January 13, 2026
AI vs Code Bros: Gloves Off
The Insecure Evangelism of LLM Maximalists
Veteran coder says AI fanboys are projecting; comments explode over slop vs speed
TLDR: A seasoned dev says building software by chatting with AI is slow and sloppy, and claims the loud hype is insecurity. Comments split between “AI is average and best for simple tasks” and “adapt or be left behind,” with real stakes for job security and code quality.
In today’s tech soap opera, a veteran coder says he tried “agentic” LLM (large language model) coding—basically telling chatty AI to build and edit code—and found it slow, wrong, and soul‑sucking, then called the hype insecurity dressed up as futurism. The comments went full cage match. budududuroiu argues the career ladder often rewards “slop” that looks busy while hiding bugs—LLMs might just turbocharge that. rudedogg fires back that “These things are average text generators,” great when you don’t know a language, but rarely better than a focused human. IncreasePosts brings receipts: mitchellh vibe‑coded a one‑off visualization and shipped it. kiernanmcgowan adds nuance: AI nails cookie‑cutter pages but melts down on messy, bespoke tasks like tricky data migrations. Then accrual drops the gladiator line: adapt to LLMs or get left behind.
The jokes? People dunked on “token drain,” dubbed vibe coding “AI astrology,” and memed LLMs as a tireless intern copying Stack Overflow. Strongest split: Is AI a crutch for mediocre coders, or a legit sidekick for pros when used right? The vibe is defensive, flexy, and very online. Some demand proof, others post bug‑spotting humblebrags. Also: multiple folks asked if the author’s “cleanup” rates are still open
Key Points
- •The author finds LLMs useful for clerical support tasks but limited for coding unless context is small and instructions are precise.
- •Agentic, prompt-driven coding with LLMs required extensive supervision and produced slow, often incorrect code changes in the author’s experience.
- •The author acknowledges LLMs can enable less-experienced developers to build projects they might not otherwise create.
- •Vocal LLM proponents are described as framing adoption as inevitable and skeptics as fearful of becoming obsolete.
- •The author suggests some evangelism may stem from insecurity, while leaving open the possibility that effective agent use is a skill they may yet learn.