Comments about the News & Views Article in Nature: Large language models help computer programs to evolve

Following is a discussion about this article in Nature Vol 625 17 January 2024, by Jean-Baptiste Mouret.
To study the full text select this link: https://www.nature.com/articles/d41586-023-03998-0 In the last paragraph I explain my own opinion.

Reflection

Introduction

Romera-Paredes and colleagues’ work is the latest step in a long line of research that attempts to create programs automatically by taking inspiration from biological evolution, a field called genetic programming
The key question in genetic programming is how to represent programs so that they can be modified easily, but meaningfully, by random variation. In other words, what is the ‘DNA’ of a computer program.
General speaking to modify any program, written in Assembler language (machine code) or high level language is very tricky. You have to understand the purpose of the program.
For instance, adding random letters to a program etc. which means that the vast majority of modified programs cannot be executed by the computer, and are therefore useless.
Correct.
To approach this problem, genetic‑programming researchers have taken inspiration from compilers, which are programs that transform text written in a programming language into code that a computer can interpret.
All programs written in a high level language require in some form or another a compiler to translate the high level language in machine code. Remember it is possible that your compiler is build in directly in what is called the firmware of your PC, but functionaly there is still one.
Compilers represent programs in the computer’s memory with an abstract ‘tree’ structure.
Generally speaking a compiler translates line by line of the source (in a HLL) in machine code. Exceptions are possible.