I’ve been using GPT-4 for various tasks, and while it’s generally impressive, I’ve noticed a recurring issue that seems to arise during longer conversations. After extended use, particularly when the model is tasked with searching the web, it occasionally becomes unresponsive or encounters errors. These problems seem to occur more frequently the longer the conversation continues, leading to significant delays and disruptions in the workflow. Has anyone else experienced this, and are there any known solutions or workarounds to prevent these interruptions?
1 Like
Welcome to the community!
As the page gets longer, your browser needs enough RAM to handle it, so it can get clunky with super-long threads.
Best to make a Custom GPT and start new threads for each task, maybe?
Thanks for the insight! That makes sense—RAM limitations could definitely be contributing to the issue, especially with longer threads. However, I’ve noticed that once the model becomes unresponsive, it often becomes completely unusable. I’ve tried clearing the page’s memory to see if it helps, but the problem still persists. Additionally, I’ve encountered another issue where responses from the model disappear the next day, leaving only the questions I asked without any replies.
I’m curious, are these problems common across GPT-4, or could they be specific bugs that the model currently can’t resolve? It’s been quite frustrating, and I’m wondering if there’s any way to manage or avoid these issues, or if it’s something that needs to be addressed in GPT-4 itself. Any further suggestions or insights would be greatly appreciated!