[desmondwillow]

I'm quite surprised to hear that "programming languages have reached a terminal state": there are (imo) at least four in-progress movements in the industry right now:

1. Memory safety 2. Better metaprogramming capabilities 3. Algebraic effects 4. Solver/prover awareness

Even if LLMs become capable of writing all code, I think there's a good chance that we'd want those LLMs writing code in a language with memory safety and one amenable to some sort of verification.

[hippo22]

I didn’t quite mean that programming languages have reached their terminal state, although I understand how my comment was interpreted like that. I meant that programming languages, as we known them today, have reached a terminal state.

Using Rust as an example: Rust aims to provide memory safety at the expensive of developer ergonomics. Personally, I shy away from Rust because I don’t like fighting the borrow checker.

However, with AI agents, Rust could make a lot more sense. Strict errors at compile time are helpful to an agent, which is more than happy to smash its head against the wall until it reaches a working solution.

Following this logic, we could see languages develop that are extremely impractical for humans to use yet provide benefits like memory safety or correctness. But these languages might not look anything like the languages we’re currently used to.