@pozo70 To me, a programming language is just a tool, what matters is how well it works with AI in my workflow, regardless of whether the tool itself was built by humans or AI, using Zig with AI and deciding what contributions to accept into Zig itself are two separate questions
Not automatically. Zig’s testing allocator only tracks allocations made through it. If a C library supports custom allocation callbacks, you can hook those up to it. If the library calls malloc internally, you need something like LeakSanitizer to check for leaks
My experience is that AI changes how I build confidence in code. When I’m no longer reviewing every line myself, I put much more emphasis on having AI write extensive tests, including edge cases and failure paths, and repeatedly run and refine them. Combined with Zig’s testing allocator and the right checks for C dependencies, that gives me a practical way to catch many memory issues before they reach production
I’ve also found that a capable AI working through a solid test and fix loop can produce considerably more consistent and thoroughly tested code than a typical manual workflow. I’d expect substantially fewer memory bugs to survive that process. Tests can still miss bugs, of course, but that’s why iteration speed matters so much to me: the faster the feedback, the more opportunities the agent has to test, find problems, and fix them. That combination makes Zig very attractive to me for AI driven development
Introducing Omarchy Pocket. Real Omarchy Linux on a $99 dual-screen handheld.
We rebuilt the Wayland GUI with PocketJS Applets, made it fluid on hardware that can’t even run Android, and rethought how Linux should work across two screens.