If you tell it to write a spec -> then write the tests -> then implement, the LLM should be able to pretty much one-shot a compiler frontend. LLMs really benefit from the kind of task that has a built-in validation loop.
I'm working on something similar, but unlike the author, my progress has been pretty slow. It's tough. I do write about a fifth of the code myself, but I keep getting stuck on the rest.
GPT/Fable/what have you is not building this language, as its laid out, and "working" in a true sense that fast.
Either OP is completely full of it, the language didn't actually work (and likely still doesn't), or the language is far less sophisticated than it seems from the examples - it's and the examples are minimal, so it's kind of hard to tell what it actually does...