Fast slop or slow slop? One tweet nails AI coding's quality dilemma
MikkoH · x · 2026-09-23
Mikko offers a tongue-in-cheek choice: do you want to produce slop quickly or slowly? Fast slop means letting an LLM generate the entire codebase at once; slow slop means writing small, tidy PRs yourself, then watching AI agents gradually degrade the codebase through their reviews. The joke captures the paradox of AI-assisted coding: quality erosion happens either way — only the speed differs.
More from coding & agent
- Dev builds 100% local metasearch engine with DiffusionGemma on vLLM — bodonoghue85 · 2026-09-24
- Voice Agents Can't Tell Who's Talking: Voice Isolation Cuts WER from 23.3% to 6.2% — thetripathi58 · 2026-09-24
- Project Reportedly One-Shot by Claude Opus 5.5 With One Bug-Fix Prompt — Robert_Palgrave · 2026-09-24
- Dev spots aria-checked on a <button>: evidence vibe-coded LLM output isn't being challenged enough — jh3yy · 2026-09-24
- serve-sim adds iPhone Duo support with 3D model, camera dev, and agent-accessible tree inspection — Baconbrix · 2026-09-24
- Open-source Nautilo lets your Genie agent orchestrate Codex, Claude Code and Hermes — Dan_Jeffries1 · 2026-09-24