Much as you don't generally go auditing the bytecode or intermediate representation generated by your compiler, I think the idea of manually reviewing LLM-written code will fall by the wayside too. Like it or not, these agents are the new compilers, and prompting them is the new programming. Regardless of what happens with any AI bubble, this is just how things will be from now on; we've experienced a permanent, irreversible increase to the level of abstraction. We are all assembly programmers
If you have a fediverse account, you can quote this note from your own instance. Search https://mastodon.social/users/stroughtonsmith/statuses/116030136026775832 on your instance and quote it. (Note that quoting is not supported in Mastodon.)
RE: https://mastodon.social/@stroughtonsmith/116030136026775832
This is one of the worst takes from LLM enthusiasts.
Compilers are deterministic, extremely well tested, made out of incredibly detailed specifications debated for months and properly formalized.
LLMs are random content generators with a whole lot of automatically trained heuristics. They can produce literally anything. Not a single person who built them can predict what the output will be for a given input.
Comparing both is a display of ignorance and dishonesty.