A Gentle Introduction to the System Prompt

In partnership with

The #1 AI Newsletter for Business Leaders

Join 400,000+ executives and professionals who trust The AI Report for daily, practical AI updates.

Built for business—not engineers—this newsletter delivers expert prompts, real-world use cases, and decision-ready insights.

No hype. No jargon. Just results.

Subscribe free—trusted by leaders

Hello everyone and welcome to my newsletter where I discuss real-world skills needed for the top data jobs. 👏

This week I’m discussing the system prompt on large language models.  👀

I’ve become one of the top authors in the artificial intelligence space. Each month, approximately 200k people read my content. You’ve made this possible and I’d like to thank each and every one of you. Thank you.

Not a subscriber? Join the informed. 

Thank you. 🎉

Introduction

Prompt engineering is an important skill that enables us to interact with large language models and achieve whatever outcomes we specify.

Prompt engineering is the practice of crafting precise prompts to help large language models correctly respond to questions and perform a wide range of tasks. This practice improves the model's ability to produce accurate and relevant responses.

Prompts

There are two types of prompts in large language models.

• System Prompts

• User Prompts

System prompts serve as the foundational instructions that dictate a model’s behavior. They establish the framework for how the model will interact and respond, similar to a job description for an employee.

System Prompt: You are a college professor. You are arrogant and condescending and have little to say that’s consequential. Use words few people understand to impress your students.

User prompts are the specific instructions or questions a user provides to a model to elicit a desired response. These prompts are dynamic and vary with each interaction, reflecting the user's immediate needs and goals. They can range from simple requests for information to complex instructions for generating creative content.

User Prompt: Write a 1500 word essay on the downfall of modern colleges. Please explain why large language models are already better educators than professors.

System prompts define the models overall behavior and role, while user prompts provide specific instructions or questions for a particular task or interaction.

Consider you are briefing a new employee on how to answer customer queries. You might tell them to be polite, provide detailed answers, and avoid using technical jargon. Similarly, a system prompt gives the machine learning model similar guidelines. System prompts can dictate whether the model should be formal or informal, concise or detailed. It can even include specific points the AI should cover in its responses.

Here’s an example of a system prompt you’ve created for a travel planner. Answer questions with detailed information about destinations, tips for travel planning, and recommended activities. Maintain a friendly and informative tone.

This prompt sets clear expectations, helping the model generate responses that are relevant and helpful for someone seeking travel advice.

System prompts are important because they set the stage for the entire interaction. Without them, the model might provide responses that are too generic, off-topic, or not go with the user’s expectations. By using a well-crafted system prompt, users can improve the quality of the models outputs.

Custom Instructions

If you’re using ChatGPT you also have a feature called custom instructions. Custom instructions are a powerful feature in large language models like ChatGPT. It is designed to further refine and personalize the interactions between the user and the model.

While system prompts provide a general framework for the models responses, custom instructions allow users to customize the models behavior and focus even more precisely to meet their specific needs.

The role of custom instructions is particularly important in scenarios where specific outputs are required. For example:

• Educational Content: Teachers can use custom instructions to ensure that the model’s responses suitable for different learning levels, from elementary school to advanced students. 

• Customer Support: Support teams can guide the model to use a compassionate tone and provide step-by-step solutions to common issues, improving customer service.

• Content Creation: Content writers and marketers can direct the model to hold specific guidelines. It ensures consistency across various content creation pieces like blog posts, social media updates, and advertisements.

Implementing system prompt in ChatGPT involves creating clear and effective guidelines that direct the AI’s behavior and responses. Well-crafted system prompts are essential for achieving accurate, relevant, and useful interactions. Here’s a guide on how to create effective system prompts and avoid common pitfalls.

Define your objective

Start by clearly defining what you want to achieve with the mode’s interaction. Whether it’s generating marketing content, or providing educational assistance, having a clear objective helps in crafting a precise system prompt.

Identify Key Elements

Identify the key elements that should be included in the effective prompt. These can be tone, style, specific content points, and any other relevant factors. For example, if the objective is to provide customer support, key elements might include empathy, clarity, and step-by-step instructions.

Be Specific and Clear

Write the prompt in a way that leaves no room for irrelevancy. Specific and clear instructions help the AI understand exactly what is expected. For example, instead of saying Provide good customer service, specify - Use a polite and friendly tone to answer questions about product returns.

Use Examples

Include examples within the prompt to show the desired type of response. This can help the model better understand the context and format of the expected output. For example, Respond like this:

I’m sorry to hear that you’re having an issue with your order. Let me help you with the return process.

Test and Refine

Implement the system prompt and test the models responses. Evaluate if the output meets the objectives and make adjustments as necessary. This iterative process helps in refining the prompt for optimal performance.

Common Pitfalls and How to Avoid Them

Pitfall: Irrelevant instructions lead to inconsistent and unclear model responses.

Solution: Ensure your system prompts are detailed and specific. Avoid general statements and include precise guidelines.

Pitfall: Overly complex or lengthy prompts can confuse the model, leading to errors or irrelevant responses.

Solution: Keep prompts concise and focused on the essential instructions. Break down complex tasks into simpler, manageable parts.

Pitfall: Without context, the model may provide responses that are off-base or irrelevant.

Solution: Provide enough background information and context within the prompt to guide the AI effectively.

Pitfall: Inconsistent language in the prompt can lead to mixed tones or styles in the model responses.

Solution: Use consistent language that matches with the desired tone and style throughout the prompt. This helps maintain a uniform response.

Pitfall: Failing to consider user feedback can result in persistent issues with AI responses.

Solution: Actively seek and get feedback from users to continually improve the effectiveness of your system prompts.

The system prompt and custom instructions play a great role in shaping the future of AI interactions. It offers a pathway to enhanced personalization, efficiency, and user engagement. As technology continues to evolve, adopting advanced strategies will enable organizations to reach the full potential of AI models like ChatGPT.

Thanks for reading and have a great day.

There is only one entry level data role and it’s the data analyst.

I’ve created the courses, study guides and a GPT for preparing for this role. All you need to do is follow directions.

Here is my platform.

STEP 1:

• Take the Transact-SQL course.

• Download the must know interview questions for data roles study guide and fill it in. (Included in the course)

• Download and install the GPT and learn all the top SQL interview questions. A GPT is a model I created that is part of ChatGPT to help you ace the SQL Server interview. (Included in the course)

SQL is the single most important skill for all the data roles. Spend a lot of time here.

STEP 2:

• Learn PowerBI. Take the PowerBI course.

• Study the exam cram guides included in the course for PL-300.

• Use the interview prep sheet for acing the interview. Attain PL-300. That’s the PowerBI certification. I know it works because almost every one of my students has passed the exam on the first try.

You’ll need to learn as many recent version as possible with a focus on the Fabric implementation.

STEP 3:

• Learn Fabric. It’s Microsoft new data centric cloud ecosystem.

• Take my Fabric course and pass DP-600. That’s the Fabric certification. Yes, I’ve included the cheat sheet for that certification also. It’s at the end of the Fabric course. Study these, you’ll pass the exam I promise.

🥳 Here’s a code to receive $20 off the first month. Yep, that means the first month is only $30 dollars. This price includes:

• All the courses

• The study guides

• The GPT

• The exam crams

CODE: MEMDAY25

Take control of your future. Nobody else is going to. If you don’t fail often you are doing something wrong. 👏