Inputting prompt from scratch

Inputting Prompt From Scratch

ChatGPT allows users to chat with an AI by typing messages or selecting from suggestions. This patterns provide a natural conversational experience like in other chat applications.


Perplexity has difference in the way it the way users inputing the first message and the following ones. In the first message to initiate the conversation, they build similar UX to search or query products where they use suggestions to show similar query, this is a good way to help users form a better initial prompt. After the first message, the UX comes back to traditional conversation app.


Github Copilot is an AI pair programmer that helps developers write code faster. It provides a chat-based UX to let users ask quesion right in VS code, this pattern let users chat with AI right in VS Code without switching to other app, maintaining the flow of coding.

Github Copilot on VS Code

Bing Copilot allow users select the conversation style when inputing the first prompt, this UX enhance the user experience by providing a more personalized experience.

Copilot on Bing

Problem: Beginning interactions with AI without a clear starting point can overwhelm users, making it challenging to initiate meaningful exchanges.

Example: A user approaching a creative writing AI without any prompts may struggle to articulate their needs, resulting in vague or unhelpful responses from the system.

Usage: The "Inputting from Scratch" pattern is crucial for applications requiring user-generated prompts, such as creative writing tools, brainstorming assistants, and problem-solving bots, facilitating a user-friendly way to start interactions.


To effectively implement the "Inputting from Scratch" pattern, several strategies are adopted to simplify the initial interaction phase for users:

  • Guided Prompt Suggestions: Offering users examples or templates of prompts to help kickstart their interaction with the AI.
  • Incremental Prompt Building: Allowing users to build their prompts step by step, with the AI suggesting additions to refine or expand the user's input.
  • Interactive Tutorials: Providing short, interactive guides that demonstrate how to craft effective prompts, enhancing users' understanding and confidence in initiating exchanges.
  • Feedback Loop for Improvement: Incorporating a mechanism for users to receive immediate feedback on their prompts, with suggestions for improvement or clarification.


The "Inputting from Scratch" pattern is designed with the understanding that a clear and accessible starting point is essential for engaging and productive interactions with AI:

  • Facilitates Ease of Use: Reducing barriers to entry for users new to AI interactions, making the technology more accessible and inviting.
  • Enhances User Confidence: By guiding users in creating effective prompts, they feel more in control and assured of the relevance of the AI's responses.
  • Promotes Creative Exploration: Encouraging users to experiment and explore different prompts opens up a broader range of possibilities for AI engagement.
  • Improves Interaction Quality: Well-crafted initial inputs lead to more accurate and satisfying AI responses, enhancing the overall user experience.
Contact us to build your next product or to get consulting on your current project.