问HN:大型语言模型的未来是什么?

1作者: apatheticonion5 天前原帖
我一直在使用 DeepSeek v4 的闪存功能来进行引导代理工作流程(在 IDE 中,提示/审查差异),并发现使用 Haiku、Opus 和 Sonnet 在这个工作流程中并没有明显的好处。<p>每天花费大约一美元,我的生产力可以翻倍。前沿模型是什么?它们试图解决什么问题?<p>这些模型是为了单次应用而设计的吗?我们是否期望能够“随意”提示大型语言模型(LLM),让它们在无人监督的情况下完成工作?
查看原文
I&#x27;ve been using DeepSeek v4 flash for guided agent workflows (in-IDE, prompt&#x2F;review diffs) and have found no noticeable benefit to using Haiku, Opus, Sonnet in this workflow.<p>For around a dollar a day, I am able to more than double my own productivity. What are the frontier models for&#x2F;what are they trying to solve?<p>Are they designed to one shot applications? Is the expectation that we want to be able to &quot;yolo&quot; prompt LLMs and have them complete work unsupervised?