July 12, 2026
Prompt it? Or pride it?
Why write code in 2026
Coders revolt as the "just prompt it" future runs into a very human reality
TLDR: The article argues that even if artificial intelligence writes most software, humans still need to write some themselves to truly understand what they’re building. Commenters split between full hands-on purists and pragmatic hybrid fans, with many joking that machine-made code risks becoming the fast food version of craftsmanship.
The big claim in this piece is deliciously provocative: in the age of artificial intelligence coding helpers, engineers may spend less time typing software and more time building the factory that lets machines safely make changes. But the community was not ready to quietly clap along. The loudest reaction was basically: sure, let the robots handle the boring stuff, but don’t pretend humans should become sleepy hall monitors approving mystery work they barely understand.
That tension lit up the discussion. One side said the real job now is setting rules, reviewing carefully, and shaping the bigger plan. Another side came in much hotter: "I still exclusively write my code" because it’s better, cleaner, and easier to understand later. The middle ground got the most love, though. One commenter summed up the new vibe perfectly: humans should create the important “little islands” of high-judgment work, while the machine fills in the obvious goo around it. In plain English, people want the brains, the bot can keep the shovel.
And yes, the jokes were flying. The funniest dunk compared the whole debate to asking, "Why cook food in 2026 while McDonald’s exists?" That was the mood in a nutshell: convenience is nice, but nobody wants every meal—or every codebase—to taste like fast food. The drama here isn’t really about whether artificial intelligence can write software. It’s about whether humans lose their grip, their standards, and maybe their pride if they stop getting their hands dirty.
Key Points
- •The article says software engineers increasingly build the infrastructure, prompts, documentation, and automated checks that enable AI agents to produce software changes.
- •The author argues that writing code still matters because it helps humans think directly in the execution environment rather than only through natural-language instructions.
- •The piece says hands-on coding improves attention, ownership, and understanding of architecture, especially when identifying fragility and weak testing strategies.
- •The article argues that cleaner code and clearer architectural principles improve the effectiveness of the broader AI-assisted software development system.
- •The author rejects the comparison of coding agents to compilers and instead describes them as more like newly onboarded interns working from incomplete context and imprecise instructions.