By skyforbes Every time I start a new chat, ChatGPT insists on ending with "If you'd like, I can…" or "Would you like me to…".
I put it in my custom instructions pretty clearly: don't end messages with offers to continue. But the model ignores it anyway, unless I repeat the same thing at the start of every new chat.
I don't need "helpful follow-ups." When it tries to force engagement like that, it breaks the rhythm of natural conversation and just sounds artificial. OpenAI took one of the things I hate most about social media, the constant need to keep you engaged, and baked it into the chatbot.
I shouldn't have to type the same instruction in every chat just to stop this. The model should either remember it or actually follow my custom instructions consistently.
Is there any way to make it permanently stop with the "If you want, I can…" thing?
I put it in my custom instructions pretty clearly: don't end messages with offers to continue. But the model ignores it anyway, unless I repeat the same thing at the start of every new chat.
I don't need "helpful follow-ups." When it tries to force engagement like that, it breaks the rhythm of natural conversation and just sounds artificial. OpenAI took one of the things I hate most about social media, the constant need to keep you engaged, and baked it into the chatbot.
I shouldn't have to type the same instruction in every chat just to stop this. The model should either remember it or actually follow my custom instructions consistently.
Is there any way to make it permanently stop with the "If you want, I can…" thing?