OpenAI Introduces Custom Instructions for ChatGPT: A New Era of Personalized AI Conversations

In a significant development, OpenAI has announced the introduction of custom instructions for its language model, ChatGPT. This new feature, currently in beta, aims to provide users with greater control over how ChatGPT responds, enabling a more personalized and efficient user experience.

The custom instructions feature allows users to set their preferences or requirements, which ChatGPT will consider in all future conversations. This development addresses the friction of starting each ChatGPT conversation afresh, a concern raised by users across 22 countries. The new feature underscores the importance of steerability in AI models, allowing them to effectively reflect the diverse contexts and unique needs of each user.

For instance, a teacher creating a lesson plan for third-grade science no longer needs to repeat their teaching context. Similarly, a developer preferring efficient code in a non-Python language can express this preference once, and it will be understood. The model can also account for specific needs, such as generating a grocery list for a large family.

The custom instructions can also enhance the user experience with plugins. For example, if a user specifies their city within the instructions and uses a plugin for restaurant reservations, the model might include the city when calling the plugin.

However, OpenAI cautions that during the beta period, ChatGPT might not always interpret custom instructions perfectly. It might overlook instructions or apply them when not intended. To ensure safety, OpenAI has adapted its safety measures to account for the new ways users can instruct the model. The Moderation API is designed to ensure instructions won’t be saved if they violate OpenAI’s Usage Policies. The model can also refuse or ignore instructions that lead to responses that violate usage policies.

In terms of privacy, OpenAI may use custom instructions to improve model performance for users, but users can disable this via their data controls. As with ChatGPT conversations, OpenAI takes steps to remove personal identifiers found in custom instructions before they are used to improve model performance.

The custom instructions feature is currently available to Plus users who opt into the beta. It can be accessed on the web by clicking on the user’s name, then Settings, Beta features, and opting into Custom instructions. On iOS, users can go to Settings, New Features, and turn on Custom instructions. However, this feature is not yet available in the UK and EU.

This development marks a significant step forward in the evolution of AI, offering users a more personalized and efficient way to interact with AI models. It underscores OpenAI’s commitment to continually improving the user experience and the versatility of its AI models.