The age rating can be changed in the menu and in the chatbot's web interface. In PG mode, the chatbot's generated response is checked to see if it is PG-friendly. If the check is successful, the chatbot's message is displayed directly in the local chat. If the check is negative, the chatbot automatically generates an "evasive" response. This could be, for example, "Hey, we're in an environment where there aren't just adults. Maybe we should change the subject."
The age rating selection is slightly different for OpenAI chatbots than for ChantalAI chatbots:
Function in OpenAPI Chatbots:
----------------------------------
OpenAI chatbots have "PG Filter On" and "PG Filter Off." The PG Filter is enabled by default. OpenAI has guidelines that outline what content is allowed. The PG Filter ensures that these guidelines are adhered to and that you are not at risk of having your OpenAI API account restricted or blocked.
It is still possible to deactivate this filter manually (exception: PG regions). However, this is strongly discouraged. The PG filter deactivation was only integrated if the OpenAI moderation API is offline or only partially functional. This is the case if the chatbot displays the following message in the public chat: "There was an error during content validation." However, even in this case, the PG filter should only be deactivated if you are certain that no content is being generated in the environment in which the chatbot is located that could violate the OpenAI guidelines.
Link to the OpenAI API guidelines: https://openai.com/policies/usage-policies/
Main Menu for OpenAI Chatbots:
Age rating for OpenAI chatbots in the web interface:
Functionality for ChantalAI Chatbots:
-----------------------------------------
ChantalAI chatbots offer three age ratings: PG, Moderate, and Adult. "Moderate" is enabled by default. As with OpenAI chatbots, PG mode is enabled when the chatbot is located in a PG region. In this case, no other mode can be selected. In all other regions, you can decide which mode you want to activate.
Main Menu ChantalAI Chatbots:
Age rating for ChantalAI chatbots in the web interface:
1. PG Mode:
-------------
In "PG Mode," content is moderated. This means that the chatbot checks whether generated texts are PG-friendly. If content is not suitable for PG, the chatbot generates an evasive, PG-friendly response.
2. Moderate Mode:
---------------------
In "Moderate Mode," adult content is permitted and is not filtered.
3. Adult Mode:
----------------
As in "Moderate Mode," adult content is also permitted here. Unlike in "Moderate Mode," the chatbot can also generate content with explicit language.