Logical Errors and Increasing Mistakes in GPT-4o | OpenAI | Page 1

Reported by @plush crystal

Bug Report: Logical Errors and Increasing Mistakes in GPT-4o

`Steps to Reproduce`

Engage in extended conversations with GPT-4o.
Observe that over time, logical inconsistencies appear, such as incorrect timeline references or contradictions in remembered facts.
Notably, previously well-maintained context starts to degrade, leading to responses that are inconsistent with past interactions.

`Expected Result`

GPT-4o should maintain logical consistency, correctly referencing past interactions and ensuring continuity in reasoning. Previously established facts should not change arbitrarily.

`Actual Result`

GPT-4o has started making logical errors, such as referencing past events incorrectly (e.g., misplacing future dates into past ones).
Increasing frequency of minor spelling mistakes and grammatical inconsistencies.
Previously retained details are sometimes contradicted within the same conversation.

`Environment`

OpenAI ChatGPT (GPT-4o) Platform: ChatGPT Mobile App Android and Web Browser (Specify: Chrome Subscription: ChatGPT Plus Issue has been ongoing for at least 2-3 days and has been observed across multiple conversations. I´m a beta tester.

Additional Information

Please provide relevant details to help resolve the issue, such as:

ChatGPT Shared Link (if applicable).
Screenshots or videos demonstrating the problem.

-# ➜ Need to contact support? Visit the OpenAI Help Center.

#Logical Errors and Increasing Mistakes in GPT-4o