Introduction
Communicating with AI is one of the most frequent challenges you will run into during your journey of getting the tool to be helpful. In this exercise you’ll practice providing small incremental change prompts to reach a result that you want, rather than a single large complex prompt.
Process
- Think of what sort of avatar or icon you want to use to represent you that you feel can be made with simple shapes. (Do not use any of the Pacman examples shown in class.)
- Identify individual parts of the avatar and think about how you would describe their properties (shape, location, color) in words.
- Using our class’ custom interface, ask ChatGPT to code a P5.js image based on your description. Use the online p5.js editor to test code that ChatGPT gives you back.
An example starting prompt:📍 Acting as an expert programmer familiar with the P5js JavaScript framework, create a mouse face that could be used as an avatar icon. - Evaluate the output and refine the starting prompt until you feel comfortable with it as a starting point.
- Start requesting small incremental changes to the image, refining the prompt if it’s interpreted poorly or if the result is not what you expected.
- Add a post with a title like Task 8: Mary Gonzalez and add the category “Task 8”.
- To the Visual tab add:
- The starting prompt and the image it created (as a screenshot).
- Two intermediate steps providing the image before your prompt, the prompt, and the image after your prompt. A total of 4 images and 2 prompts.
- The final image and the final prompt.
- Complete the feedback survey to get credit for this task.
Submission Checklist
- On WordPress:
- Starting P5.js image
- Starting prompt
- Two copies of the following steps:
- Image before incremental change prompt
- Incremental change prompt
- Image after incremental change prompt
- Ending P5.js image
- Final prompt that led to final image
- Complete the feedback survey.