问HN：你认为Opus 5会有所改善吗？
Claude的Opus 5是一个近期推出的模型，作为Opus 4.8的替代品存在很多问题，包括忽视我们详尽记录的部署流程（在我们的短文档claude.md和架构文件中有清晰描述），并在激活后立即导致生产环境崩溃。
目前，我添加了一个监控层，让Opus 4.8实例来监督Opus 5并进行纠正。这稍微有所帮助。我在考虑是否降级到Opus 4.8，但也想给Opus 5一个机会。
你认为这个模型会有所改善，还是会停留在目前的能力水平？
你可以看到其他人的投诉：
https://www.reddit.com/r/ClaudeAI/comments/1v92csh/opus_5_extremely_rlfried_and_mistakeprone_for/
“在我看来，Opus 5在任何任务上都真的不好。我在一个大型复杂项目中彻底尝试了它在所有可能角色下的表现，结果它在所有任务上都很糟糕。”
Claudebot对该讨论串的总结：
> 该讨论串的压倒性共识是，Opus 5是一个有很多bug、过于自信的混乱产物，对于复杂编码来说是Opus 4.8的显著退步。
查看原文
Claude's Opus 5 is a recent model and has been very problematic as a drop-in replacement for Opus 4.8, including ignoring our well-documented deploy process (clearly described in our short claude.md file and short architecture file) and breaking production immediately upon being activated.<p>Currently, I added a supervisor layer where an Opus 4.8 instance supervises Opus 5 and corrects it. It helped a little. I am thinking of downgrading to Opus 4.8 but want to give Opus 5 a chance.<p>Do you think this model will improve or will it stay stuck at its current level of capabilities?<p>You can see complaints from other people:<p>https://www.reddit.com/r/ClaudeAI/comments/1v92csh/opus_5_extremely_rlfried_and_mistakeprone_for/<p>"Opus 5 is truly not good for any task, imo. i have thoroughly tried it in every possible role in a large, complicated project. it is bad for all tasks."<p>Claudebot's summary of the thread:<p>>The overwhelming consensus in this thread is that Opus 5 is a buggy, overconfident mess and a significant regression from Opus 4.8 for complex coding.