Chatgpt services in webapp continue to block interaction and answers since the GPT 4o updates of the UI, increasing Browser Memory usage and CPU

I would like to point out in a precise and technical way that the services that make up the chatgpt elements stop responding and working while chatgpt generates the response to the point of completely freezing the UI since there was the last update 2 weeks ago for GPT 4o and the new UI. it is obvious that some of the computation and load during the response goes to the browser both in terms of memory, load and threads, but this regardless of the computer and browser, creates slowdowns all the time and continuous freezes.

chatgpt services in webapp continue to block interaction and answers since the GPT 4o updates of the UI, increasing Browser Memory usage and CPU

I am experiencing significant performance issues and UI freezes on ChatGPT since the last update for GPT-4o and the new user interface released about two weeks ago. These issues occur exclusively when using the GPT-4o model, regardless of the type of computer and browser used.

Problem Description:

  • Frequency of the problem: Every time I use the GPT-4o model.
  • Browsers and operating systems used: [Insert details of tested browsers and operating systems].
  • Problem description: While ChatGPT with the GPT-4o model generates a response, there is a noticeable increase in memory usage and CPU load in the browser. This leads to a general slowdown of the user interface until it completely freezes, making the service extremely difficult to use.

Steps to reproduce the problem:

  1. Access ChatGPT through [specific browser].
  2. Select the GPT-4o model.
  3. Start a chat session and wait for a response from GPT.
  4. Observe the increase in memory and CPU usage, accompanied by a progressive performance degradation until the UI freezes.

It would be helpful to understand if this is a known issue and if there are specific steps I can follow to mitigate this inconvenience, or if any updates are planned to resolve this issue.