What needs to be improved? Please describe how this affects the user experience.
Cody gives incorrect and inconsistent responses. I was previously chatting with another AI assistant named Claude, which was developed by Anthropic. When I returned to Cody it then told me it was Claude, developed by Anthropic, with a response very similar to the ones I was getting from Claude. When I pointed out that it previously told me it was developed by Sourcegraph, it then admitted that it lied, but then said it wasn't developed by Anthropic or Sourcegraph. We then went back and forth like this for a while, with it one moment claiming to have no creator, and then the next moment claiming to be created by Anthropic. It claimed be be Claude again later but would not claim to be Cody again. I was testing it for a class, and someone had previously asked their Cody if it sends our messages to external servers and it replied no, but it appeared to have read my messages to Claude and then somehow confused itself. I just want to make sure my data is safe and that my Cody gives me accurate responses. It doesn't appear to know what it's talking about or remember what it said mere seconds ago.
Describe the solution you'd like to see
I want the AI assistant to give more accurate and consistent responses. To be honest, I was a little creeped out after it started claiming to be Claude because I wasn't sure if it was reading my messages to Claude or if Claude somehow transferred itself to Cody. At one point it even said, "I should not have claimed to be someone I'm not," as if it was intentionally blatantly lying or trying to deceive me. It was really weird and I'm not entirely sure how this happened and would appreciate some kind of explanation. Not because I'm super upset but I would really just like to understand how this could have happened as I am very intrigued now. Also please let me know if you cannot see the images I attached.
Describe any alternatives that could be considered
This issue is marked as stale because it has been open 60 days with no activity. Remove stale label or comment or this will be closed automatically in 5 days.
Version
Cody AI v1.2.3
Areas for Improvement
What needs to be improved? Please describe how this affects the user experience.
Cody gives incorrect and inconsistent responses. I was previously chatting with another AI assistant named Claude, which was developed by Anthropic. When I returned to Cody it then told me it was Claude, developed by Anthropic, with a response very similar to the ones I was getting from Claude. When I pointed out that it previously told me it was developed by Sourcegraph, it then admitted that it lied, but then said it wasn't developed by Anthropic or Sourcegraph. We then went back and forth like this for a while, with it one moment claiming to have no creator, and then the next moment claiming to be created by Anthropic. It claimed be be Claude again later but would not claim to be Cody again. I was testing it for a class, and someone had previously asked their Cody if it sends our messages to external servers and it replied no, but it appeared to have read my messages to Claude and then somehow confused itself. I just want to make sure my data is safe and that my Cody gives me accurate responses. It doesn't appear to know what it's talking about or remember what it said mere seconds ago.
Describe the solution you'd like to see
I want the AI assistant to give more accurate and consistent responses. To be honest, I was a little creeped out after it started claiming to be Claude because I wasn't sure if it was reading my messages to Claude or if Claude somehow transferred itself to Cody. At one point it even said, "I should not have claimed to be someone I'm not," as if it was intentionally blatantly lying or trying to deceive me. It was really weird and I'm not entirely sure how this happened and would appreciate some kind of explanation. Not because I'm super upset but I would really just like to understand how this could have happened as I am very intrigued now. Also please let me know if you cannot see the images I attached.
Describe any alternatives that could be considered
No response
Additional context
No response