Armin Ronacher — longtime Python and Rust developer, creator of Flask and Jinja — pushes back on the “programming is solved” meme with a half-agreement: the part that’s solved is learning a language. For an agent, picking up Rust or Zig costs almost nothing, which makes language choice “much less consequential than it used to be.” So people increasingly choose languages based on marketing, and two vibe shifts are driving that choice toward fast, hard software.

What he’s seeing in the wild:

  • LLM-assisted projects pick “hard languages.” Cloudflare’s new Artifacts service uses a pure-Zig Git-protocol engine compiled to a ~100 KB WebAssembly module; Vercel’s fx is a Zig coding agent marketed on being small and fast. Both are, as far as he can tell, largely LLM-assisted.
  • The fast-software crowd leads the way. People like Mitchell Hashimoto, Charlie Marsh, Jarred Sumner, and Daniel Lemire carried the obsession with performance — and they’re all receptive to agents writing code. Tools like autoresearch mean you don’t need to know every optimization trick yourself: “you just need to put an agent on it — though knowledge greatly helps!”
  • Harder technologies are opening up. He’s suddenly seeing people do impressive work with DWARF debug data, eBPF, custom network drivers, custom crypto, and old computing hardware — things that were previously off-limits, and in crypto’s case, actively gatekept by the people in the know.

The bet underneath all of it: LLMs change the economics of expertise. If agents absorb the cost of learning and the cost of optimizing, then the barrier to “hard” engineering collapses — and developers who want things fast and small can actually go build them. His closing line: “maybe the world will have more slop, but it might also have more developers in it, that want things to be fast and small.”

Why it matters: this is the flip side of the “AI makes everything generic” story. The same technology that floods the web with mediocre output also hands serious engineers tools to reach into territory that used to take a decade of specialized study.