As an ADHDer, I subscribed to the plus version as soon as I could (although the free version would have likely sufficed).
Customized it in Android app, turned memory on and planned to use that function a lot, to keep track of all my neurodivergent, scattered but brilliant ideas and projects.
Then realized I’m not really great at prompting, so I installed the GitHub APK and started using voice.
Much to my disappointment, and despite the voice FAQ reading that “In both standard and advanced voice modes, you can create and access memories, as well as access custom instructions”, this turned out not to be true.
Voice keeps creating standalone chats, which is very frustrating because it completely defies the purpose.
To make things worse, what’s now happening is that depending on how I give my input, from app chat or voice command directly, generated responses come out with 3 different voices.
This is terrible for me! I love Cove, and get really upset by that squeaky female or the young and idiotic-like boy voices that sometimes come out!
How can there be no consistency in voice model and memory access?
I raised the issue before, I was told to clear everything out and I did, but the loss of a year worth of conversations didn’t solve the problem.
I can’t bear to use it like this, it’s really not helping, and actually quite unnerving.
I understand that staff can’t keep up with millions of users, but there has to be a solution!
Why is this not working for me?