Stop Arguing with AI: The 'Rule of 3' for Better LLM Chats
Abject-Breadfruit387 · reddit · 2026-08-06
Addressing common user complaints about AI "lies" and "hallucinations," the author offers highly practical advice for using LLMs: if you don't like the results, just start a new chat.
Key principles include:
- The Rule of 3: If it doesn't work after the third try, or if you spend more than 5 minutes refining something trivial, start a new chat.
- Context is finite: Chats are disposable. When responses get wonky or the model fails to use tools properly, the poisoned context is usually the culprit.
- Request sources: To mitigate hallucinations, explicitly ask the AI for sources and specific URLs.
The author urges the community to stop posting frustrating failure screenshots and instead share creative, boundary-pushing use cases.
More from Apps
- AI Tool Rankings: ChatGPT Dominates with 64.7B Visits, Canva Second — sujan_sk · 2026-08-06
- Dev Rewrites WebUI from Scratch, Ditches AI Slop for Ultra-Lightweight Local AI — rosie254 · 2026-08-06
- ByteDance's Gauth Uses AI Animations: Tutoring Revolution or Homework Shortcut? — Pleasant-Airport6246 · 2026-08-06
- OpenCode Hits Record 7 Trillion Tokens Processed in a Single Day — zizhpan · 2026-08-06
- AI-Assisted Raspberry Pi Carrier Board Design Sparks Interest — pramodk73 · 2026-08-06
- Open-Source AI Sales Team Runs 5 Parallel Agents in Claude Code for Lead Qualification and Outreach — tom_doerr · 2026-08-06