Switching Back to GPT from Gemini
- 发布
- 09/05 03:37
- 收录
- 09/05 06:00
- 来源类型
- 开发者社区
- 档位
- 社区
- 信源状态
- 正常
讨论趋势
百分比基于采集到的讨论信号，不代表新增评论数或独立参与人数。曲线仅用于同一话题在不同时段的比较。
一位用户在使用Gemini一个月后表示非常失望，认为与GPT相比，Gemini效率低下且不可靠。用户提到，Gemini经常无法遵循指令，例如多次要求生成PDF文件，但它几乎每次都表示无法完成并建议使用Google Docs。用户总结说，尽管对Gemini与Google生态系统的整合抱有更高期望，但实际体验令人失望，Gemini在工作流程中表现出较低的可靠性和效率。
Honestly, Gemini has been extremely disappointing for me lately. I’ve been using it for about a month, and even basic things often require me to explain the same instructions over and over again.
For example, I asked it around 10 times to create a PDF file for me. About 9 times, it simply told me that it couldn’t do it and suggested that I use Google Docs instead. The one time it actually managed to generate a PDF, the result was basically unusable and far from what I asked for.
The same problem happens with other tasks as well. It frequently forgets or ignores instructions that I’ve already explained multiple times, which makes working with it unnecessarily frustrating and time-consuming.
On top of that, the amount of usage it consumes feels unnecessarily high, especially considering how much I actually get done with each request. I often feel like I’m wasting my usage limit just trying to get Gemini to understand or correctly execute relatively simple tasks.
What makes the comparison even worse is that I’ve been using GPT extensively for similar work, and I haven’t even hit the usage limit once. GPT has been able to handle essentially all of the tasks I’ve needed, including more complex ones, without requiring me to repeatedly explain basic requirements.
So after using both for a while, my experience has been that Gemini is not only less reliable for my workflow, but also significantly less efficient with its usage. I expected Gemini to be much better, especially because of its integration with Google’s ecosystem, but so far, the actual experience has been pretty disappointing.