GenAI coding assistants are only as good as the data they are trained on. Less-used proglangs make up a tiny fraction of the available data, or may even be completely absent. There is a reason coding assistants give convincing results with Python and JS/TS, but underperform even on relatively up-and-coming langs like Rust.
I know devs writing in it making over 200k per year.
Ai isn’t that useful unless you can correct for it’s mistakes, which requires some experience with the language.
Probably no longer possible now that we have generative AI, a coder can now be archived alongside the codebase itself.
GenAI coding assistants are only as good as the data they are trained on. Less-used proglangs make up a tiny fraction of the available data, or may even be completely absent. There is a reason coding assistants give convincing results with Python and JS/TS, but underperform even on relatively up-and-coming langs like Rust.
I know devs writing in it making over 200k per year. Ai isn’t that useful unless you can correct for it’s mistakes, which requires some experience with the language.
Maybe in another 10 years.
The reason banking still uses Cobol is the same reason they would never trust AI