停止让模型变得更聪明
我喜欢使用“更简单”的模型（比如 Composer 1.5，可能是经过微调的 Qwen），尽管这需要我提供更具体的指导。<p>Claude Opus 4.6，可能是目前最“聪明”的模型，在“一次性完成”方面表现出色——但我发现我很少想要一次性完成任何事情，除非是非常小的离散元素，而简单模型也能一次性完成这些任务。<p>有趣的例子是告诉 Claude：“制作一个关于达达主义青蛙的游戏，使用一种投注机制，你试图成为池塘中最富有的青蛙。不要犯错。”而 Claude 会执行出一个可以运行的东西，或多或少符合你的要求。<p>在实际工作中，我喜欢简单模型的知识面不那么广。我注意到，由于缺乏知识，它们在网络搜索方面的使用更为积极。（尽管通过系统提示可以改善工具的使用。）我还觉得简单模型的错误观点较少。Claude 往往会草率下结论——因为它更聪明（而且有点过于自信？），几乎需要更多的“保护措施”。<p>我很好奇其他人的体验如何。
查看原文
I like to use 'dumber' models (like Composer 1.5, which is possibly a fine-tuned Qwen) even though it requires much more specific outlined guidance from me.<p>Claude Opus 4.6, which is probably the 'smartest' model out there, is great at "one-shotting" - but I find I rarely want to one-shot anything, unless its a really small discrete element, which a dumb model can one-shot as well.<p>There are funny examples of telling Claude "Make a game about Dadaist frogs that uses a betting mechanic where you try to become the richest frog in the pond. Make no mistakes." And Claude executes something that runs, and more or less is what you asked for.<p>For real work, I like that the dumb models don't know as much. I noticed they use web search more agressively because of their lack of knowledge, for example. (Although tool use can be improved via system prompts.) I also feel that dumber models have less bad opinions. Claude tends to jump to conclusions - it almost needs <i>more</i> guardrails because it's smarter (and a little overconfident?)<p>Curious what others' experience is.