Use GPT-4+DALL-E 3 to teach history

Using GPT-4+DALL-E 3 for historical simulation teaching

The author used GPT as a text-interactive learning tool to create a simulation of a “day in the life” of a temple scribe in the city-state of Ur in 2254 BC in his history course.

This simulation teaching uses GPT to generate reasonable historical environments and characters based on original text fragments. Then use DALL-E 3 for image generation to make teaching more immersive.

Main process:

1. Select the historical environment and characters: In this example, the author chose a temple scribe in the city-state of Ur in 2254 BC as the subject of the simulation. This requires some historical understanding of the time and place.

2. Use ChatGPT as an interactive tool: The author uses a large language model (such as GPT-4) as an interactive text generation tool. This means students can explore this historical setting by engaging in dialogue with the model.

3. Generate environments and characters based on original texts: Through large-scale language models, authors are able to generate reasonable historical environments and characters based on original text fragments (such as ancient documents, letters, etc.).

4. Application of multi-modal AI: In addition to text, the author also mentioned the use of multi-modal AI (such as DALL-E 3) to generate simulation-related images. This increases the immersion of the simulation.

Through this simulation, the authors hope to not only give students a better understanding of history, but also inspire them to empathize and understand other cultures and ways of life.

Not only does this increase student engagement, it also provides a more comprehensive and in-depth learning experience.

Detailed: resobscura.substack.com/p/simulating-h

More AI News