Comment by znnajdla

Instead of empty theorizing, we should have benchmarks for this. There is at least one benchmark which suggests that LLMs are better at writing Elixir than most other languages: see the AutoCodeBenchmark.