Unfortunately, for the last couple of months, ChatGPT-4o and 4 have both been facing a serious answer quality issue.
For example, I’m asking it to organize a text in sectors, not removing information, and it removes a whole lot. I ask about it, it acknowledges, and does the same but with other pieces of content.
Basic actions are being a big too much for it. It can get some serious codes for some tasks, but cant blend a text, or organize it.
There was a time where I could trust ChatGPT to blend two texts, for example, without losing info. Not anymore.
That being said, it’s been two months having stress with OpenAI. After a few tests, I now have to admit:
We can’t rely 100% on ChatGPT to do basic tasks (I dont mean only getting hard answers etc). I’m signing up Claude, it’s getting us better responses. It understands better, and answers better also.
( I dont even want to mention the daily bugs we’re facing since november, like when where we are typing some prompt for like 5min, 10min, and all of sudden the Chat field generates an error and we lose our prompt lol )