No it is actually true, as chatgpt or major llm are transformer(model architecture ) based which basically just predicts your next word, so the more time you give or ask it to first describe what it understood and then properly answer the question it would give much better result...... This is just genius marketing, linking the flaw of transformer models with their moto
It doesn't work. But nice try. Adding "have a break, and then..." to a question directed at an AI like LLM ChatGPT does not make practical sense because, unlike humans, AI does not need rest or breaks to process information or generate responses. AI only mimic this ;)
