January 30, 2026
Code’s cheap, clapbacks aren’t
Code is cheap. Show me the talk
Dev world meltdown: AI spits code, humans mop the mess
TLDR: An essay says AI code generators mean the old way of programming is ending. Commenters fired back: cheap prototype code isn’t the problem—quality, process, and care are; some cheered that knowing syntax is “a dead business,” others warned human judgment still does the heavy lifting.
An essay declared the old way of building software is over, citing Linux legend Linus Torvalds and the rise of LLMs—“large language models,” AIs that can spit out code on command. The author points to Torvalds merging AI-written bits and says code is now cheap; the real work is talking through plans. Cue chaos. Commenters split into camps: the “code is cheap, cleanup isn’t” crew, and the “syntax is dead, skills shift” believers. One veteran snapped that truly good, reliable code still comes from humans fighting entropy, not magical bots.
Another blamed years of sloppy software culture, saying AI merely turns the volume up on bad builders. A third torched the headline as clueless, reading it as “code is worth less than talk.” Big Tech vibes rolled in too: for big products, writing code is a tiny slice; testing, safety checks, and rollout eat the calendar. The memes? “AI intern” jokes, “ship now, fix later” bingo cards, and bold claims that remembering syntax is “a dead business.” The vibe: code may be cheap, but quality, accountability, and taste are still very expensive. Meanwhile, the author’s bold “end of history” vibe drew eye-rolls and applause in equal measure from tired engineers.
Key Points
- •The article argues that LLM-assisted coding has fundamentally changed software development practices.
- •It contextualizes Linus Torvalds’ 2000 maxim about code’s primacy and cites his recent use of AI-generated code in a toy project.
- •Historically, software development is described as high effort and constrained by human and coordination limits, making prototyping costly.
- •The author asserts that traditional approaches are “over,” with AI tools lowering the cost of trying and building software.
- •A historical overview traces major shifts in internet and software: connectivity, languages, web tech, browsers, mobile OS, open-source platforms, and tooling.