It seems to me that the 2-20% of use cases where ChatGPT Agent isn't able to perform it might make sense to have a plug-in run that can either guide the agent through the complex workflow or perform a deterministic action (e.g. API call).
It seems to me that the 2-20% of use cases where ChatGPT Agent isn't able to perform it might make sense to have a plug-in run that can either guide the agent through the complex workflow or perform a deterministic action (e.g. API call).