No Sloptober: A Month Without AI Help for Developers
No Sloptober is an October challenge to stop using LLM-based tools for the full month, at home and at work No Sloptober. LLM means large language model, the type of AI behind chatbots and coding assistants.
The site describes a suggested hard core option for people who want the strictest version. It is defined as using no AI or LLM tools whatsoever at home or at work. It is presented as a suggestion.
For that hard core version, the page lists what to avoid. The list includes no AI search summaries, no chatbot chats and search, and no code review. It also names Claude, OpenCode, Pi, Agents, Models and related systems.
The page also states a principle it calls Onarheim's Law: "Agents can only maintain or INCREASE entropy in a system, humans are uniquely capable of decreasing it." No Sloptober
Participation is meant to be public and written up afterward. The page encourages participants to post with #no-sloptober on socials and to write a blog post about their No Sloptober experiences.
The broader context here is how routine this AI help has become in software work. Chat search has taken the place of manual documentation lookup and forum threads. Autocomplete tools now fill in routine code and test scaffolding, the basic structures developers build on. Agentic coding loops, tools that can carry out multi-step coding tasks on their own, draft fixes, explain error messages, and summarize proposed changes for review. Removing those layers for a month does not bring back an older normal. It shows which parts of the code a developer truly understands without assistance.
In my view, the telling result will not be lines of code shipped. It will be where time goes. Without automatic summaries, gathering context is slower. Reading source code directly, tracing how functions call each other, reproducing bugs on a local machine, and waiting for a colleague to review work all take attention. The tradeoff is closer contact with failure modes and design limits that summarized answers tend to smooth over. For senior staff, that friction can show dependence on prompts in areas assumed to be solid knowledge.
Looking at what this means for teams, the hard core rule has a practical advantage. Partial limits need constant interpretation. A complete ban for a set period is easier to follow and easier to compare across participants. The hashtag and the follow-up post matter here. They turn separate efforts into shared notes on AI-assisted development, review quality, and debugging stamina. The long-term payoff is not abstinence itself. It is better judgment afterward about when to hand triage and drafting to models and when to stay close to the system being changed.
