I want to revisit this topic since it hasn’t been talked about much since 2023. According to GPT4 it assists in content moderation with human interaction. However, given the speed of response this is unlikely. While I am not generating porn or NSFW literature it even flags normal human interactions as “possible violations.” While I was willing to accept image generations (common to a paperback novel) might be too risqué for the model. Trying to write dialogue that isn’t being flagged is getting to be annoying. Right now I’m having to use other platforms any time two adults of the opposite gender enter a room. This is silly. Worse is that even ChatGPT4 is recommending to use other vendors. Is this really the best case business model?
While I am not generating porn or NSFW literature it even flags normal human interactions as “possible violations.”
I was doing quadratic equations practice, and GPT was laying the cheerleading on a bit thick. It posed a problem, saying something like “Let’s see what we can squeeze out of x! What’s your solution to this, lay it on me!” and that was enough to get the red “possible violations” warning. /smh
Literally, algebra is too risqué… ha ha!
1 Like