请问HN：有没有人正在尝试不同的方式使用大型语言模型（LLMs）进行编程？
我对使用大型语言模型（LLMs）进行编程时感到有些烦恼，因为我们似乎有些停滞不前。我使用Claude Code和Codex，但我还无法像手动编写代码时那样进入流畅的状态。
这让我觉得有些讽刺，因为人工智能应该是思维的助推器，但现在感觉就像是一辆每隔几分钟就会突然刹车的自行车。我停下来，等待，回顾，然后再次提示。
有没有人正在探索与我们今天的提示响应循环根本不同的东西？
我实际上认为标签模型的想法在方向上比提示响应更好。
我很想听听关于任何初创公司、个人实验等的消息。
查看原文
I'm a bit annoyed by the feeling that we're kind of stuck when it comes to using LLMs for programming.<p>I use Claude Code and Codex, but I haven't been able to enter flow state like I can when I hand write code.<p>This is kind of ironic to me since AI should be a bicycle for the mind, but right now it feels like a bicycle that just brakes abruptly every couple minutes. I stop, wait, review, prompt again.<p>Is there anyone exploring something fundamentally different than the prompt response loop we have today?<p>I actually think the idea of a tab model is directionally better than prompt response.<p>Would love to hear about any startups, personal experiments, etc.