Seeing how differently people prompt LLMs is funny

A Reddit user and their brother employ contrasting methods for prompting LLMs in coding. The brother uses GPT-6-Astra and frequently berates the AI for errors, believing it makes the model more compliant. In contrast, the user, who is autistic, uses a local GLM 5.3 Flash instance (q4 qat) and communicates with precise, detailed requests, viewing the smaller model with more patience and understanding. This highlights the diverse and often humorous approaches individuals take when interacting with different LLM technologies.