OpenAI has recently announced a new feature for ChatGPT: custom instructions. This feature allows users to set their preferences or requirements for ChatGPT, which the AI will consider when generating responses. The aim is to make the AI more adaptable to the unique needs of each user.
A Step Towards Better User Experience
The introduction of custom instructions is a response to user feedback about the need for more steerability in AI models. Users across 22 countries have expressed the need for models that can better reflect their unique contexts and needs. With custom instructions, ChatGPT will consider your preferences for every conversation going forward, eliminating the need to repeat your preferences or information in every conversation.
Practical Applications of Custom Instructions
The practical applications of this feature are numerous. For instance, a teacher crafting a lesson plan no longer has to repeat that they’re teaching 3rd grade science. A developer preferring efficient code in a language that’s not Python can specify this once, and it’s understood. Even grocery shopping for a big family becomes easier, with the model accounting for 6 servings in the grocery list.
Improving Experience with Plugins
Custom instructions can also improve your experience with plugins by sharing relevant information with the plugins that you use. For example, if you specify the city you live in within your instructions and use a plugin that helps you make restaurant reservations, the model might include your city when it calls the plugin.
Beta Testing and Safety Measures
While the feature is still in beta, OpenAI acknowledges that ChatGPT won’t always interpret custom instructions perfectly. There might be times when it overlooks instructions or applies them when not intended. However, OpenAI has adapted its safety measures to account for the new ways that 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 these policies.
Privacy Concerns
OpenAI may use your custom instructions to improve model performance for its users, but you can disable this via your 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.
Conclusion
The introduction of custom instructions for ChatGPT is a significant step towards making AI more user-friendly and adaptable. By allowing users to set their preferences, OpenAI is making AI more accessible and useful in a variety of contexts.
FAQs
- What are custom instructions in ChatGPT? Custom instructions are a new feature that allows users to set their preferences or requirements for ChatGPT, which the AI will consider when generating responses.
- How can custom instructions improve my experience with ChatGPT? With custom instructions, you can specify your preferences once and ChatGPT will consider them for every conversation going forward. This eliminates the need to repeat your preferences or information in every conversation.
- Are there any safety measures in place for custom instructions? Yes, OpenAI has adapted its safety measures to account for the new ways that users can instruct the model. The Moderation API is designed to ensure instructions won’t be saved if they violate OpenAI’s Usage Policies.
- Can OpenAI use my custom instructions to improve model performance? Yes, but you can disable this via your data controls. OpenAI takes steps to remove personal identifiers found in custom instructions before they are used to improve model performance.
- Is the custom instructions feature available to all users? The feature is currently available in beta starting with the Plus plan, and will be expanding to all users in the coming weeks.
