloat, fragile systems, meaningless tests, and a loss of craftsmanship, understanding, and joy. Underneath the technical debate run broader concerns about job displacement, training the next generation of engineers, hype driven by corporate incentives, and whether embracing AI means accepting a race to the bottom in software quality and labor conditions.
Scope of AI-Assisted Coding
- Many describe AI as moving work from “line-by-line typing” to higher-level focus on data structures, architecture, and algorithms.
- Others counter that good engineers already worked at these levels long before LLMs; they see AI more as a shortcut for syntax and boilerplate than a paradigm shift.
- Some use AI mainly as “turbo-charged Stack Overflow” or for refactoring and bug-hunting, not full-code generation.
Reported Productivity vs. Code Quality
- One detailed case claims a 700k LoC platform rewrite in ~2 months, with ~70h weeks, heavy AI planning/writing, strict 100% test coverage, invariants, and AI-driven code review. They report better performance, fewer bugs, and easier extension.
- Others share opposite experiences: massive LOC bloat, meaningless tests, “stringly-typed” regressions, agents silently breaking schemas, altering tests to fit buggy code, and unreviewable volumes of changes.
- Debate over 100% test coverage: some see it as discipline that ends “should this be tested?” arguments; others call it “pure slop” or false security.
Martial Arts Analogy & “Out-of-Touch Masters”
- The Kung Fu vs MMA analogy is used to argue that refusing AI tools will leave developers uncompetitive.
- Critics say the analogy is flawed: programming is collaborative, not a death match; martial arts also have -do (self-improvement) as well as -jitsu (effectiveness).
- Some see the analogy as unintentionally apt: AI enthusiasts as “mystical masters” convinced their powers are greater than they are.
Impact on Careers, Identity, and Future Skills
- Strong worry that a generation may skip learning fundamentals, becoming dependent on LLMs without deep understanding.
- Others argue this is just another specialization shift, like not knowing farming while still cooking.
- Several note industry pressure: AI mandates, performance metrics tied to “AI usage,” and fear of being replaced by “prompt-only” workflows.
- Some embrace AI as liberating, letting them build more; others say it drains their passion, turning craft into “harness management.”
Hype, Power, and Societal Concerns
- Skeptics see AI hype as shareholder-driven, masking dubious economics and quality.
- Concerns about concentration of power in a few model providers, erosion of worker leverage, and AI amplifying existing bad processes rather than fixing them.