With replacement of ChatGPT-o3 with ChatGPT-o4 and replacement of ChatGPT-4 with ChatGPT-4o
It seems that we reached a point that only dumber models are left until plus subscribers are only entitled models of similar performance to ChatGPT-3.5.
Is OpenAI trying to convey the message that they no longer find the subscription beneficial for them and we would be better spending that money on other services?
Also, did anybody find success in getting better responses of those new models?