Process vs. Result
Yesterday, while talking with a friend who works in AI (NLP) about the astonishing capabilities of ChatGPT, I was struck by how it evolves and upgrades like a monster. No matter what kind of question you throw at it (except for those heavily subjective), it can respond convincingly, and the better your prompt, the more accurate its answer. A few days ago, I upgraded to version 4.0, which supports online document reading, images, etc., and it has already started becoming more ecosystem-like. I realized that much of my work will be replaced by it. I jokingly said that if I keep using it, I might become a person who has lost the ability to think independently. However, I also noticed that while its answers are good, I don’t get that exhilarating feeling of having truly achieved something. When my friend said: "ChatGPT, to a large extent, uses human feedback reinforcement learning through dialogue, which diminishes our direct experience of achieving something." I suddenly remembe...