Skip to main content
HomeTutorialsArtificial Intelligence (AI)

How to Use ChatGPT Custom Instructions

Explore ChatGPT Custom Instructions feature. Learn how to fine-tune responses, explore use cases for teachers, entrepreneurs, and content creators.
Updated Aug 2023  · 7 min read

OpenAI ChatGPT is taking the world by storm. Week after week, the updates are continuously raising the bar of what you can do with this state-of-the-art language model.

Here, we take a deep dive into the recent announcement made by OpenAI on Custom Instructions for ChatGPT. This feature is initially available in beta to users with the ChatGPT Plus plan and will gradually be expanded to other users over the next several weeks. As per OpenAI’s release announcement, this feature is not yet available in the UK and EU.

You can learn how to use ChatGP, discover best practices for writing prompts, and explore common business use cases for the powerful AI tool by taking Introduction to ChatGPT on DataCamp.

Understanding ChatGPT’s Custom Instructions

Custom instructions feature makes ChatGPT more powerful by incorporating your preferences or specifications that you wish ChatGPT to take into account each time you generate responses. The model will consistently consider these instructions whenever it responds, eliminating the need for you to repeat your preferences or information in each interaction.

An example use-case can be, imagine a teacher devising a lesson plan without the need to repeatedly mention that they're teaching 3rd-grade science. Likewise, a developer striving for efficient code in Python can state their preference just once, and ChatGPT will consistently consider this.

The development of this feature was directly influenced by user feedback, which emphasized the significance of steerability for users. It serves a critical purpose in enhancing adaptability, enabling models to effectively adjust to various contexts and cater to individual requirements.

Setting Up ChatGPT Custom Instructions

You must be a ChatGPT Plus member to enable this feature. If you have a Plus account, you enable the feature:

1. Click the three dots next to your name, then click "Settings and Beta."

ChatGPT Settings

2. Enable custom instructions feature by enabling the “Custom instructions” toggle button.

ChatGPT Settings Custom Instructions On

3. Click the three dots again, and select "Custom instructions."

image4.png

4. Set up custom instructions

Custom Instructions Menu

Making the Most of Custom Instructions

In this section, we’ve outlined some examples and ideas to inspire you to use the Custom Instructions feature in ChatGPT.

1. Chat GPT Custom Instructions for teachers

For our first ChatGPT Custom Instructions example, let’s explore how someone working in academia could use the tool.

First, we give ChatGPT some context about our role, in this case, as a university lecturer. Next, we outline the tone and style we want it to use in its replies, as well as some guidance on the features it should include.

Custom Instructions in Use

Now, let’s see how ChatGPT responds when we ask it to perform a simple request:

Simple request using custom instructions

Notice how it has an abstract at the beginning and further reading at the end of the lesson. Without custom instructions, you will have to make this information part of the prompt every time.

In our free Webinar, A Beginner's Guide to Prompt Engineering with ChatGPT, Adel Nehme, Head of Media & Product Marketing at DataCamp and co-host of the DataFramed podcast, covers the importance of constructing clear and concise prompts that are tailored to meet specific workflows. Moreover, he shares best practices for maximizing the output of ChatGPT alongside use cases and case studies of how ChatGPT can be leveraged in marketing, sales, data, and other functional workflows.

2. ChatGPT Custom Instructions for Entrepreneurs

Now, let’s look at how a business leader or entrepreneur can use ChatGPT custom instructions to get the most out of the powerful feature. Below, we’ve included an example of the context and response guidelines that can help:

Alternative Custom Instructions

Learn how to use ChatGPT in a real-life end-to-end data science project. Check out our blog “A Guide to Using ChatGPT For Data Science Projects” by Abid Ali Awan on DataCamp.

Alternative Custom Instructions Response

3. ChatGPT Custom Instructions for content creators

For our final example, let’s explore how a content creator might use ChatGPT to get content ideas and guidance. Again, we can specify things like the appropriate tone, the aim of the content, and how we want readers to respond to it:

Second Alternative Custom Instructions

Second Alternative Custom Instructions Response

However, you’ll notice that the tone of the above is probably a little too extreme (quite sarcastic and arguably too conversational). In this instance, it might mean refining the custom instructions a little more, making them more suited to the output that you want. You can read our guide on ChatGPT prompt engineering for more information about how to refine your inputs. 

