Are hallucinations in code generation still a problem? I thought with linters, type checkers, and compilers especially as strict as Rust, LLM agents easily catch their own mistakes. At least that's my experience: the agent writes code, runs linters and compilers, fixes whatever it hallucinated, and I probably get a working solution. I tell it to write unit tests and integration tests and it catches even more of its own mistakes. Not saying that it will always produce code free of bugs, but hallucinations haven't been an issue for me anymore.

> With AI lifting legacy code bases into Rust got a whole lot easier, and purging the blight of C from the world, excepting the most deeply embedded of applications, got a whole lot closer.

You probably don't realise this, but AI written Rust (or any language) will have much more undefined behaviour than human-written C.

AI coding brings non-determinism to every language; with AI, now every language can have Undefined Behaviour.

Rust has a very strict type system and an ecosystem that often utilizes the type system well.

Many things that would only be caught at runtime in other languages are caught at compile time in Rust, making coding agents iterate until things compile and work well.

Rust also has great error messages, which help the agents in fixing compilation errors.

The compile errors are great. I can change one function signature and have my output fill up with compile errors (that would all be runtime errors in python). Then I just let claude cook on fixing them. Any time you have to run your program and tell claude what’s wrong with it you’re wasting time, but because claude can run the compiler itself and iterate it’s much more able to complete a task without intervention.

I think relative to the typical Rust code it likely does worse than AI relative to the typical Python code. But due to the compiler, it's possible you might get more correctness out of AI-generated rust code on average.

I can't give you examples, but my experience is that AI does very well with Rust except for cases where a library has a constantly changing API/ has had recent breaking changes. I find that AI does extremely well at "picking up" a Rust codebase, I suspect due to the type information providing context but I couldn't say.

I think the argument is more that working rust code is better than working Python, and AI assistance makes it more tenable for average developers to successfully produce working rust code, and in particular is helpful for navigating the gap between "code written" and "code compiling" (eg why is the borrow checked mad at me).

Even if it writes the same or even somewhat worse rust than python, assuming the output is the same you are likely to get a speedup + a better distribution story.