If ChatScript C++ may generate LLVM bitcode, then I find what Mohammad is proposing to be intriguing. The fact that Mohammad opened a dialog was considerate in manner. And it is a smart idea, that may be tried with any open source C++ chatbot, that can generate LLVM.
For the sake of conversation only, with no legal advice intended, suggested or implied. Based on other famous systems that have been source to source compiled successfully through Emscripten, I believe all credit will naturally stick like glue to the author of the original source.
For example, I believe Emscripten reuses the same data files with little or no change. It just targets only the Low Level Virtual Machine code, for lack of a better term, and sort of redirects it from C++ to JavaScript. Umm, in easy language, you can say, “Compiler, spit me out some LLVM please.” So, it’s like everyone knows the original source was not JavaScript, right? That is why they call it a source to a source compiler.
Lastly, listen closely—these comments are all very friendly toned… I am NOT, repeat NOT criticising or trying to correct Dave Morton’s opinion. I am just trying to discuss how running open source code through different compilers, needn’t be any worry to the original author.