Use Case Title:

Dual code your flashcards with custom DALL-E 3 images


This Use Case allows users to quickly create custom dual-coded flashcards in the Anki app using DALL-E's image generation and ChatGPT's FlashcardsGenerator Plugin. It streamlines the process for flashcard users, eliminating the need to search or design images manually.

Tutorial for Use and Best Practices:

Step 1: Use the FlashcardsGenerator plugin of ChatGPT 4 to create a flash card for a specific purpose with an image. Step 2: Download the generated flashcard and import it to the Anki App. Step 3: Generate an appropriate image using DALL-E 3 in ChatGPT 4 either through a user-generated prompt or through a dual code prompt for DALL-E generated by ChatGPT based on the flashcard. Step 4: Make modifications to your prompt until you find an appropriate image. Step 5: Use the image from Step 4 to replace the default image of the flashcard that was imported in the Anki App in Step 2.

Best Practices:

  1. Be specific with your prompts for generating the flashcards and the DALL-E image, making sure to provide details for the custom image for an effective dual code for your flashcard.
  2. Initially, use the default feature of the FlashcardsGenerator plugin to create a corresponding image for the flashcard. You might get an appropriate image for your flashcard and might not have to use a DALL-E-generated image.

Impacts on Learning:

Dual coding is a theory of cognition introduced by Allan Paivio in the 1970s. It posits that humans process and represent information in two distinct ways: verbally and visually. It helps improve the learning experience in the following ways:

  • Multiple Representations: By using both verbal (like text) and visual information, learners can create multiple representations of the same information in their minds. This redundancy increases the chances of recall and understanding.
  • Cognitive Offloading: By presenting information both verbally and visually, learners can offload some cognitive processing to the visual system, reducing the cognitive load on the verbal system and making processing more efficient.
  • Enhanced Memory: Visual images can serve as mnemonic devices, making information more memorable. For instance, a diagram or a relevant picture can help solidify a concept in a learner's memory.
  • Increased Engagement with flashcards: Using text based flashcards can get boring after a while, combining words with visuals can make learning materials more engaging and interesting, which can motivate learners and hold their attention longer.
  • Enhanced Creativity: Using different prompts to generate images in DALL-E can increase the creativity of the user by forcing them to think about something from different perspectives. This also increases learning by the principle of generation effect. Without this Use Case, people who want to make dual coded flash cards will spend considerably more time finding or creating images for the flashcards and also will not be able to use the impact of dual-coding theory on learning.

Limitations and Ethical Considerations:


  • It is not possible to import the flashcards of the FlashcardsGenerator Plugin of ChatGPT to apps or platforms not supported by the PlugIn.
  • The solution is not integrated. We need an flashcard organization app like Anki for its use. Ethical Considerations: Biases: The data AI models are trained on data that often reflects historical and cultural biases. For example, if historically certain groups of people were underrepresented or misrepresented in media, the training data might carry those biases, leading the model to generate images that perpetuate those stereotypes. Using those images for flashcards can perpetuate those biases to the user. Accessibility: The Use Case uses ChatGPT 4 which is paid and hence affects accessibility for people who cannot afford it.

Link to the Guide:


As a student at Minerva University, I have been an avid user of flashcards for learning. Flashcards help with memory retention through spaced practice, an evidenced-based learning technique based on the principle of active recall. After using this learning technique for a long time I have realized that it can be made better if we can create custom dual-coded flashcards. Finding or creating an appropriate image for a flashcard can be a time-consuming process and hence it adds friction to the process of integrating an effective learning principle into flashcards. Text-to-image models like DALL-E offer a solution to overcoming this problem.

What it does

It helps you build custom dual-coded flashcards in the Anki App that can be used for effective learning and memory retention. Students can use it to create flashcards for various purposes like language learning or subject-specific conceptual learning.

How we built it

To make custom dual-coded flashcards we first use the FlashcardsGenerator Plugin in ChatGPT to create flashcards that can be imported to the Anki App and then use DALL-E to create a custom visual image for the flashcard generated by FlashcardsGenerator Plugin and use that image for dual coding the flashcard. The prompt to be used for image generation by DALL-E can be user-generated or we can also use ChapGPT to find an appropriate visual prompt for dual coding the flash card.

Challenges we ran into

  1. We might have to try multiple prompts to generate an appropriate image for a given flashcard.
  2. This is not an integrated solution. We have to first generate the basic flashcard using FlashcardsGenerator Plugin and then use DALL-E to generate an appropriate visual to integrate it with the flashcard in the Anki App.

Accomplishments that we're proud of

  1. Finding a way to make custom dual-coded flashcards for learning.
  2. Making the process of flashcard creation more fun.

What we learned

  1. The use of gap analysis for finding a novel problem and coming up with a solution.
  2. How to combine multiple apps and AI tools to solve a problem.
  3. How to make current learning strategies more effective using science of learning principles.

What's next for Dual code your flashcards with custom DALL-E 3 images

  1. Create an integrated ChatGPT PlugIn or an app to create the flashcard in one place.
  2. Find ways to integrate other learning principles like chunking and associative chaining into flashcards.

Built With

  • anki
  • chatgpt4
  • dall-e
  • flashcardsgenerator
+ 84 more
Share this project: