Well, as I said in a comment in another thread, an LLM that doesn’t descend into complete gibberish after around 25 messages would be nice, as would one that consistently understands at least the system prompt. Until we can do at least those two things this will only ever be a toy IMO.

Yeah, ok. Wake me when there’s at least one open-source LLM that doesn’t descend into complete gibberish after around 25 messages or so, and consistently understands at least the bloody system prompt.