Astra is good but its not AGI
一位用户在连续使用两天后，认为Astra是一个非常有能力的模型，并赞扬其简洁性优于其他模型。然而，他们指出Astra犯了一些基本错误，并且没有展现出OpenAI所宣传的惊人记忆力。尽管该用户计划继续将Astra作为他们的首选模型，但他们不认为它已经达到了通用人工智能（AGI）的水平。
- 发布
- 2026年9月6日 12:45
- 来源类型
- 开发者社区
- 档位
- 社区
- 信源状态
- 正常
时间以 UTC 显示
更多信息
After two days of solid use, here are a few of the basic things Astra got wrong.
- I would give it a command and it would tell me how it was going to implement it but then…. Nothing. I would ask it again and it would agree that it didn’t start the process and yet it still would not start the process. It would take several rounds to finally get it to start doing something.
- I had to redesign a website and overall I was impressed, but there were tons of copy errors, spacing issues, color mismatches, layout mistakes. And I’m not talking about complicated things here I’m talking about very obvious stuff.
- it would get confused about what it had just done only a few moments earlier, it forgot about the whole website redesign and started making changes to the old site, even though the new one had been merged to main.
I think its a very capable model but I have not witnessed this incredible memory that OpenAi is bragging about. I do like its brevity compared to other models and frankly is fantastic. I will continue to use it as my top model for sure but AGI? I havent seen it.