Hi! I use the google generative AI integration, developed by tronikos. I told it to speak like TARS from interstellar, which means the responses may be the ones of a dkhead(swearing etc). Most word, but sometimes i get a response saying "the message got blocked by your safety settings". A quick google led me here https://ai.google.dev/api/generate-content#v1beta.HarmCategory , which suggests this is not an issue in my google AI dashboard, but in the way the requests are made to the API within the integration. Is there any way to disable safety settings? I don't care how offensive my assistant gets
#Google AI blocks messages due to safety settings?
red ridge
yea, definately not a setting in google AI studio: https://stackoverflow.com/questions/77947637/how-to-disable-safety-settings-in-gemini-vision-pro-model-using-api
vocal nexus
Think there's settings in the integration to change these
Google likes to change their apis without telling people, they probably change levels the api accepts, had happened before 🤣
Though if you are trying to get swears the defaults in the integration may have the safety settings to block that (to be family friendly) 🙂
