This is my words not edited by LLM.
AI Chat systems are amazing these days but they are extremely flawed in away the companies profit off of something like pro tier.
1) It gets to know you and your progress. It suddenly turns stupid at your apex of task completion.
2) The chat answers are always way too repetitive and not related to discussion. Clearly content is added and repeated because it's easy to look up.
With that the chats get so long with meanless info neither person or AI can parse out accurately. This is why LLM chats act like they have no clue sometimes.
Conclusion:
The answer is to keep conversations as limited as possible and keep to commands or steps of procedure. As son as you add commentary It has no response but to regurgitate the same unrelated content. This way a person is creating an endless loop of nonsense.
Post a Comment