I built a programming language to test how well LLM-edited code survives

1. 1Design a statically typed programming language to test LLM code edits
2. 2Build a compiler and runtime environment for the language
3. 3Test LLM-edited code snippets for syntax and semantic errors
4. 4Compare the results with human-written code
5. 5Refine the language design based on the findings

Software engineers and AI researchers can benefit from understanding how LLM-edited code can be improved and validated

💡 A custom programming language can be designed to test and improve the survival of LLM-edited code

🤖 Can programming languages be designed for LLM code edits? 🤔