My bold opinion:

The people who are now wildly criticizing Claude are the same ones who previously wildly criticized GPT, just like the behavior of criticizing Android while using an iPhone, or wearing Adidas and criticizing Nike.
Now that large models are so powerful, every time a new model is released, people test it with a quick brain teaser or just use it for a couple of days to do a subjective evaluation, and then conclude: xx is trash, xx is a god.
In fact, all models have their own strengths, and their capabilities have already far exceeded the boundaries of 99% of people's abilities.
Some use AI to solve difficult math problems, like Brother Jie using AI to pick very impressive stocks, but most people use it for basic tasks, including myself, just to satisfy curiosity or write some code.
In this situation, what's the point of fussing over who is better or worse?
As long as you know how to use it, both GPT and Claude can meet your needs.
The boundary of AI's capabilities has long been outside of AI itself, and depends on the user.
I have been very convinced of this view since Claude 3.7.
There’s no real difference between the Heaven Sword and the Dragon Slaying Saber; for us who burn firewood with golden nanmu wood, just using what’s convenient is enough.
At most, find a cheap one that matches your aesthetic, and that’s enough.