OpenAI keeps making the product worse. Here are the specific ways it has gotten worse in just one year:Ads were introduced for free accounts (that part is fine, it is clear why).Voice translation into foreign languages used to work normally, but now it does not. GPT stubbornly does not understand the command “only translate in this chat.” It starts spouting nonsense and adding things on its own. So, in fact, the function used to work, and now they have broken it (perhaps even intentionally to save resources). This happened because of an update.The introduction of limits on image generation for paid accounts. This is complete nonsense. 🙄 Your GPT produces, at best, one decent image out of 30 generations. That is, the limit is not, for example, on saving an image, but basically on nothing. The principle seems to be: “We have become more popular, so now we do not have enough money, and we are going to charge you for whatever we can. Next time, we will probably make a Pro plan for the number of letters you can type.”Censorship, censorship, and censorship again. And very often it is false censorship, triggered when you are simply asking ordinary questions.Conclusion: the owners are making the product worse, saving resources, introducing restrictions, while at the same time raising prices. In the end, I recommend not fixating on GPT just because of its good advertising, but considering other good products as well. Pay attention to Chinese and European developers. It is not worth it: paying more and getting less, and getting worse quality.