wraith rivet
Very frustrating that it just lies about data it doesnt know, and come on guys, literally this is what you are telling companies the value of your software is but then you yourself dont use it this way.
Answering questions about openai API, gpt says lies, 100% lies that it knows is lies, and all it does is waste my time. For example, I just want to know how to use the new API because changing the model to what is listed in the model list isnt enough, I think I have to change completions to chat as well, and maybe some other stuff but chatGPT said
"Choose the GPT-3.5 engine ID that you want to use. The available engine IDs for the GPT-3.5 API are:
text-davinci-002: The standard GPT-3.5 engine for text generation tasks."
And it said "You are correct that the engine ID for the GPT-3.5 "Turbo" model is not davinci-003, as I previously stated. The correct engine ID for the GPT-3.5 "Turbo" model is davinci-005"
If you are trying to showcase how this can alleviate companies support needs on new releases, this is an epic fail.
PLEASE train GPT on your API updates before you release them to the public, its better for your customers, its better for your image.
fallen kernel