Quality Answer BAD, Result not good

To the OpenAI Customer Support Team,

I would like to file a complaint regarding the issues I have encountered while using the ChatGPT model versions o1 Mini, o1, o1 Pro, specifically the Pro version that I am currently using. Previously, I was able to rely on this model to provide accurate, complete, and well-structured code solutions. However, I am now facing several issues that significantly affect the effectiveness and quality of use.

My previous experience was very satisfying. The model was able to provide excellent results, with clear and detailed explanations for each part of the code functions. However, it seems that this capability is no longer available, and the results provided are now far from my expectations.

The main issues I am facing are as follows:

Poor Quality of Generated Code: The code provided is often poorly structured and full of syntax or logical errors, making it difficult or impossible to run correctly.

Inaccurate Calculations and Analysis: When I ask the model to perform calculations or analysis, the results are often not in line with my expectations and are highly inaccurate. This indicates that the model is not following instructions correctly when performing evaluations or calculations.

Incomplete or Irrelevant Code: When I request the model to provide complete code or follow instructions with clear steps, it often only provides partial code or code that is not relevant to other functions. This disrupts the programming flow and makes it difficult for me to continue my work.

Programming Language Inconsistency: Sometimes, despite requesting code in a specific programming language (e.g., Kotlin), the model instead provides code in another language (such as Python), which does not meet my needs, causing confusion and requiring time-consuming manual conversion.

It seems that this issue may be related to resource or module limitations used by OpenAI, which results in the model’s inability to accurately follow instructions and provide relevant and high-quality solutions.

I kindly request assistance from the OpenAI team to evaluate this issue and provide clarification regarding any limitations that may be affecting this model. I hope this problem can be resolved promptly so that my experience with the model can return to being optimal and in line with the expectations I had previously.

Thank you very much for your attention and assistance.

I experienced exactly the same thing since I upgraded to Pro (a week ago) and I was never able to get a solution, I don’t think it happens on all accounts as otherwise it would be more widespread but I gave up trying to use Pro temporarily, the results I had using o1 with Plus were much better than o1 on this plan.

There are some posts about it, but they didn’t have much impact:

After a week of trying in vain to get support to analyze the case, I could never understand why things that I see others doing with o1 Pro (even replicating steps I see in Youtube reviews), in my case they behave differently, which somewhat supports my hypothesis that it is a particular problem with some accounts, but anyway as I said, I never even managed to make any real progress with support, they just don’t care.