No Sloptober

No Sloptober is a month‑long challenge to abstain from large language model (LLM) tools entirely as a deliberate “fast for your mind.” The rule set ranges from a hardcore ban on any AI/LLM search summaries, chatbots, code review tools, agents and hosted models to softer guidance about avoiding being a “meat proxy” who launders work through chatbots without editorial thought. The argument rests on several concrete claims: agents tend to increase system entropy (Onarheim’s Law), LLMs can hinder learning, create security and malware risks, foster sloppy work, and contribute to mental-health issues or “AI psychosis.” The initiative frames abstention as a way to rebuild skills, regain flow, and better judge when automation actually adds value, while allowing a human translation exception and acknowledging corporate constraints.

Practical advice is tightly specific: review and clean up any existing LLM‑generated cruft in codebases; practice code katas; learn a new programming language (examples: Zig, Rust, Go, Lua, TypeScript); adopt tools like Neovim; try new frameworks (Astro/Mastro, Effect.ts); or build a game with Excalibur or Godot. Participants are encouraged to run experiments at work - measure velocity, incident rate, and costs - to quantify tradeoffs and use the iron‑triangle argument (good/cheap/fast). The campaign asks people to document results publicly, share badges or screenshots, and reflect honestly on capabilities regained or gaps revealed.