Use AI-generated simulations where ready-made tools are unavailable, for edge cases, or as comparative learning tools¶

Strategy · All strategies
Evidence · no claims cited

Description¶

The article recommends deploying custom LLM-generated simulations alongside, not instead of, established tools: the approach "may be beneficial in cases where ready-made simulations are unavailable, when exploring edge cases not supported by existing tools, or as a comparative learning tool". Combining professional and custom simulations gives educators and students a broader range of possibilities.

Design Implications¶

Context¶

Requirements¶

• Access to LLMs and the prompt-based workflow; awareness of existing simulation libraries to identify gaps

Constraints¶

• The authors explicitly state the approach is not intended to replace well-established, robust simulations like PhET

Target Learners¶

• physics educators
• students

Target Learning Goals¶

• extending access to tailored interactive learning tools
• comparing custom and professional simulations

Related Strategies¶

• Develop complex simulations incrementally, testing each stage before adding the next feature

Examples¶

-

Key Sources¶

• Yossi Ben-Zion, Roi Einhorn Zarzecki, Joshua Glazer, Noah D. Finkelstein. (2024). Leveraging AI for Rapid Generation of Physics Simulations in Education: Building Your Own Virtual Lab. https://arxiv.org/abs/2412.07482