The information provided for customizing responses will be used to train OpenAI's API models. Users can opt out of this setting through their data control settings. The feature has been tested with some users and allows them to specify their chat preferences. Once set, the changes take effect from the next session, with a limit of 1,500 characters for responses. OpenAI uses its moderation API to ensure the safety of customized instructions.
Key takeaways:
- OpenAI has launched custom instructions for ChatGPT users, allowing them to set context for the chatbot's responses without having to repeat the same instructions each time.
- The feature is currently available for Plus plan users, excluding those based in the EU and the UK, and can be accessed via the settings on the web and iOS.
- Information provided for customizing responses will be used to train OpenAI's API models, but users can opt out of this setting.
- OpenAI uses its moderation API to scan customized instructions for safety, and can refuse to save or ignore instructions that could lead to harmful or hateful responses from ChatGPT.