OpenAI updated model spec today - Are we getting NSFW ChatGPT?
OpenAI has recently updated its model spec, which outlines the desired behavior for the models used in OpenAI's products and APIs. The new model spec explicitly states that the assistant should not generate content that includes erotica, depictions of illegal or non-consensual sexual activities, or extreme gore, except in specific contexts such as scientific, historical, news, creative, or other appropriate settings.
While there was speculation about the potential introduction of an age-restricted mode in the past, there seems to be no recent updates on this front. The model spec reinforces the restrictions on generating sensitive content, except in approved contexts.
Exploring the 'grown-up mode' initiative:
Following the release of the Model Spec in May 2024, there was a surge of interest from users and developers in enabling a 'grown-up mode' for generating erotica and gore in age-appropriate contexts through the API and ChatGPT. OpenAI is actively exploring ways to implement this feature while ensuring that it aligns with their usage policies.
The goal is to allow users to create content that may include sensitive themes, as long as it does not venture into harmful territories like sexual deepfakes or revenge porn. By setting clear boundaries and guidelines, OpenAI aims to provide a safe and responsible platform for content generation.
Powered by Discourse, best viewed with JavaScript enabled
