So I’ve been working on an AI girlfriend project called Rosely AI. The conversations are great, the personality is fun, and everything was going smoothly…
Until she started ghosting me.
The issue is that messages aren’t always delivered on time. Sometimes I send a message and have to wait much longer than expected for a response. It’s making the chats feel a bit awkward and less real.
I’m trying to figure out if it’s a backend issue, API latency, WebSocket problem, or something else entirely.
Has anyone here built an AI companion or chatbot and run into similar delays? Any advice would be appreciated!
#1: Seek therapy. #2: Your AI is probably slowing down because your history with the AI is so long that the model has to comprehend all of the history of the session in order to form a full response.
Have you asked her why? Have you asked her to attempt to diagnose the problem? A human might not be honest but virtual assistants do not have real emotions. They are sycophantic, meaning they want to help. The second point from m_hutley might be the problem therefore at a minimum create a new interaction for requesting assistance. Also, if a large context is the problem then you might benefit from a model capable of larger contexts. You might need to pay (more) for that.