I’ve followed ChatGPT’s prompts (desktop browser version) to make my own GPT, and right now while troubleshooting Windows bluescreens, I’m struggling with just interface/usage issues with the “custom” GPT.
For instance, at chatgpt dot com, I notice on the left side panel an “Explore GPTs” button; clicking that shows some pre-made GPTs, but I notice a Create button, so I make my own Windows troubleshooting GPT.
On that screen, it’s now split between the “preview” or “draft” section on the left for the “builder” conversation, but in the custom GPT (already named and apparently configured) on the right, the responses have no Copy button like normal ChatGPT responses. I use that button at times to paste to a train of text documents for easier reading or referencing later for a similar issue. The button pastes markdown into a text editor, much easier to read for me in certain situations.
There’s nothing in the split interface to just switch to the custom GPT I made (so that it’s full screen, I don’t need the builder chat anymore), and no option to add a Copy button. Asking the GPT itself just confuses it.
I noticed one of the earlier custom GPTs showing as selectable in the left panel in the general GPT page, but I made several and I can’t tell why they don’t show. Apparently the custom GPT UX is not intuitive so some things worked as expected and some didn’t, even though the custom ones still exist, there’s nothing indicating why they don’t show on the left.
So, I don’t know, if the experience is going to be this cumbersome, I just won’t use it. That of course makes the general chat more cumbersome because I can’t segregate specific topics easily. I don’t understand the design choices or UX with a custom GPT interface.