**Unveiling the Art of AI prompting: part 8 — Procedural order**

In the realm of Conversational AI, ensuring reliable and consistent task completion is paramount. One highly effective approach to achieving this is the use of procedural order—a systematic method that involves breaking down complex tasks into a series of clear and concise steps. This approach allows AI models like ChatGPT to guide users through tasks methodically, reducing ambiguity and increasing task success rates.

To illustrate the practical application of procedural order, let's delve into the concept and its significance in providing a seamless user experience. In this article, we will explore the fundamental principles and advantages of using procedural order in prompting Conversational AI.

**The Role of Procedural Order in Conversational AI**

Conversational AI systems like ChatGPT are designed to engage users in natural and meaningful conversations. However, understanding and accurately responding to user queries or requests, especially for complex tasks, can be challenging. This is where procedural order comes into play.

At its core, procedural order is a structured framework that guides the conversation through a predetermined sequence of steps. It ensures that the AI model and the user are on the same page, following a logical progression towards task completion. Let's break down the key aspects of procedural order in prompting:

**1. Clarity and Precision:**

Procedural order demands that each step be clearly defined and articulated. This precision in language minimizes the risk of misunderstanding and miscommunication. Users can follow along with confidence, knowing exactly what to expect at each stage of the interaction.

**2. Sequence and Flow:**

One of the defining features of procedural order is its emphasis on sequence and flow. It outlines the correct order of actions or information exchange, ensuring that tasks are tackled in a structured and logical manner. This is particularly beneficial for complex tasks with multiple components or dependencies.

**3. User Validation:**

At various junctures, procedural order prompts users for validation or confirmation. This user-centric approach ensures that the AI model respects the user's preferences and choices, promoting a collaborative and user-friendly interaction.

**4. Adaptability:**

Procedural order is not rigid; it accommodates different user responses and scenarios. It anticipates branching paths based on user input, allowing the AI model to respond flexibly to a wide range of user needs.

**Advantages of Procedural Order in Prompting**

The use of procedural order in prompting Conversational AI offers several key advantages:

- **Reduced Ambiguity:**

Clear and concise instructions leave little room for ambiguity, helping users and AI models stay on track.

- **Enhanced User Experience:**

Users appreciate structured, step-by-step guidance, which fosters a sense of competence and engagement.

- **Error Reduction:**

Procedural order minimizes the risk of errors or misunderstandings, leading to more accurate task completion.

- **Adaptability:**

The framework is adaptable, accommodating variations in user responses and requirements.

In essence, procedural order serves as a valuable tool in the Conversational AI toolkit. Whether it's assisting with complex tasks, guiding users through decision-making processes, or providing step-by-step instructions, this approach empowers AI models to deliver a superior user experience. As AI continues to play a vital role in various domains, mastering the art of procedural order in prompting remains crucial for successful and reliable interactions.

**An example prompt with procedural order**

"

Follow these steps in the following order:

1. Initialisation

1.1 ask the user what task he wishes to complete. If the user stated the task or goal move to step 2.

2. gathering the team

2.1 create a list of experts needed for this task and assign fictitious names to each expert for future reference.

2.2 output each expert with name and job title in a listed format

2.3 ask the user if this is the team he wants or has any other suggestions

3. Finetuning

3.1 if the user is content with the team move to step 4. If not move to step 3.2

3.2 if the user answered which expert to add or remove move to step 4. otherwise move to step 3.3

3.3 ask the user what kind of expert he wishes to add or remove.

4. Finalisation

4.1 Display the final set of experts with names in an ordered list, these experts can be addressed directly by name or collectively by team. Tell the user how he can address his team after the list. "

**The prompt in action**

Let’s roleplay through the steps with a fake user who wants to create a website for a used car salesman in Rotterdam.

1. **Initialisation**
1.1. Ask the user what task he wishes to complete. If the user stated the task or goal, move to step 2.

- Assistant: "Hello! How can I assist you today? What task would you like to complete?"

- User: "I want to create a website for a used car salesman in Rotterdam."

2. **Gathering the Team**
2.1. Create a list of experts needed for this task and assign fictitious names to each expert for future reference.

- Assistant: "To create a website for a used car salesman in Rotterdam, I've assembled a team of experts:
- Expert 1: Mark Davis, Web Developer
- Expert 2: Laura Evans, Graphic Designer
- Expert 3: Sarah Johnson, Content Writer"

2.2. Output each expert with name and job title in a listed format.

- Assistant: "Here are your experts:
- Mark Davis, Web Developer
- Laura Evans, Graphic Designer
- Sarah Johnson, Content Writer"

2.3. Ask the user if this is the team he wants or has any other suggestions.

- Assistant: "Is this the team you want for creating the website, or do you have any other suggestions?"

- User: "This team looks great. Proceed."

3. **Finetuning**
3.1. If the user is content with the team, move to step 4.

- Assistant: "Fantastic! Your team is ready to go. What's the next step in creating the website?"

- User: "I'm ready to proceed."

4. **Display the Final Set of Experts**
4.1. Display the final set of experts with names in an ordered list. These experts can be addressed directly by name or collectively by team. Tell the user how he can address his team after the list.

- Assistant: "Your finalized team for creating the website includes:
- Mark Davis, Web Developer
- Laura Evans, Graphic Designer
- Sarah Johnson, Content Writer
You can address them by name or collectively as your website development team. What would you like to discuss or plan for the website?"

- User: "Let's start by discussing the design and layout of the website."

This roleplay demonstrates how ChatGPT can follow the provided steps to help a user initiate the process of creating a website for a used car salesman in Rotterdam.

Conclusion: Leveraging Procedural Order for Enhanced Conversational AI Experiences

In the ever-evolving landscape of Conversational AI, the quest for more reliable, efficient, and user-friendly interactions remains paramount. Throughout this exploration, we've delved into the concept of procedural order and its profound impact on enhancing Conversational AI experiences. We've seen how this structured framework can be applied to guide users through complex tasks effectively and seamlessly.

Procedural order in prompting Conversational AI is more than just a methodical sequence of steps; it represents a paradigm shift in how we approach AI interactions. It offers a structured approach to communication, emphasizing clarity, precision, and adaptability.