Custom Instructions and Plugins

Including additional instructions can enhance your interaction with ChatGPT plugins, as it allows you to share pertinent information with the specific plugins you utilize. For instance, if you provide your city's name in the instructions while using a restaurant reservation plugin, the model may incorporate that city when accessing the plugin.

Are you familiar with the ChatGPT Code Interpreter? If you don’t know, no worries. Check out our blog on how to use ChatGPT code interpreter for more information.

Custom Instructions Privacy and Data Usage

As per OpenAI’s website, they may use your custom instructions to improve model performance, but you can disable this via your data controls in settings.

ChatGPT Data Control Settings

Current Limitations of ChatGPT Custom Instructions

Custom instructions in the beta phase may not always produce accurate and expected results. This may get better with time. However, another limitation we currently see in the functionalities is a lack of multi-persona custom instructions. You can only have one saved custom instruction at a time which can be pretty inconvenient if you are moving back and forth between different personas.

Conclusion

In the ever-evolving landscape of OpenAI ChatGPT, the introduction of Custom Instructions heralds a new era of personalization and steerability. With this feature, users gain the power to fine-tune ChatGPT's responses according to their preferences and specifications, eliminating the need for repetitive instructions in each interaction. This capability enhances adaptability and ensures that the model consistently caters to individual requirements.

To make the most of this powerful feature, ChatGPT Plus members can enable Custom Instructions by accessing the settings and beta options. Once enabled, users can set up their personalized instructions and witness the model's responsiveness in action.

Custom Instructions offer a world of possibilities across various domains. Teachers can effortlessly design lesson plans, entrepreneurs can specify their preferences, and content creators can infuse their distinctive style—all with a touch of magic from ChatGPT.

And for those who love plugins, including additional instructions can enhance interactions and streamline tasks, making the AI-assisted experience even more delightful.

Discover how to get ChatGPT to give you the outputs you want by giving it the inputs it needs with our Beginner's Guide to ChatGPT Prompt Engineering.


Photo of Moez Ali
Author
Moez Ali

Data Scientist, Founder & Creator of PyCaret

Topics

Learn topics mentioned in this tutorial!

Course

Introduction to ChatGPT

1 hr
171.2K
Learn how to use ChatGPT. Discover best practices for writing prompts and explore common business use cases for the powerful AI tool.
See DetailsRight Arrow
Start Course
See MoreRight Arrow
Related

From BI to AI with Nick Magnuson, Head of AI at Qlik

RIchie and Nick explore what Qlik offers, including products like Sense and Staige, use cases of generative AI, advice on data privacy and security when using AI, data quality and its effect on the success of AI tools, how data roles are changing, and much more.
Richie Cotton's photo

Richie Cotton

43 min

Getting Started with Claude 3 and the Claude 3 API

Learn about the Claude 3 models, detailed performance benchmarks, and how to access them. Additionally, discover the new Claude 3 Python API for generating text, accessing vision capabilities, and streaming.
Abid Ali Awan's photo

Abid Ali Awan

Fine-Tuning OpenAI's GPT-4: A Step-by-Step Guide

This step-by-step tutorial offers an in-depth exploration of how to harness the full capabilities of GPT-4, enhancing its performance for specialized tasks through fine-tuning.
Moez Ali's photo

Moez Ali

10 min

Hugging Face's Text Generation Inference Toolkit for LLMs - A Game Changer in AI

A comprehensive guide to Hugging Face Text Generation Inference for self-hosting large language models on local devices.
Josep Ferrer's photo

Josep Ferrer

11 min

Exploring Text-Embedding-3-Large: A Comprehensive Guide to the new OpenAI Embeddings

Explore OpenAI's text-embedding-3-large and -small models in our guide to enhancing NLP tasks with cutting-edge AI embeddings for developers and researchers.
Zoumana Keita 's photo

Zoumana Keita

13 min

Hugging Face Image Classification: A Comprehensive Guide With Examples

Master image classification using Hugging Face with a step-by-step guide on training and deploying models in AI and computer vision.
Zoumana Keita 's photo

Zoumana Keita

13 min

See MoreSee More