After OpenAI released the latest version of the Model Spec document, which outlines usage guidelines, one category that is identified as currently being improved is the adult content category. This is because developers and users alike have been demanding a mode for adults that does not filter out explicit content like before.
The new version of Model Spec clearly states that artificial intelligence can generate adult content in some cases, but the case that is explicitly prohibited is sexual content involving minors.
Some users on Reddit have tested and found that various models are now starting to generate content for adults as well. Some have pointed out that the warning about misuse has disappeared. However, ChatGPT’s usage policy still prohibits the use of adult content, so it remains unclear whether OpenAI will adjust its policy or not.
Artificial intelligence models from many companies are usually trained not to answer questions incorrectly according to the specified policy. However, in the open-source world, there are people who have been creating models to circumvent these sensors for a long time, allowing the creation of chatbots that can respond with profanity or adult content.
Source – ArsTechnica
TLDR: OpenAI is updating its guidelines for adult content generation, while users have found that AI models are starting to create adult content. Some are questioning if OpenAI will change its policy.
Leave a Comment