I'm not disputing that some people like user-controlled operator overloading (Zig does have + etc. for vectors), but clearly others prefer no user-controlled overloading of any kind (don't know what the LLMs prefer, if they have any preferences at all). In the safety-critical space where I once worked, where people could get hurt or die if there's a stack overflow, the more common preference was against (user-controlled) overloads or any implicitness. That's why we need options.

The example you choose is rather telling, though. It indicates that the use case you have in mind is not at all the one that Zig is targeting.

A more humble critique would be to just acknowledge that you don’t seem to be a member of Zig’s target audience, and move on.

You can do a+b in Zig with SIMD vectors. No operator overloading required