Everytime I ask Chat-GPT to write or modfy a python script it does this “analyzing” thing only to say there is an error ~75% of the time.
Feels like just a month ago everything was fine, I’d ask it for a python script and it would spit out the script.
There appears to be a bug in some new feature which is trying to execute every python script it writes while only showing the output .
I’ve noticed it too, but I’ve just changed my prompts slight to “Provide me with a script to {whatever I want it to do here}. I will execute it locally.”
Something along those lines usually works for me.
Yes I’ve added custom instructions to treat the symptoms but I wonder why this feature was added. I doubt any devs are relying on chatGPT alone to maintain their python environments. IMO seems like a completely unnecessary feature
In fact, this is annoying. ChatGPT struggles with knowing when we want code written.
A quick fix is to use ChatGPT Classic.
