If your LLM can output 10-100x the LOC output, and it's equally good at all languages, and you're not bound to an existing language choice or ecosystem, why not choose Rust?

Rust code will be faster, safer, and easier to ameliorate bugs in.

Rust seems like the best language to serialize business logic to now that LLMs are so good at it.

If the LLM makes a mistake in Javascript or Python, you literally won't know until runtime. With Rust, you'll know immediately and have really good compiler recommendations for fixes.

I think Rust is the best LLM language. I am somewhat biased: I've written Rust code for ten years, and I'm having a blast with Claude Code writing it for me instead now. But I've also used so many other tools and languages - enough to say that Rust has some unique advantages here. And also that Claude does a fantastic job emitting Rust.

LLMs emitting Python feels like building with clay. LLMs emitting Rust feels like building well-engineered steel skyscrapers.

▲

resonious 4 hours ago | parent | next [-]

I'm also having a really good time having LLMs write code in Rust. In Typescript they tend to abuse `any` or just cast stuff around, bypassing the type system at every corner. In Rust they seem to take compiler errors to heart and things tend to work well.

	▲	maxbond 4 hours ago \| parent [-]
		You might also have success asking your agent to run `eslint` at the end of every subtask and instruct it to always address lint errors, even if they are preexisting. I agree with your diagnosis; there's "implicit prompting" in Rust being more strongly typed and the agent "knowing" that going in but we can compensate with explicit prompting. (You do also have to tell it to actually fix the lints, I find, or it will conclude they aren't relevant.)

▲

maxbond 5 hours ago | parent | prev [-]

I do choose Rust. For now. I write Rust everyday. I'm generating Rust at this moment.

But when I learn a better language I will adopt it.