[theqoo] THE LATEST ON GEMINI, WHICH HAS TURNED INTO A TOTAL IDIOT

(T/N: This is the leaderboard comparing AI models, with the models ranked primarily by “Net Improvement.” It's measuring how much each model improves performance on some set of tasks, along with how often those improvements are considered successful.)

It's even getting beaten by DeepSeek, a Chinese-made AI.

To exaggerate a little, it’s practically becoming the next SimSimi at this point (T/N: SimSimi = Korean AI app).

original post: here

1. Right... I use the free version of ChatGPT to polish my writing, and the paid version of Claude for coding or accurate, up-to-date info. I only use the free Gemini for simple questions or calculations. The free Gemini doesn't even pull the latest info from web searches, so I couldn't use it as a substitute for Googling

2. It really seems to have gotten worse lately ㅠㅠ

3. I signed up for the annual plan and totally regret it

4. Even the free version of Claude is amazing at organizing documents

5. I feel like Gemini fell behind quite a while ago. They just kept churning out "Flash" models... Their image models aren't great either; comparing it to the latest versions of GPT or Claude feels almost insulting to those two. It's basically just an AI for casual chatting and creative writing.

6. Gemini is weird.. If I take a photo of my outfit and ask how it looks, it keeps generating an image instead

7. GPT is just so good...

8. It lags every time I use voice search, it’s infuriating. In the end, I just ask ChatGPT, and it gives me the answer in a second. Also, it really pisses me off when I ask something and it suddenly starts generating an image instead. Plus, there isn't much difference in information quality between the paid and free versions, so I eventually cancelled my subscription and just use the free version occasionally.

9. Seriously, Gemini has become a total moron.

10. With everyone switching to Claude, I hope Claude doesn't end up just like GPT and Gemini...

New post

✉ PC Stories