# How does the chat gpt api have so many bugs that are not fixed if they have access to the unreleased model? — retained record

> Retained reddit record from reddit:ChatGPT.

Canonical: https://highsignal.app/data/records/d2e14849f0f99fc2

[← sources](https://highsignal.app/data)


source: [reddit](https://highsignal.app/data/reddit) (reddit:ChatGPT)  published 2026-10-19 17:04 ingested 2026-10-20 02:35

[original source ↗](https://www.reddit.com/r/ChatGPT/comments/1xa2ifu/how_does_the_chat_gpt_api_have_so_many_bugs_that/)

## retained text



```
E.g. i often cannot branch chats that are inside of projects i have to remove them, then branch then put them back into the project.

And the voice(reading aloud) function doesn't work most of the time or it cuts off after about a minute of reading with no way of scrolling the bar and returning to where it left off. 

With a.i. based app development one should at least expect a quicker turnaround on such issues as it should be pretty cheap for them to fix.

Curious if anyone has actual insights on this, this is not (mostly not) a complaint post i am legitimately wondering. 
```



document fetched 2026-10-20 00:20 · canonical https://reddit.com/r/ChatGPT/comments/1xa2ifu/how_does_the_chat_gpt_api_have_so_many_bugs_that
